DVDshrink quality compared to CCE? [Archive]

0xdeadbeef

19th June 2003, 00:09

If CCE would get the uncompressed stream as input, it would of course produce always better or at least equal results compared to a good transcoder.
Then again, compressing an image which was already compressed by removing the higher spatial frequencies is a less ideal scenario. Removing the higher coefficients of the discrete cosinus transformation used by JPEG/MPEG creates the typical "ghost image" artifacts on sharp edges. These are caused by the missing higher coefficients which would normally extinguish the oscillation.
When recompressing this, the encoder will try to reproduce the edge including the artifacts. When transcoding, just some more coefficients are deleted. It is obvious that the latter approach not only creates less data but also a higher image quality.
This works fine for medium compressin ratios. However at some point only removing coefficients will create too many block artefacts. Then it would me wiser to recalculate the motion vetors and/or insert/delete I-frames at the right position and so on. That's were an encoder wins over a transcoder.
But for compression ratios of about 30-40% or less, I would bet on the transcoder. Please note that some of the current one click transcoders use pretty simple algorithms.

Zhnujm

19th June 2003, 00:22

Then all my test encodes are obviously above this limit :D
but i think the ct would have come to another conclusion if they had used their eyes instead of this comparing program. ;)

thop

19th June 2003, 06:09

The problem with CCE is that it is so slooooow compared to DVDShrink ... i can't really waste my computer ressources for a few hours (not to mention the power bill, an often neglected thing). If it was faster (<1h) i'd use it. Maybe with 10Ghz CPUs i will reconsider.

0xdeadbeef

19th June 2003, 10:53

@Zhnujm:
I think it's pretty hard to judge compression results of a whole movie if the quality between two outputs is not too far part.
Even if you make screnshots and compare them side by side, it's sometimes hard to tell which is better since one lacks here and the other there. Furthermore, this will usually be a comparison of I-frames which doesn't tell about the P-frames inbetween. It won't tell about correct motion vectors either or about the distribution of bitrate above a longer time.

Surely it's also a question of personal preferences. Current transcoders tend to cause macro blocking. Reencoding using CCE (or whatever) tends to cause blurriness and edge artifacts (noise around hard edges).

Fmazzanti

19th June 2003, 16:10

I've backed up quite a few movies and I definitely stay with CCE. And yes, quality issues are apparent when comparing CCE and IC7/DVDShrink results. If the movie is short such that you can use a low compression level in IC7, then the corresponding CCE bitrate is also high and the copy you get using CCE is almost identical to the original one. If the movie is so long that you must use a high compression level in IC7/DVDShrink, then you see blocks in the transcoded version and CCE wins again. In any case CCE wins :p
And concerning tweaking CCe, just use DoCCE4U and RoBa method to fix encoding parameters...

Slug123

19th June 2003, 16:41

Wow, lot of answers to a simple question.. I'll have a play with DVD-shrink and a few others out of interest BUT for serious quality needs will stick with CCE.

Thank you everybody..

Slug

0xdeadbeef

19th June 2003, 17:17

@Fmazzanti:
Would be nice if it were all that easy and yet, it's not.

Fmazzanti

20th June 2003, 12:14

Why not? Maybe there are specific cases when this may fail, but as a rule of thumb, my own experience says it is... of course, that's only my experience :D

ChickenMan

20th June 2003, 14:49

My experiances have been much the same as per Fmazzanti. While the one-click wonders are improving all the time, they are still way behind the quality of CCE IMHO.

As for conversion times, gee I re-encode most movies in 2.5 to 3 hr, with 20min to author, then ready to burn. IC7 is just as slow or as fast depending on which side of the fence you sit. Compared to DVD2ONE its slow, but with poorer quality results. You get what you pay for :)

0xdeadbeef

20th June 2003, 15:18

Depending on the number of passes the reencoding using CCE alone will take several times the time IC7 needs to convert the whole thing. So usually there's is a BIG time difference between IC7 and CCE conversion. Not talking of all the additional issues like demultiplexing and multiplexing which also take their time.

And then - though I dislike to repeat myself - it's simply not true, that reencoding always creates better quality. As always in life, it depends.

If the reduction needed is very low (like <15%), you can go with any transcoder as the visual differences will be neglectable. If it's more like 30% or above, IC7 is the way to go. For something nearing 50% or above, reencoding is probably the best solution.
You can surely discuss the percentages and they will vary from movie to movie. But to use CCE just in any case, because it's the most expensive, professional and time consuming method, is not sensible.

thop

20th June 2003, 15:27

I re-encode most movies in 2.5 to 3 hr1 pass? don't forget demuxing and multiplexing also needs a fair amount of time. CCE is just too slow.

mpucoder

20th June 2003, 17:06

@Fmazzanti I'll bet you work with PAL. It's a totally different story with NTSC.

Fmazzanti

20th June 2003, 20:55

Yes you're right. I'm a PAL man...

ChickenMan

24th June 2003, 14:18

Originally posted by thop
1 pass? don't forget demuxing and multiplexing also needs a fair amount of time. CCE is just too slow.
No, 2 pass and I dont do any demuxing or multiplexing, just authoring as I said. I use DVD2SVCD to do it all for me, it all really so darn simple and I do PAL mostly. The few NTSC dvd's I have were a pain to backup, they were 30-40% slower.

CM

monomer

24th June 2003, 16:48

Originally posted by 0xdeadbeef
Depending on the number of passes the reencoding using CCE alone will take several times the time IC7 needs to convert the whole thing. So usually there's is a BIG time difference between IC7 and CCE conversion...

True unless you accept the fact that IC7 doesn't do sizing very well and you subscribe to the belief that the closer you can get to using all the 4.37Gb available on a DVD the better the picture quality will be. IC7 often gave me file sizes between 3.2Gb and 3.8Gb the first time around... that's a LOT of wasted space. If fact it usually took me 2, 3, 4 and even sometimes 5 tries to get it in the 4.1Gb to 4.37Gb range. Calculate the time on that scenario! Under these conditions I would says it's close to dead even between using IC7 and CCE... 'time-wise' that is.

Originally posted by 0xdeadbeef
And then - though I dislike to repeat myself - it's simply not true, that reencoding always creates better quality. As always in life, it depends.

If the reduction needed is very low (like <15%), you can go with any transcoder as the visual differences will be neglectable. If it's more like 30% or above, IC7 is the way to go. For something nearing 50% or above, reencoding is probably the best solution.
You can surely discuss the percentages and they will vary from movie to movie. But to use CCE just in any case, because it's the most expensive, professional and time consuming method, is not sensible.

Again true... except instead of using IC7 between 30%-50% I just use CCE (via DVD2DVD-R and Scenarist) and above 50% I just do a 1:1 split across 2 discs. (Why use CCE over IC7? see the answer above)

BTW, I can see none of these 'edge' artifacts spoken of. And I'm viewing on a ~114" wide screen with a front projector feed by my computer set-up for a 1:1 pixel map. However I will go back and look very closely to see if I can discern any 'ghost' images. But if I understood you correctly, the problem occurs if the movie that's being encoded had already been 'stepped on' once (previously compressed)? If that's so, then the answer would be to always re-code from the original whenever possible.

Fmazzanti

24th June 2003, 17:25

Originally posted by 0xdeadbeef
And then - though I dislike to repeat myself - it's simply not true, that reencoding always creates better quality. As always in life, it depends.

If the reduction needed is very low (like <15%), you can go with any transcoder as the visual differences will be neglectable. If it's more like 30% or above, IC7 is the way to go. For something nearing 50% or above, reencoding is probably the best solution.
You can surely discuss the percentages and they will vary from movie to movie. But to use CCE just in any case, because it's the most expensive, professional and time consuming method, is not sensible.

Uhmm... I'd like you to post that in other forums in this site, to see what others have to say. Your statement is very nice, but I'd like to see what you have there to back them up. At least in the PAL case (which is the one I work with always) I can tell you that there are significant differences between transcoded/encoded material with compression rates of 30%. At compression rates <15% differences are minor, but still visible after close inspection.
maybe NTSC is different? Can be... NTSC is poorer in any aspect compared to PAL, so maybe you guys are accostumed to less quality than we europeans (and remember I wrote 'maybe' :D )...

0xdeadbeef

24th June 2003, 19:35

@monomer:
Of course, every movie on DVD is already compressed. Compressing it again will enhance noise and create new artifacts. Transcoding - if done properly and not pushed to far - will only enhance artifacts already there. It might even reduce noise. Both will reduce detail to some degree.
About IC7 target size: my tests with the newest version always ended very close to 4.3GB and took about an hour to transcode. Of course, the variable compression rate makes it pretty complicated to predict the target size compared to fixed compression trancoders as DVDShrink. Maybe they should introduce an optional multipass option. But up to now, it worked for me without it.

@Fmazzanti:
I don't feel like doing hours of CCE transcoding when it's hot like this. But I could offer a transcoder test I did some days ago.

Beware, the picture is about 900kb:

http://home.nexgo.de/0xdeadbeef/brazil_comparison.bmp
(better download it and watch it in 2x magnification)

This is an example from the PAL (!) movie Brazil. I transcoded only the main movie and removed the Spanish sound track. The resulting compression factor was about 32% (based on the original size).

As you can see, the inferior transcoders create block artifacts. IC7 however maintains image quality. Sure there are also some differences, but I'd bet you wouldn't recognize them without the original next to it. Not to talk of an usual 32" tube.
I don't think CCE would create better results. Surely not FAR better results. Prove me wrong.

Fmazzanti

24th June 2003, 19:53

I'll take a look at the picture tomorrow morning where I can download it at work (today i sholiday here in Spain :D :D ) but sure you don't have to convince me that a compression rate <15% is gonna give good results... my point is that CCE will do it equal or better (that's a personal impression based on experience which is of course far from being a scientific argument). However I must admit that IC7 beats the hell out of ALL other transcoders in quality. I'd like to use it more often, but there are several drawbacks:
1- Sizing issues. You say with latest release you've ended quite close to the 4.3Gb limit. You're a lucky man. I also use the latest release and still get 4.0-4.1Gb (this morning I've backed up a movie setting the bar to 4.83Gb and ended with 4.24Gb, and this was the second trial!). I bet I'll have to go for a third one... in the end this takes as much as a good CCE encoding. I'd be happy to know you 'sizing secret' that let's you reach the 4.3Gb limit at once :)
2- Menues with IC7 are usually screwed up when seen on my Marantz player. Other users whi own Phillips/Toshiba players have reported similar issues in this forum. And this happens even if I do not compress them.
Anyway, I understand that one thing is the theroretical discussion of transcoders vs. CCE and another the issues IC7 presents... different stories I must admit.

0xdeadbeef

24th June 2003, 20:05

@Fmazzanti:
1)In this very example, I'm talking about 32%, not 15%.
2)I'm aware hitting the target size might be a problem with IC7.
I would however recommend to use the IC7 Setting editor (http://www.mrbass.org/instantcopy/) to increase the size of the video to be analyzed. This should also increase accuracy of size prediction.
3)I agree there a some bugs in IC7. Hope they will be fixed soon. There seems to be a workaround for the menu bug though.
4)I will still remove the Spanish sound track :)

I would also prefer if IC7 was just a transcoder (same for CloneDVD) since having three or more burning drivers on my PC makes me nervous.

ulfschack

25th June 2003, 16:54

I just gotta butt in here cause I simply love it when people go against the stream and can suport it with good argumnets too :)

[b]Oxdeadbeef[b/], didn't you know that it's almost against forum rules to claim that CCE isn't the best solution for everything!:devil:

I've made similar discoveries and found clear instances where transcoders beat encoders (yes even CCE). What I find strange is peoples unwillingness to accept this. (Can you say group preassure).

We might not have it yet but certainly theoretically it should be obvious that transcoders could do a better job than any encoder since the main effort in finding quantifier and motion vector values is allready done!

The extra work that is done by an encoder as compared to a trancoder is roughly: IDCT to RGB/YUV to DCT. The last step is hardly bijective (1:1), in fact this is the very heart of the encoding and the main reason for encoders having a difference in quality (and price). It can in short be done in an astronomical amount of ways and only one is optimal (if you think CCE can do a better encoding than the uncompressed stream that is fed to it...think again).

Now I admit we could argue that there might exist a movie that has such crummy motion detection but the bitrate is so high that the picture looks good anyhow. Then if we were to compress it to lets say (for arguments sake) 30% of the original size CCE would make it look almost as good due to a blatant waste of bits in the original, whereas a transcoder leaves the MV alone and only operates on reducing frequency coefficients by 70% ending up wiht crap. Do we think that this occurs often? If you answer no to this you're voting for the transcoder.

Another argument would be: Would the authoring house have chosen a completely different way of encoding if they had (most often a little) less bits to play with? If you answer no to this you're again voting for the transcoder.

Then again trancoders might have in the future (or allready do have) functionalities to recalculate MVs and new differences for B and P frames based on the new bitrate... who knows? The line of encoders and transcoders then starts to get fuzzy. But I don't care, cause the transcoders still have the advantage of knowing every little parameter from the original movie and should in theory be able to outperform any encoder. Can anyone really object to this?

All respect to [b]Oxdeadbeef[b/] for standing his ground on this. :cool:

cheers

mrbass

25th June 2003, 19:22

Originally posted by ulfschack

I've made similar discoveries and found clear instances where transcoders beat encoders (yes even CCE). What I find strange is peoples unwillingness to accept this. (Can you say group preassure).

http://forum.doom9.org/showthread.php?&threadid=55844
read the 6th post by mpucoder....he says something to this effect that CCE doesn't always win. I guess CCE wins only on longer movies so I retract my statement "But honestly nothing beats CCE ". I'm not sure exactly at what point since they're so many variables.

Zhnujm

25th June 2003, 22:19

@0xdeadbeef:
I wouldnt even think about using CCE or TMPEG if my test encodes looked like the picture you have posted. But they looked far worse.
Maybe T2 Special Edition and Conan the Barbarian are not so good test movies.
I try if i can post some pics the next days.

btw, i think TMPEG is better than CCE :D

0xdeadbeef

25th June 2003, 22:34

@ulfschack:
Finally some support ;)

@Zhnujm:
I don't claim IC7 is a perfect transcoder, it's just the best one I know at the moment. Maybe it fails for some movies even with percentages it should be able to handle. The success of IC7 will surely depend on the encoding strategy of the original mgp2 encoder. If it uses a strange frame order or whatever, IC7 might fail.
But I'm sure IC7 will evolve and other transcoders will come up or mature.
Maybe I should try T2 myself...

Fmazzanti

26th June 2003, 09:02

Well as a physicist I do believe theories are worth a damn if they disagree with experimental measurements. And that's what I think it happens in this comparison of transcoders vs. CCE. As you see, people here only dare to defend IC7, since others (DVDShrink, CloneDVD... which are transcoders also) deliver a quality that is under comparison levels, let's say. So my conclussion is still the same:
maybe a transcode strategy can be better from a theoretical point of view in certain cases... Maybe... have I seen a single case where IC7 delivers better quality than CCE? NO. Maybe I can find cases where the original, the CCE encoded and the IC7 transcoded projects are almost identical, but I would like to see a single case where IC7 performs manifestly better than CCE. And I'm not talking about a very rare case specially choosen for the proof, I'm talking about common situations one encounters often when backing up movies...
I insist, you can mess up whatever you want with Fourier coefficients and filters and similar things, but still, I talk about actual programs and their result, not about theories.
BTW, although you may find it strange, I also use IC7 from time to time, because I also believe IC performs great... only that CCE does better :D You may say CCE takes loooong to finish, and then I'll argue you have to do every project two or three times with IC7 to make it fill a DVD-R and in the end that's pretty much the same time...

monomer

26th June 2003, 16:48

I too must say that even though it may be theoretically possible for a transcoder to deliver better picture qualilty than CCE, I personally have not experienced such. For the past couple of weeks or so I find myself encoding almost any movie above ~5.3Gb with CCE. So the only movies I've compressed with both CCE and a transcoder (DVD21 or IC7 or DVDShrink) were a while ago. I had several very large movies (Grey Zone, Mulholland Drive, Apocalypse Now Redux, etc.) that all came out looking REALLY BAD!!! using the transcoders. If you have any doubts, I can give you a full listing of the movies I transcoded and you can try one of them for yourself. These movies were so granulated and pixellated (DVD21 added 'pulsing focus' artifacts as well) that there's really no discussion... these movies were just unwatchable. You didn't have to squint or pause the movie or frame advance or zoom-in or do a 'side-by-side comparison' to the original to see these artifacts. I mean the picture quality from the transcoders were REALLY BAD!!! Then I 'discovered' CCE and there was NO COMPARISON... NOT EVEN CLOSE. This is why I can't understand what Oxdeadbeef and ulfschack are talking about unless... they are referring to movies requiring a relatively small amount compression (say anything below ~6GB). (Of course there may be occassional exceptions depending on bitrate so I'm speaking of these limits 'in general' and not as being exact absolute limitations.)

Confession: My real life job has nothing to do with video encoding, programming, or even computers in general. This is all just a hobby for me and as such I come by my knowledge by participating in a number of forums and by actual trial-and-error experience. It would be great if others would also honestly state their source(s) of knowledge as well. I am always eager to learn more but I also know not to blindly trust that which I cannot see and reproduce myself. So, can someone suggest a movie greater than 5.5Gb that I can do for myself in which I will see better picture quality by transcoding. (I don't actually trust 'still shots' of movies delivered over the web for a mirade of reasons... but that would be a whole other post.)

Again as I've stated before... this is all just a hobby for me and I'm eager to learn more... this is not a 'challenge' but rather an honest request... please someone give me a movie to try and I will post back with my honest impressions in a few days. (BTW, the only transcoders I possess are DVD21, IC7, and DVDShrink... I have all the latest versions and up-dates.) Thanks to anyone that responds.

0xdeadbeef

26th June 2003, 17:32

@Fmazzanti:
As one of my professors at university once said: "science is not only paying attention to correlation but also to causality.".
[Though offtopic I might add a quote of my physics professor as well: "some people seem to be surprised that a cat has holes in her fur where the eyes are." :D ]

@monomer:
To quote myself from some postings above:
If it's more like 30% or above, IC7 is the way to go. For something nearing 50% or above, reencoding is probably the best solution.
6GB needs only about 28% compression to get to 4.3GB, so that is not near the limit, where IC7 usually starts to create bad quality. My example above was 32% or about 6,4GB. I think even 40% should be ok, assumed the main movie isn't to long (about 90 minutes). Then again, if it just has an immense bitrate, then IC7 might be able to succeed beyond 40%. Dunno, never tried this. If it is poorly encoded or extremely long, than 30% might be too much already.

Indeed, nearly all of the (PAL) DVDs I tried have a playtime of about 90 minutes (+- some minutes) and a main movie size of about 5-7 GB. I even encountered some below 4.3GB.
So in this range <40% IC7 usually creates great quality (as my example shows). No need for CCE here as far as I'm concerned.

BTW: nobody commented on my example picture. But just ignoring it doesn't help ;)

monomer

26th June 2003, 20:18

I apologize for not having made my request as crystal clear as I generally aspire to. I believe the discussion centers around transcoders actually producing better picture quality than CCE on some occassions. I assume someone has actually experienced this and thus is speaking from 'first-hand' experience. I would like to replicate the experience on my home theater computer and then have an informed opinion. Could the person(s) who has experienced this 'first-hand' please announce the title of the movie in which the transcoded movie was better (ie. fewer compression artifacts produced) than the CCE re-encoded version... as I wish (and others may as well also) to duplicate your experience. Thank you to any who respond.
BTW, I'm from the land of NTSC so any PAL titles will not be possible for me... however someone else in PAL-land might want to try and replicate and then corroborate your experience so go ahead and respond anyway.
Again, I am asking for the title of a movie in which the transcoded version has fewer artifacts than the CCE re-encoded version.

Fmazzanti

26th June 2003, 20:57

@Oxdeadbeef,
I'm also a physiscs professor at the university but think quoting myself would be too pedantic :D :D
Now back to the stuff. In short, the pictures you posted show clearly that only IC7 does a good job. In this case, an excellent job I must admit. But I believe this was not the discussion... we were (are actually) discussing if IC does better or not than CCE. At this stage the picture doesn't help since I don't have this movie to compare to your results. I bet (but that is what I believe only) CCE would produce similar results... At this point then it is a matter of choice. If you like IC7 better than CCE, then use IC7 and be happy :)
Still I'd like to see a picture where IC does a better job than CCe. That is, a clear comparison between CCe and IC7. I can (and I think you will admit that easily) do the reverse: show you pictures where CCE does better than IC7...
Still I repeat myself: IC7 does sometimes a great job and I do use it actually from time to time, so I'm not closed-minded.

snidely

26th June 2003, 21:01

I agree with monomer and think this his request is extremely reasonable. Oxdeadbeef, for example, would you kindly provide us with an ACTUAL DVD TITLE where we could buy the DVD and then perform the experiment for ourselves. I, like monomer, would like to personally experience a title where transcoding with IC decidedly beats out reencoding with CCE, rather than just accept it as a "scientific possibility". Like monomer, I am not putting out a challenge, but just wish to see the results for myself rather than take anyone's word for it. As the old saying goes "a picture is worth a thousand words", so please point me to that picture that I can provide for myself (instead of viewing your stills). I mean, in order to have made the assertion in the first place I would think that someone - anyone has a DVD movie that comes to mind that clearly demonstrates the "hard facts".

snidely

26th June 2003, 21:25

***Edit: I see that Fmazzanti is also in the same camp as monomer and myself. Sorry to leave you out, Fmazzanti!

0xdeadbeef

26th June 2003, 22:14

Ah, now the CCE addicts become angry :D
That was not my intention.

First of all: yes, I didn't show proof that "IC7 beats out CCE", agreed. But I never even claimed that, if you read my postings carefully. I also wonder why I need to show proof for my modest claims while people alleging "CCE ALWAYS WINS" don't. Kinda strange, isn't it?

Secondly, I showed some proof that IC7 does a very good job in transcoding at 32% which some of the guys here denied. The screenshot also showed that IC7 is clearly the best transcoder at the moment, which is pretty important, since people tend to speak of transcoders in general.
Yes, I could process the same movie with CCE but do you real expect the outcome would be SIGNIFICANTLY better? Some of you seem to think that CCE would not change the picture at all. Besides: if I get the above quality with IC7 in an hour with no hassle, why would I wait for CCE hours and hours even if it WOULD come out some pecent better?

Thirdly, I commented about reasons, why (properly implemented) transcoding is not only a "quick and dirty" approach as the general opinion amongst the CCE freaks seems to be, but has substantial benefits. I don't see what to proof here. I mean, it's just as it is.

Last but not least, as I a already said, it's pretty hard to judge a movie by some frames. So even if IC7 would win in a selected movie, I think we all agree, it would be a pretty close victory over CCE. So proving the victory would be a pretty complex matter.
On the other hand, as some seem to claim that IC7 creates horrible picture quality compared to CCE in about every movie, this should be pretty easy to prove. So why don't you just prove ME wrong?

BTW: I must insist on the correct spelling of my name. It's 0xdeadbeef, not Oxdeadbeef. Im a hexadecimal number in C syntax, not a neutered bull, though the pun was intended of yourse.

snidely

26th June 2003, 23:15

Ah, now the CCE addicts become angry

Nope, there is no anger here at all. I am learning, and I would like to learn as much as possible, and have the information be as accurate as possible. I just don't understand your reluctance to provide just one title, just one single little title, where we can actually see your hypothesis put into action.

I take two approaches with reducing the size of a movie :

1. If it requires very little compression, then use one of the "fast" compressors (DVDShrink or DVD2One, for examples).

2. If the movie requires significant compression, and a "fast" job won't do, then reencode the movie using CCE for maximum quality.

I have used InstantCopy (which is anything but "instant"), and I feel that that program is in the middle of "nowhere man's land". That is, it is way too slow to be used as a "fast" compressor, and the results that I have seen were not impressive enough to consider it on par with CCE, so it hasn't qualified as a "quality" compressor (yet). That is why I asked for a specific example - I would like to see for myself just once where IC can provide comparable results, never mind better results than CCE. I have an open mind and would be more than willing to embrace IC if I were to actually see for myself how well it can transcode, and that it is in actuality a better product than I had thought.

BTW, I don't know how you are arriving at "one hour" for IC to transcode a movie. On my machine, a P-IV 2.4 ghz, IC required about 3 hours to transcode Die Another Day, while CCE reencoded the same movie (ready to burn) in about 4.5 hours. DVDShrink and DVD2One required about 19 and 17 minutes each (respectively, as memory serves), so IC is a lot closer to CCE in time than it is to the fast transcoders. I didn't actually measure these times, but I will make it a point to log the exact numbers the next time I try this experiment.

First of all: yes, I didn't show proof that "IC7 beats out CCE", agreed. But I never even claimed that, if you read my postings carefully. I also wonder why I need to show proof for my modest claims while people alleging "CCE ALWAYS WINS" don't. Kinda strange, isn't it?

Hmmm...I never made any such statement, so I guess you are not talking to me here. Please understand, someone can come in here and state "the moon is made of green cheese", but if they can't provide any proof, then it is just a theory, don't you agree?

Secondly, I showed some proof that IC7 does a very good job in transcoding at 32% which some of the guys here denied. The screenshot also showed that IC7 is clearly the best transcoder at the moment, which is pretty important, since people tend to speak of transcoders in general.

Maybe I have the wrong person, but I thought it was you that stated something to the effect of "single screen shots demonstrate very little in the way of quality differences. You really have to watch the video in motion to evaluate the quality of the compression or reencoding". Please note, this is not an actual quote, but rather how I understood the quote. If I have the wrong person, or I have misunderstood the intentions of what you have said, please set me straight.

Yes, I could process the same movie with CCE but do you real expect the outcome would be SIGNIFICANTLY better? Some of you seem to think that CCE would not change the picture at all. Besides: if I get the above quality with IC7 in an hour with no hassle, why would I wait for CCE hours and hours even if it WOULD come out some pecent better?

My contention here is that IC does not take "one hour" and in fact takes several hours, and if I am going to dedicate that kind of time, then I am going to go all the way and use CCE. Remember, I prefer to have my compression either "fast" (like DVDShrink or DVD2One) or "excellent" (like CCE), where the time means very little. IC, at least in the past, has not been a winner in either category. Please prove me wrong!

Last but not least, as I a already said, it's pretty hard to judge a movie by some frames.

Ah, I guess I got the right person after all. Did I misunderstand your meaning?

On the other hand, as some seem to claim that IC7 creates horrible picture quality compared to CCE in about every movie, this should be pretty easy to prove. So why don't you just prove ME wrong?

Nope, not horrible at all, so we agree here. I just have not seen a case where IC outperformed CCE even once. My only statement concerning the IC results was "(they) were not impressive enough to consider it on par with CCE". I have not been the one posting the horrors of IC and the wonders of CCE (or vice versa) - I have only requested proof that the moon is made of green cheese. :)

BTW: I must insist on the correct spelling of my name. It's 0xdeadbeef, not Oxdeadbeef. Im a hexadecimal number in C syntax, not a neutered bull, though the pun was intended of yourse.

Gotcha, 0xdeadbeef. Ok, I'll byte (pun intended). What is the significance of the number in C++? :D

monomer

26th June 2003, 23:35

Originally posted by 0xdeadbeef
Ah, now the CCE addicts become angry :D
That was not my intention...
...I must insist on the correct spelling of my name. It's 0xdeadbeef, not Oxdeadbeef. Im a hexadecimal number in C syntax, not a neutered bull, though the pun was intended of yourse.

I get it, that's very funny... I promise I won't mispell it again.

I didn't mean to come off as being belligerent. I really am curious as I want to make the best compression back-ups possible and I'm not very knowledgeable about the technology (though I have recently gained much by way of experience with almost 200 DVD back-ups so far and another 50 or so to go... the goal is to back-up my entire DVD collection to date.) If I am creating inferior back-ups with visual compression artifacts now and later I discover these flaws when I up-grade my display system I would rather know about it now so I won't have to 'redo' my entire collection all over again later. Give me some sample movies I can try now to see if this is going to be a real problem later on. Thanks.

If you would like to see the quality differences I'm referring to between CCE and IC7 ... might I suggest Apocalypse Now Redux... and I would appreciate your honest impressions of the difference between the two final versions after you view both of them. I'll be looking forward to that post.

Take care now...

mrbass

27th June 2003, 02:56

time to chime in here to the ToughMan thread:-)

I'm too lazy do test out myself but from what I can remember if you do want to try to find where DVDShrink 2.3 and IC7 can compare close / or better / or just a tad worse....in other words very very close to CCE encode, I'd say don't lower either program by more than 15%.

--DVDShrink choose level 3 (15%) or level 2 (10%).
--Instant Copy set slider to 85-90% (15-10% compression)
anything above these compression ratios with these transcoders then CCE will come out on top IMHO.

I'm not gonna say which is better in a given situtation...just trying to advise if your going to test for yourself.

monomer

27th June 2003, 03:52

Originally posted by mrbass
...I'd say don't lower either program by more than 15%.

--DVDShrink choose level 3 (15%) or level 2 (10%).
--Instant Copy set slider to 85-90% (15-10% compression)
anything above these compression ratios with these transcoders then CCE will come out on top IMHO...

Good advice I will heed mrbass...

actually during the last several weeks I've found myself using the transcoders (DVD21 and DVDShrink) for anything between 4.37Gb and ~5.2/5.3Gb (equates to ~16%-18%) and using CCE for anything larger. Though I do trust CCE with very large movies (>7.6Gb) I have still been doing those at 1:1 and split across 2 discs.

I will be the first to admit IC7, in my opinion, is the best of the transcoders but the 'unpredictable sizing' issue means I never use it anymore... for the extra time it takes to re-run IC7 over and over, I'd rather use CCE once. (Besides, on my system I 'peg' IC7's PQ limits at around 5.8Gb-6.2Gb tops).

Fmazzanti

27th June 2003, 09:12

Originally posted by mrbass

I'm too lazy do test out myself but from what I can remember if you do want to try to find where DVDShrink 2.3 and IC7 can compare close / or better / or just a tad worse....in other words very very close to CCE encode, I'd say don't lower either program by more than 15%.

Ahaha! That is EXACTLY my point, and I said it several times.
At low compression rates, probably both work great, but if you need to compress more... it is obvious to me who wins, and it seems I'm not alone in this side :D

0xdeadbeef

27th June 2003, 09:34

Ok guys,

last night I reencoded Brazil using DVD2DVD-R just for the sake of personal interest. I used CCE, 2pass, MMX SSE
since I wanted to get it finished till this morning and 'coz Chickenman said he used 2pass. Conditions were the same
as in the transcoder comparison posted above.
Honestly, I expected it would come out about the same quality as IC7.

Imagine how surprised I was to get see my theories proved not just in a selected frame, but in each and every frame of the movie. So I didn't even have to select a new frame or example in this frame but could just compare the IC7 screenshot made for the transcoder comparison:

http://home.nexgo.de/0xdeadbeef/brazil_ic7_vs_cce.bmp (beware, 500k)

Pixel noise everywhere ... better than IC7? No way!

Of course you will argue I should have taken 3pass or more or even floating point. Maybe this would really improve
the noise issue a bit. But the whole process took already 5 and a half hours! The encoder itself needed nearly 5 hours...
Compared to IC7 this is about 5 times the time needed and a worse quality (Athlon XP 2000+, 512MB RAM).

I don't dare to generalize this. Brazil is a movie with pretty bad picture quality compared to other DVDs. Maybe it's
a different story with NTSC. And agreed, one frame (in this case even a selection of this frame) is no real proof.
But the pixel noise is in every frame, trust me.

If I'll encounter a case where IC7 fails for me, I will retry CCE, but until then: why should I?

Fmazzanti

27th June 2003, 10:04

Let's see: you don't have to do anything you don't want. In this sense, if you like IC7 more than CCE then... go ahead with IC7 and be happy!
I downloaded your file and yes... pixelation is evident and yes, IC7 does better under this conditions. For me that means only that:
1) 2 pass is too low. I never used 2 pass VBR. I usually do 4 or 5 passes.
2) DVD2DVD-R does not look for the best settings in CCE. I know this for sure because I have used it several times.
But that's the MAGIC of CCE! You can increase quality by using more passes and tweking the parameters. And that's a big difference compared to IC7, where you can only play with compression rates.
You want to do a good comparison? Go with ReAuthotist and DoItFast4U and DoCCE4U and you'll see what I mean.
For me, quality is more important that speed, so I can afford 7+ hours processing if the resulting movie looks better. You may not agree on that, but that's a personali choice. Furthermore, you keep insisting in that IC does a complete backup in one hour or so, but other people here (inlcuding myself) have a different experience: 2 or 3 hours to do a complete dvd backup.
Could you please explain what do you do to get a reduction factor of 1/2 or 1/3?
Thanks...

Kosh N

27th June 2003, 11:33

If I'll encounter a case where IC7 fails for me, I will retry CCE, but until then: why should I?
I agree that on your pictures, IC is better. So far, I have only encoded Buffy DVDs. I have had to use 40-45% compression, though, and on those I have compared, CCE has come out better. The picture quality was pretty much equal, but what annoyed me with IC was a lot more floating pixels than with CCE. You can't see those on a still image comparison.

I'm using the DoCCE/DIF4U/Reauthorist method, with 3 passes.

0xdeadbeef

27th June 2003, 13:57

@Kosh N:
No "floating" pixels for me. I would be somehow surprised to see problems with the motions vectors as a transcoder doesn't touch them. Maybe something with the difference macroblocks? I would also guess, such an episode DVD is pretty stuffed. Something like 240 minutes? Maybe a bit too much even for IC7.

@Fmazzanti:
I answered all of this in one or another way in this thread and I'm fully aware that, whatever proof I will show for my claims, it won't satisfy some of you. Yet I eagerly wait for a proof for "CCE ALWAYS WINS!". Now it's "CCE maybe wins with 5 passes and tweaking (?)". Pretty clever, even if I would come up with an example using 5 passes, surely my tweaking would be wrong. Yeah, well.

Just to show that I'm willing to admit mistakes on my side: since I really didn't measure the time IC7 needed and it doesn't seem to have a log file either, I relied on my memory there. And unfortunately, it fooled me. As as exact measurement (I just did) showed, the whole process took a few minutes less than 2 hours with IC7 on my machine. MEA CULPA, MEA CULPA, MEA MAXIMA CULPA :o

So it's two hours agains 5 and a half. Maybe with 8 hours CCE encoding I would reach the break even. So what?

Fmazzanti

27th June 2003, 14:37

So nothing... you got me. I'm tired of discussing all over again the same thing. 4 passes in CCe is THE STANDARD. Just move to the other fprums and check... if people wants to do 2 passes, it's up to them, but it is well known that CCE does its best at 4 passes. 'Tweaking' CCE means simply USING THE RIGHT PARAMETERS. There are few of them and many people simply let's them at their default values... ok for them, but missudsing CCE's capaibilities doesn't mean CCE produces worse quality. Now you want a proof were CCE performs better... you already know. Just grab any 2h30+ movie as Lord of the Rings and do your IC stuff.
Now as I said I'm tired of this discussion so I withdraw. None of us is going to conince each other, so what's the use?

0xdeadbeef

27th June 2003, 15:04

@Fmazzanti:
I agree we disagree.

However, if you tried to convince me, I didn't notice. You claimed this and that but didn't really try to prove it. As did others in this thread.

Then again, it wasn't my intention to convince anyone here to use IC7 instead of CCE. I posted in this thread to object to simplifications like "CCE ALWAYS WINS!". Then I was asked to back up my claims, which I did. Make of it what you want.

Well, whatever, I rest my case.

monomer

27th June 2003, 17:26

Originally posted by 0xdeadbeef
... Well, whatever, I rest my case.

Whoa, hold on there... I still haven't been convinced of anything, one way or the other. First of all I'd like to thank you for responding to my request. So the movie is Brazil. I have just ordered it and expect to receive it by next Thursday or Friday. Of course its NTSC but I'm thinking I should still be able to replicate your comparison. I will immediately stop backing-up my DVD collection until I can get this all sorted out for myself. Now there is something to be said about using the correct settings in DVD2DVD-R (this I had to learn the hard way). I will just use 2-pass VBR, like I always do. BTW, do not be offended that I'm not bothering to look at the stills you've posted... as I've said before, there are many reasons why stills can be very misleading. I will definitely post back here with my honest impressions by next week Saturday. It would be nice if someone else could also give it a try (especially the PAL version). As I've said before, I am very interested in learning... and one of my best modes for learning is to experiment and judge for myself. Again this is not a contest or a slug-fest... I'd like to think of it more as science and thus I find I can be very objective.
Who else is game to give this movie a try? BTW, I've never seen this movie before... I hope it's a good one to add to my collection. And again 0xdeadbeef, thank you for the DVD title.

Edit: Have you tried Apocalypse Now Redux? Or you might try Grey Zone, it's a much smaller movie (thus faster to process) and probably much easier for you to find than the aforementioned.

Fmazzanti

27th June 2003, 20:18

Originally posted by 0xdeadbeef
@Fmazzanti:
I agree we disagree.
However, if you tried to convince me, I didn't notice. You claimed this and that but didn't really try to prove it. As did others in this thread.

Take it as you want... if you feel like a still image made by you using DVd2DVD-R with default settings and 2-Pass VBR is enough to prove CCE quality, then that's it. Your still doesn't prove anything to me. My point is that you did it the wrong way. Or put it in another way: you didn't do it the way one gets the best of CCE...
In any case I said this was over for me and actually it is. take it as if you won and let's all be happy :D

Zhnujm

27th June 2003, 20:46

I have now done "Conan" with InstantCopy (before i only used DVD2One and CloneDVD on this movie) and i agree that InstantCopy would be a good choice for this movie.
The differences between CCE and InstanCopy are very small, DVD2One can not compare to these. The Pictures even do not show how bad DVD2One looks, you have to see it moving.
And i must say that i had to search for pictures that shows a difference between CCE and InstantCopy. The Movie is 7.1GB.

You can TRY to download them here:
Picture 1- ftp://zhnujm.dyndns.org/Image1.png
Picture 2- ftp://zhnujm.dyndns.org/Image2.png

Top Left - Original
Top Right - DVD2One
Lower Left - CCE
Lower Right- InstantCopy

If i get my original of T2 back i try it also, but i have done it before with InstantCopy, and it came out not that good.

snidely

27th June 2003, 22:41

Hi guys,

Sorry, but I have been tied up for the last 12 hours with a major fire that was just a few hundred yards away from my business. It seems to be under control and it looks like my building will be staying up for now. :)

Anyway, I am very glad to know that 0xdeadbeef has provide a title for us to try, as that was all that I had been requesting...THANK YOU! Like monomer, I decided to purchase Brazil for myself and run the same comparisons so that I can see how the two videos compare while in motion. As another person in NTSC land, I hope there are no major differences between the 2 versions other than the obvious, and as soon as I recieve my copy and run the comparisons, I will also be glad to report my honest findings, good or bad.

Oh, and as far as the time issue is concerned, if I were to choose a dividing point, it would be one half hour. That is, the "fast" compressors take less than a half hour to do their work, while the "slow" compressors take over a half hour. Basically, am I willing to wait right in front of my computer for the compressor to finish, or am I going to go to bed and wake up the next morning to find out what has happened? That type of division clearly puts IC and CCE in one category, and DVD2One and DVDShrink in the other, and the differences in time between CCE and IC are meaningless, or at least for my purposes. Others may feel that 2 hours still represents "fast", and that 4.5 to 5 hours is "slow", so each person will need to decide for himself what kind of time he is willing to dedicate for what kind of quality.

mb1

28th June 2003, 00:33

@ Oxdeadbeef
To my view at your comparison images CCE image is sharper than the original.
As everybody knows that CCE slightly softens images there is only one conclusion (at least to me):
either DVD2DVD-R or you used any sharpening filter ...
Do you have the used dvd2dvd-r script for us ?

This is in no way a handmade optimized cce encoding.

I personally don't use dvd2dvd-r (never !) as I know I do it better myself.

Looking at Zhnujms images CCE clearly wins (after the original of course).
Even in IC7s images there are lots of 8x8 blocks recognizable (even to the blind) whereas CCE images show very pleasant bitrate reducing.

To my experience CCE wins every comparison with transcoders doing progressive dvd source.
Of course CCE has to be setup correct which is problematic for all non-advanced users (I wonder why, there are only few options ?).
And there is no need for 4 passes (which means 5 in real).

And another statement:
No other transcoder reaches up to now IC7s' quality (if the movie is shrunk the same compression).

Now the third statement ;) :
Doing interlaced source (no matter what source) Canopus ProCoder beats every other encoder and transcoder by ...... very far.
And no matter what bitrate.
I can see the differences on any bitrate (as I know the programs weaknesses) even 9.500 CBR.
And I verified that for others who did test me :)

MackemX

28th June 2003, 03:05

my eyes hurt after reading this thread :eek:

as for DVDShrink quality compared to CCE, I'll use what I've used before regarding Premiership football teams

let's just say that CCE is the MAN UTD (currently one of the best and beat nearly everything else) of backup tools, DVShrink is Newcastle :p:spit: (the best on odd occasions and rarely beats MAN UTD) :D

mrbass

28th June 2003, 05:58

Brazil is a 1985 movie....not the best to start doing tests on. How about something like 1998 or later.
@mb1 ...yea that canopus pro coder $600... I plan on using that when I get into mini-dv conversion to dvd. It's actually just ligos mpeg encoder which I think they sell for $70 or so ..can't remember.

ok found my post back in feb 2003
http://forum.doom9.org/showthread.php?&threadid=45476
Canopus Pro Coder 1.2 is over $600...nice cuz it allows you to stitch together various sources. But in the end though for SVCDs and DVDs (mpeg2) all it is using is the $79 Ligos LSX MPEG Encoder ligos.com

monomer

28th June 2003, 07:26

Confession: I did 'sneak' a peek at 0xdeadbeef's 'stills' and I had to really zoom-in to actually see anything. In those captured 'stills' nothing's very sharp at all but it would definitely appear that IC is softer than CCE. I zoomed-in quite far and found it was curious that the IC labeled picture never showed any macro-blocking structure at all... if anything the CCE labeled picture began to show the 'blocking' first. But then looking at Zhnujm's 'still' shots it's very much the reverse. His 'still' captures are both very sharp and quite detailed. Without even having to zoom-in its obvious that both the transcoders have the obvious block structures I've seen before on my own transcoded movies while in the CCE 'still' the blocking is practically non-existent (only faintly 'hinted' at in a few selected locations).

Why the stark difference between these two comparisons of CCE and IC?

Could it be due to the degree of compression? Zhnujm's Conan @ 38% vs 0xdeadbeef's Brazil @ ??% (I'm assuming its much less than 38%).

The quality of the original? the Brazil shot was very 'fuzzy' as if it were out of focus or very far away... however the Conan pictures were very sharp and well defined...

I'm stumped as to why these two comparisons appear to suggest opposite conclusions. In any case, I don't believe 'stills' are always accurate in gauging the viewing quality of a movie.

So, Conan is next on my list after Brazil.

Edit: In one of 0xdeadbeef's earlier posts, he states that Brazil was compressed 32% (that would make Brazil ~6.4GB)... so I guess the degree of compression is probably not a factor in why those comparisons disagree.

MackemX

28th June 2003, 09:40

Originally posted by monomer
In any case, I don't believe 'stills' are always accurate in gauging the viewing quality of a movie.
Exactly!!

you could look at other stills and the order could be different

the best method is to do it yourself (if possible) and make your own mind up using your own preferred viewing setup and use whatever DVD Backup method is most pleasing for yourself and not someone else

why is it there are picture adjusments you can make to your viewing setup?. It's because people have different viewing preferences regardless of the quality of the image, so even this is a factor in a way

'to each his own' :)

quantum

28th June 2003, 16:23

Originally posted by monomer
But then looking at Zhnujm's 'still' shots it's very much the reverse. His 'still' captures are both very sharp and quite....
Why the stark difference between these two comparisons of CCE and IC? I don't know about beef's stills. I don't see the point in those myself, but Zhnujm's stills are a good example of my own experiences with CCE, IC, and DVD2one at higher compression rates. I like the way he put those together with the 4up format.

snidely

28th June 2003, 17:26

Looking at Zhnujms images CCE clearly wins (after the original of course). Even in IC7s images there are lots of 8x8 blocks recognizable (even to the blind) whereas CCE images show very pleasant bitrate reducing.

Oh my God! You are absolutely correct! Unless my computer is doing a really terrible job at scaling, then the IC stills look almost as bad as the DVD2One stills. CCE is clearly the best, but I still insist that stills do not tell the whole story. I will have to see the video in action for myself to make any kind of accurate assessment. I'm going to order a copy of Conan the Librarian and try that title out as well. :)

Zhnujm

28th June 2003, 20:39

Originally posted by snidely
I'm going to order a copy of Conan the Librarian and try that title out as well. :)

Dont know if thats a good idea, i think there are more than one versions out, mine was labeld "Digitally Remastered" and was definitly only for the german market as it has forced german subtitles if you switch to english :rolleyes:

And as said above, i had searched for the pictures to show a clear difference between InstantCopy and CCE, others looked better. If you see it in motion you can clearly see that InstantCopy does a much better job than DVD2One.

Just for fun i now encode a newer and smaller (5GB) movie, if InstantCopy/DVD2One can look better than CCE it should be in this case.

Zhnujm

29th June 2003, 14:32

Now a new and small movie. I created a 5GB movie only version out of Shang High Noon and here are the results:

Picture 3- ftp://zhnujm.dyndns.org/Image3.png
Picture 4- ftp://zhnujm.dyndns.org/Image4.png

Top Left - Original 5GB
Top Right - DVD2One 4GB
Lower Left - CCE 3.9GB
Lower Right- InstantCopy 4GB

I fail to see a difference between CCE and InstantCopy here.
So my result is that InstantCopy can look as good as CCE with short movies but can not beat it.
End of story again (but only for me i think :D).

0xdeadbeef

29th June 2003, 16:50

@Zhnujm:
Though I don't intend to restart any discussion or even take part in what's going on now: just for my personal interest, it would be nice if you added some more details about Conan.
I think we talk about "Conan the Barbarian", the 1st part, right? Did you remove any sound track? What was the exact size of the main movie before starting transcoding? What was the size after transcoding? Did you transcode only the main movie even with IC7 which doesn't allow this as standard option?
Without this information, any picture comparison is pretty useless.

[EDIT]
And one more remark: IMHO the "guy in armor" picture in the upper left (which you say is the original) has clearly more blocking than the one in the lower left (which you say is CCE encoded). This seems a bit unlikely to me. Sure you didn't mix up these screenshots?

Last but not least I's like to mention that all your screenshots show aready clear blocking in the original. Sure a transcoder can only make this worse.

mb1

29th June 2003, 17:06

Image 3: Look at the horse with its 8x8 blocks (IC7 and D2o). None (almost) for CCE.
In fact this is not visible on tv sets. But as one needs no magnifying to recognize that (at least me) this could be recognizable very well on beamers ...

Image 4: Quite calm one. D2o looses details for the background mountains and there are some small blocks left of the trees in the left center.
But again no problem for tv sets.

@mb1 ...yea that canopus pro coder $600... I plan on using that when I get into mini-dv conversion to dvd. It's actually just ligos mpeg encoder which I think they sell for $70 or so ..can't remember.
mrbass, that's a widely done mistake conclusion.
For versions 1.0 to 1.25 procoder used Ligos decoding technology. We both know that this is different to encoding ;)
The encoder is completely canopus.

Versions 1.5 upwards (present version is 1.50.22 for 8 weeks now) can use canopus own developed mpeg decoder, too, which is now part of the package.

What boosts 1.5 interlaced mpeg2 encoding is the now correctly implemented "always field structure" mode. Quality increase is significant to normal always frame structure encoding.
Field and frame structure have nothing to do with interlacing.

It is an encoding option which no other (good) encoder allows at the moment. CCE and Tmpeg and Mainconcept and Ligos only encode "frame structure" always.
But that's not the thread to discuss about that ... ;)

GlitchOfDoom

29th June 2003, 18:27

I just burned Gladiator with DVD-Shrink under level 7 compression and it came out perfect, what I think compression does is skip frames when ripping the disk, not compressing the actual video like MPEG does.

Zhnujm

29th June 2003, 21:34

0xdeadbeef:
I check this immediatly, the top left picture show clearly more Macroblocks.

-edit

No, its really the original,i loaded it now directly from the DVD. Very funny. The frame before and after this frame does look good, its the only frame that has such blocks. Maybe that shows that the source is already bad compressed, but thats the reason why i use it for tests, it seems to be a very hard source for transcoders. DVD2One and CloneDVD fail completely on this movie, were InstantCopy does a (for me) surprisingly good job.
But the more interesting question would be why this blocking is not in the CCE picture ? I have to say that i have done this CCE encoding some time ago, maybe i have used the build in and normally enabled filtering. Dont know if the filtering can have such an effect. (i disabled them for "Shang High Noon")
To be sure i will do this encoding again.

And yes it is Conan the Barbarian part 1, i have done a full copy (7.1GB) the movie is 5.1GB, of course the CCE version is also a full copy (in fact a DVD2One copy with a replaced main movie), all three versions are around 4.2GB, thats why i didnt mentioned it.

@mb1:
You are right about picture 3, i was to focused on the water :rolleyes:

-edit again
Can somebody with more experience have a look at picture 2 ? In my eyes the cce picture looks better than the original, but the picture top left is REALLY the original. Can this be caused by some filtering in CCE ?

quantum

29th June 2003, 23:23

I'm not an expert, but my understanding is the filters normally checked by default under the "Quality" button of CCE cause a softening effect which could reduce the effect of existing compression blocks.

I was also under the impression CCE can rearrange the GOPs such that compression artifacts can be moved around to different frames.

I also did some tests with dvd95copy today and found (as others suggested can happen with transcoders) that certain frames are very close to the original, while other frames show heavy block artifacts.

Zhnujm

1st July 2003, 15:19

But shouldnt the smoothing be more visible in the other parts of the picture ? I was thinking that there must be a greater difference. But then, im not an expert. ;)

digitalman

2nd July 2003, 14:11

Has anyone tried using the Effect Restricted Vertical Filter(Anti-Noise Filter) with setttings between 8 and 16 in CCE. I get uncomparable better results than all of the one click DVD backup solutions. I know most guides say to disable the filter, but I think it really works well.

redeemer-dk

2nd July 2003, 17:07

mb1:
CCE and Tmpeg and Mainconcept and Ligos only encode "frame structure" always.

That is not true. By unchecking "progressive frame flag" in CCE it goes into interlaced mode, and it does both frame DCT coding and field DCT coding for macroblocks when encoding field based material.

Kika

2nd July 2003, 17:26

@redeemer-dk
Field-Structure-Mode is a completely different thing.
Only ProCoder can encode in this mode.

@digitalman
If you like unsharp Videos, turn this Filter on. If you like sharp Videos, turn it off.

redeemer-dk

2nd July 2003, 17:32

What is field structure mode then? Does it have something to do with the picture_structure field?

Kika

2nd July 2003, 17:41

@redeemer-dk

Yes, that's right. It's a different way to handle the Fields.
Other encoders do use field DCT, that's right, but the fields are not separated. So you ever get a Frame Structure. ProCoder can handle the Fields separatly, so you get a Field Structure.

redeemer-dk

2nd July 2003, 17:52

Oh, I see. I just checked an interlaced stream encoded with CCE SP where progressive frame flags was off and the picture structure was still 3, which is frame picture.
I'm actually trying out ProCoder now, and the first impression is that there are very few settings you can change, is that true or are they just hiding somewhere?

Zhnujm

2nd July 2003, 18:22

btw, i encoded the movie again without filtering and it looks the same as before...

newguy007

6th July 2003, 07:49

Zhnujm: how did you take those screenshots (what software) and structure them the way you did? I'd like to do the same comparison with some of my dvd's and post them here.

Xayd

6th July 2003, 15:26

I've tried IC, which most of you consider to be the best one click solution quality wise, and use it to compress extras still, but it has a serious dropoff in quality below ~60% of the original size of the video stream.

CCE on the other hand I've used to compress video streams down to 35 and 40 percent of the original size in some cases with no noticable loss in quality.

And I only do single pass encodes with CCE, too, at a "Quality Priority" setting of 5. It's rare that I see enough difference between the original and the transcoded video to warrant a second pass after scanning the 'rough' scenes on a computer monitor.

Dark movies are a bad judge of these one-click softwares' abilities.

Try a movie such as Goodfellas and watch the opening scene with the red tail light over the actors' faces, or the pool hall scene in The Usual Suspects with red painted walls in the backgrounds. Bright scenes are where your blocks and jitter will show up.

Zhnujm

6th July 2003, 18:58

@newguy007:
VirtualDubMPEG2. Just load the vobs and copy&paste the pictures to your favorite Image Editing Software.

monomer

7th July 2003, 19:06

I just got back from a little "July 4th" vacation with my wife and both movies Brazil and Conan had arrived in the mail... well there must be a significant difference between the PAL versions and the NTSC versions. Both NTSC movies are well under 4.37Gb... therefore no compression is needed... WHAT A BUST! I keep learning new things all the time. This time I've learned that there is no relationship between any version of any given movie... it must be that every version is originally "scanned and compressed" from the actual film and the bitrates and such are at the complete discretion of the person doing the transfer. Therefore the next time I try to replicate someone's movie compression experience I will need to get the exact same version (making sure the ISBN code matches). So the bottomline here is I still have to go by my own past experiences with DVDShrink, DVD21, IC7 and CCE unless someone can provide me with a NTSC movie example of the transcoders 'beating out' CCE. My own experiences have shown me that all the transcoders have practical limits in terms of picture quality (which seems to vary somewhere between 5.5Gb and 6.4Gb) but CCE seems to be without PQ limitations (up to 8Gb anyway, since I've not had the occassion to go beyond this point). So far, my experiences have been very similar to the examples Zhnujm has produced... however, as a general rule, I don't usually compress a title by several different methods, so I am not able to make many direct comparisons. I just use CCE for anything larger than ~5.3Gb and use the transcoders for anything smaller (and I don't use IC7 anymore at all because of the sizing issues, which I do regard as being a major issue).
BTW, Conan I've seen before and liked but Brazil is... uh... well... a little bizarre... but that's another discussion for another time.

mrbass

7th July 2003, 20:36

Originally posted by monomer
(and I don't use IC7 anymore at all because of the sizing issues, which I regard as a major issue).

welcome back to reality monomer. Same here with IC7. Did you get a chance to test Matrix NTSC R1 level 2 10% with dvdshrink and compare it to CCE? I'm too lazy.

AM Tiainen

8th July 2003, 01:45

for Zhnujm, thanks for images that you have posted. And also that is funny that people think that you may mixed images ;-) That is not any strange that image in CCE looks better. CCE always make littlebit smoother (less detail) image quality than others software like TMPGEnc. But difference between Original and CCE is so little that i amazed that someone notised it easily.

But i have surprised that IntantCopy is so good as it is. Have any tested Intervideo DVD Copy? I have yet only SVCD videos but i have say it that made pretty good quality. Also some Germany Magazine have tested "OneClick DVD copy software". CloneDVD and DVD Shrink have get as much points as IntantCopy on picture quality.

0xdeadbeef

8th July 2003, 08:19

If it's really filtering that removes the macroblocks than it's quite intensive blurring. Fortunately, there isn't much detail in this frame, so the only detail that's lost is the blocking.
Then again I wonder if this amount of blurring in each and every frame is what I want in my DVD backups.
Interesting enough that everybody is happy that Zhnujm's screenshots prove their beliefs, while my screenshots could be easily ignored. Well, probably psychology.
Whatever: I've done some more IC7 backups, tried to use compressions factors below 30% and the picture quality is impressive each and every time, so I decided to ignore Zhnujm's results ;)

monomer

8th July 2003, 17:32

Originally posted by mrbass
welcome back to reality monomer. Same here with IC7. Did you get a chance to test Matrix NTSC R1 level 2 10% with dvdshrink and compare it to CCE? I'm too lazy.

mrbass...
Very clever how you brought this thread back around to resembling its title: "DVDshrink quality compared to CCE?"

First a little background... last week mrbass requested through a PM that I try compressing THE MATRIX NTSC R1 (fortunately I do own this exact title) using DVDshrink at level 2 (10%), which means 'shaving' off the ending credits and the WB's intro logo. Then try the same movie using CCE, which I'd did using the DVD2DVD-R/Scenarist method (thanks to mrbass)... I used the usual 2-pass VBR including all of the intro and credits in the compress. I then burned these 2 DVDs and took them on vacation for viewing. Well, I forgot to view them until I got back. Anyway, last night I spent a couple of hours switching back and forth between both versions and this is my opinion...

After careful examination I find myself 'imagining' that in the DVDshrink version sometimes the background looks a little grainy compared to the CCE and the original but unless I can get my wife or sister to do a series of 'mystery' swaps to see if I can guess it better than 50/50, this may very well just be a case of my 'wanting' CCE to be better... thus inconclusive. BUT WAIT... THERE'S MORE...

I made another observation that can easily survive the mystery swapping test. This was most unexpected. The DVDshrink version has quite a few 'jerky' transitions. Careful scrutiny of 3 separate locations in the movie (totalling about 30 minutes of the whole movie) I found 5 instances of this occuring. By comparison neither the CCE version nor the original display any hint of this. Its as if the movie stops for a split second before resuming (and then it seems to maybe drop a few frames to catch-up). It's very noticeable during the loud action scenes because an audio 'hick-up' occurs concurrently. Just to eliminate a badly manufactured DVD-R from being the cuprit, I will redo and re-view the movie again. But I would appreciate it if someone else would try this movie with DVDshrink2.3 and watch the first 20 minutes and tell me 'if' and 'when' you see/hear the hick-ups. I find there are 2 un-mistakeable 'hick-ups'. Let's see if we agree.

Thanks to anyone who participates...

EDIT: the ISBN code for my version is 0-7907-4261-6

monomer

8th July 2003, 17:56

Originally posted by 0xdeadbeef
...Interesting enough that everybody is happy that Zhnujm's screenshots prove their beliefs, while my screenshots could be easily ignored. Well, probably psychology...

Come on, let's be honest here... basically, your screenshots sucked. They were very small and so blurry and indistinct that not much could really be deciphered in a comparison. Whereas, Zhnujm's screenshots were large, sharpy detailed and showed very obvious PQ flaws. The difference between the transcoders and the CCE/original was stark and clearly evident. The macro-blocking all but 'kicks you in the teeth'... without any zoom-ins, it is very, very, VERY obvious. By contrast, the difference between the CCE and the original itself are very slight and hard to see even with zoom-ins. In this case (and often in general), the 'softness' of CCE is probably a good thing. Face it, there's no psychological forces at work here, its just so plain and the only one who can't see this is you.

Just get over it and let's all move on...

0xdeadbeef

8th July 2003, 19:50

My screenshots were small, agreed. But in no way they were more blurry than Zhnujm's. In case you didn't notice: I cut out identical squares to make comparison easier. Agreed: no perfect choice, but neither are Zhnujm's.
In very large screenshots, it's hard to see what you're looking at. This may explain why nobody (but me) noticed that the original picture has more blocking than the one recorded with CCE. The other explanation would be psychology ("Ah, I knew it was true. The world is flat!"). In any way it's unfair to complain about blocky transcoding if even the original is blocky.
And about the blurriness: of course this screenshot looks better with CCE. Who could deny this? It even looks better than the original as I told some postings before!!! However, I bet I could find a picture in the very same movie were IC7 looks great while CCE looks blurry.
It's a shame that you gave up testing yourself but decided to follow the flock. Then again you didn't like Brazil. Hm...
And I will never get over having my own mind. Sorry about that.

mpucoder

8th July 2003, 20:24

One of the problems with comparing screenshots is the software used to resize the image, which usually resides in the display driver or card, and varies by manufacturer. You should be looking at unresized 720x480 or 720x576 bitmaps extracted straight from the vob, but again, the rendering of mpeg to bmp varies according to the software performing it. And, of course, bmp's are too big, so to save bandwidth jpg's are used, adding another process that can alter the results.

quantum

8th July 2003, 21:06

One of the problems with comparing screenshots is the software used to resize the image, which usually resides in the display driver or card, and varies by manufacturer. Zhnujm used virtualdub. No resizing going on unless he did it deliberately in an image editor afterwards.
You should be looking at unresized 720x480 or 720x576 bitmaps extracted straight from the vob Which is what he did.
And, of course, bmp's are too big, so to save bandwidth jpg's are used Zhnujm used png's, which are lossless.

Zhnujm

8th July 2003, 22:15

Around 180000 frames in this movie, why o why did i use that one ? :rolleyes:

@0xdeadbeef:
If you want to see blocks that are not in the original you can look at picture 1+3.
Or look at this one ftp://zhnujm.dyndns.org/Image5.png
And as i say now the third time, the difference cce vs instantcopy with this movie is small, most times they looked much more the same than the examples. But dont say you can find a frame in the ic encoding that looks better than the cce one if you have not looked at it for yourself.

And btw, im not interested in letting cce look better than the other programs.
If you are interested, send me a list with DVDs you have and we can encode the same one, if i have it too.

0xdeadbeef

8th July 2003, 23:28

@Zhnujm:
I don't doubt IC7 creates blocks under some circumstances. I just try to point out that this is an extreme example and not typical for a two hour movie. Plus it's absolutely weird that - in this case - unwanted smoothing improves image quality. In a "normal" test case, this kind of smoothing would degrade image quality.

But lots of people tend to simplify and don't even read your posts completely (not to speak of mine). So some guys here might find the "CCE always wins" nonsense justified by looking at your pics, compress with one or two passes and lose picture quality. Or at least they won't win any considerable amount of quality with multiple passes but waste hours of time. Then again, if they want, they shall.

Just for the sake of interest, I'll try to rent Conan and do some testing on my own. Maybe in this case, the compression factor was just too high. Dunno. Indeed, most movies I tried lately (about 15) easily fit to DVD-R with just 15-20% compression. Brazil was the only one reaching 35% even after removing Extras.

MackemX

9th July 2003, 01:01

Originally posted by mpucoder
And, of course, bmp's are too big, so to save bandwidth jpg's are used, adding another process that can alter the results. if people zip up the bitmaps, they can be made a lot smaller

but personally I aint bothered about other people's opinions anymore regarding quality anymore cos why should I be?

I don't need someone telling me A is better than B cos it's whatever I feel is most pleasing on my eye that I'm bothered about about and screen captures are basically not the best comparision methods anyway

I feel if you are that fussed about quality then you dont need someone else to tell you which is the best cos you should be able to see it for yourself, preferably doing your own small tests using a DVD RW

it looks like this thread is turning nasty just the same as all the rest have and do :rolleyes:, so I guess in time it will die with the rest of them

p.s. Long Live IC LOL :D

p.p.s. at last I've got the 1000 up :)

Zhnujm

9th July 2003, 17:37

Originally posted by MackemX
it looks like this thread is turning nasty just the same as all the rest have and do :rolleyes:, so I guess in time it will die with the rest of them

You should think more positive, at least its already on page 5 WITHOUT ending in a flamewar. :D

digitalman

10th July 2003, 20:32

What about the near 1:1 encoding speed Mainconcept Encoder(http://www.mainconcept.com/products.shtml) compared to CCE, and IC7? :) j/k

Zhnujm

11th July 2003, 22:15

Just for fun i did the test with Mainconcept and also added TMPEGEnc.
As far as i know the Mainconcept Encoder does not support 2-pass encoding (please correct me, i did not find a setting for that), so the results are in some scenes very good:
ftp://zhnujm.dyndns.org/Image1a.png
ftp://zhnujm.dyndns.org/Image2a.png
ftp://zhnujm.dyndns.org/Image5a.png

but it seems Mainconcept completely fails at high action scenes:
ftp://zhnujm.dyndns.org/Image6.png
ftp://zhnujm.dyndns.org/Image7.png

order:
Original DVD2One Mainconcept
CCE InstantCopy TMPEGEnc

beware the pictures are 2-3MB

Please post some settings if you think the Mainconcept results are bullshit, so i can redo it. ;)

14th July 2003, 19:32

By comparing different movies results you are comparing apples and oranges.
I found some movies compress extremely well while others do not. I do movies only, no menus.
For example:
Merlin is a 3hr movie needing ~40% compression. The results were great with everything, IC7 was the best, even DVDShrink was watchable.
Titanic also 3:06hr movie needing ~40% compression. IC7,DVD2one,DVDshrink, was terrible. CCE was good but had an error in the last 6 minutes. I gave up and just split into 2 DVDs with DVDfab.

Some movies will come out better with CCE some with IC7. IMHO
I don't have the time needed for CCE 4-5 passes on a daily basis.

timekills

24th August 2003, 00:34

Shadow of the Vampire looks decidedly better using IC (albeit version 8) than it does with CCE (2.66). It also had markedly higher Q-factors upon Bitrate Viewer measurement with CCE than with IC or DVDShrink (3.0b5)

Of course, minus the Q-factor measurement it's all just IMHO. :sly:

freaky_00

24th February 2004, 14:02

Now that InstantCopy 8, DVDShrink 3.1.5 and CCE 2.67.00.23 are out I'd like to know your opinions about them. What about we COMPARE these Apps. I think we pick a DVD title. One for PAL and one for NTSC. It needs to be either a perfectly digitalized movie or one which has always been digital (after 1999/2000) so that transcoding and recoding do not end up in very bad macro-blocking. We should pick one title wich needs 15% compression, one wich needs 30% and one above 40%. And for the result to be absolutely compareable we should use only the main movie (because many of the foreign releases miss supplements and the size of the movie is always different from their US equals) and MUST USE EXACTLY THE SAME TWEAKS AND OPTIONS!!! I prefer quality and do not mind the time needed. As I am like otherse here from the PAL Region I have a couple of DVDs. All PAL and german releases.

So what do you think about this?

daehkcid

24th February 2004, 18:22

Cool.. How about Finding Nemo?

Buddy Casino

24th February 2004, 22:33

The point of such a comparison escapes me. Why use the special case "CGI movie"? Since there is no noise in them, all transcoders should perform pretty well, and CCE too, of course. I mean, you can do such a comparison, but I don't think the result will be valid for non-cgi movies.
And btw.: if you prefer quality, CCE is the way to go (at least in 99% of all cases). With IC8, you often get comparable quality in less time, but I've never seen it perform better than CCE. :D

U977

26th February 2004, 08:45

Hi all,

Wow, it took me such a long time to read (almost) the whole thread :-)

But after all this, don't you think it's another discussion about quality of the images you get, and then again, it depends on your own taste?

My personal experience is the following one. At the beginning, I was using the Big3 method because I was convinced that reencoding was the way for best quality.
But once, I was not satisfied (Spiderman movie, PAL) because I was getting blocks with CCE. Just because it was fast, and because I was curious to see the difference, I gave a first shot with DVDShrink 3 beta 5, then. Less blcoks, image was pleasing me more. Since then , I use DVDShrink when I don't need to compress the movie much (less than 10 or 15%).

When I've read some posts in this thread, some other posts about quality of the different MP4 codecs came back into my memory: DivX 3.11, XviD, DivX 5, RV 9? I know it's a totally different story, but yet discussions about the "quality feel" were often similar to some of the posts here.

I admit that personally, I hardly trust images posted on the net to compare the results produced by different tools.
mpucoder already gave one of the reason, and I remember I already read somewhere hereabove the second reason I also have in mind: a screenshot is just a screenshot. I prefer to see the frame integrated in its video stream.

Richk50

27th February 2004, 19:09

I may be totally wrong, but isn't CCE used to create commercial DVD's after film has been converted to a digital format.

Kedirekin

27th February 2004, 21:47

I've seen a couple of DVDs that I thought we done with CCE (they had the characteristic mosquito noise), but I think most DVDs are done with dedicated hardware encoders.

terminator911

8th March 2004, 17:09

CCE is the king quality wise! I have 3 computers and leave them running overnight! Use DVD2DVD-R along with CCE and AnyDVD and you won't even have to rip anything to your HD.

Insert movie, type in a few numbers and let it run.... it does it all for you, it even authors the new DVD, what else can you ask for? You can even have DVD2DVD-R shut down the computer for you once its done.

I use this method all the time and get the best results on movies that would otherwise require too much compression and the end result would be something that looks like one of my old VHS movies :o

Cheers,
Terminator

MackemX

8th March 2004, 17:47

Originally posted by terminator911
Use DVD2DVD-R along with CCE and AnyDVD and you won't even have to rip anything to your HD.
I haven't used it in ages but I thought DVD2DVD-R worked direct from the DVD anyway without the need for AnyDVD or do the later versions not work anymore?

terminator911

8th March 2004, 19:52

HHHHMMMMMMMM come on MackemX at least let me live in my fantasy that I have something going here :o HEHEHEHEHEHEEHHE!

You have me thinking now, I'll have to try it and see, I don't remember trying to use directly from DVD before as I always asummed that it couldn't do it!

Cheers mate!
Terminator

MackemX

12th March 2004, 16:54

does anyone know if DVD2DVD-R does work from the DVD? (I have version 1.3.5!)

p.s. dream on terminator911 ;), sorry to spoil your fantasy :D

MedicineMan

13th March 2004, 01:44

It only works if your movie isn't copy-protected.

jorel

13th March 2004, 04:44

i'm not expert but did my tests:

using toy story 1 & 2 in dvdshrink:
compression settings-->automatic
toy story 1 main movie,only one ac3-5.1 ---> 52,2%
toy story 2 main movie,only one ac3-5.1 ---> 49,7%
for a full dvd size target (DVD-5 (4.7GB)adjusted in preferences)!

the same with CCE 2660107 using --> Q10:
toy story 1 main movie,only one ac3-5.1 ---> ~1,70Gb
toy story 2 main movie,only one ac3-5.1 ---> ~2,15Gb
got ~3,90Gb in hd...(i don't burn cos i'm doing menus)!

CCE give less size, better quality(sharpness,bright,details without blocks)
using filters(undot,deen,asharp,levels,dctfilter) in avisynth script!

:cool:

edited>
more details:
was used the vmesquita template for cce266:
http://www.jltoca.uaivip.com.br/vmesquita.htm
following his guide here:
http://kvcd.net/forum/viewtopic.php?t=5142

rca29

2nd April 2004, 20:37

For the last 20 months, i've been doing a lot dvd backup's, including music concerts (which i give extra care). It can be easily understood that a music concert is a great challenge for any encoder, because:
- more than 90% of them are interlaced
- It's a high movement images, with sudden and frequent changes of lightning, strobes, etc.
- In most of the cases, you are paying a lot of 'natural' attention to details (such as fast moving fingers on a guitar, keyboard, drummer sticks, etc.)

On those early days:
- Rempeg2 = It was never a solution. Too many macroblocks and the image looked terrible.
- TempEnc = Better quality image than the later, but it was still 'unusable' for most of the cases. Even with average bitrates of 4.500 kbit, it was prone to blockiness.
- CCE = The clear winner in these demanding situations, with the 'mosquito noise' as it's only problem.

The evolution:
- I always 'looked' at the first transcoders (DVD2ONE) as a 'ghost' of ReMpeg2. I've tried more than one version of DVD2ONE and none of them convinced me.
- A few months ago, i tried dvdshrink with low hopes, and i got a (very) pleasant surprise, which led me to more testing.

Although i use the CCE/Scenarist in some instances, in most of the cases i don't bother anymore if i can get away with compression ratios of less than 20% in dvd shrink.
With these compression ratios, i don't think that dvdshrink results are worse than CCE. Maybe they're even better.
If i have to aply much more than 20%...i'll go with the CCE, because you will really notice strong artifacts on dvdshrink that you won't in CCE.
In any case, a movie with more than 3 hours and one multi-channel track has only one way of having good quality = Split by 2 dvd-r. CCE is a great encoder, but at bitrates below 3000 k/bit, i would need a miracle, not a encoder. :)

CCE is a great tool that can (and should) be used with other great tools. I've used it last time to backup 'Simple Minds - Seen The Lights'. Why ?
- The audio of this 2 hour disc is recorded in DolbyDigital 2.0 at 448 kbps. I've re-encoded the audio as 192 kbps, saving a few hundred MB
- Even with smaller audio files, the compression ratio was huge
- On top of this i kept the original menus.

The result was very nice, but i'm not going to tell you how much time i spent on this process and how many tools i've used... :rolleyes:

THAT is what i try to avoid as a regular basis...
...because dvdshrink can give me the good results i want.

ps: As a side note, i see all my movies on a 32" 100Hz TV, and i'm seated at about 2,5 meters from it.

jNA!

3rd April 2004, 20:22

http://dvd.box.sk/articles8.php

This article shows that Intervideo DVD Copy is the winner in terms of quality in encoding the matrix.
What I find weird is that it uses single pass + some unipass technology.
Since reading it I tried the demo and it was great quality and fast.
However the control freaks ;) won't be happy as there is other than audio, sub, and chapter control there is absolutely nothing.

I would like someone to encode the matrix in cce and post the same frames to compare. I've tried using DVD-RB but i'm having no luck because there is a bug in it =/

nwg

3rd April 2004, 22:51

That test has been posted on many forums including the offical Shrink forums.

The whole test is a joke. They use very old software and even get the Shrink version number totally wrong.

jorel

4th April 2004, 08:13

Originally posted by Doom9
It's best if you just try it. I'd say CCE blows away every one click tool, others will violently contest that but my eyes can definitely spot the difference, whereas other people seem to have less sensitive eyes ;)

i "round and round" but the answer is here!

HanSolo00

6th April 2004, 21:27

I'm not really partial to either method (transcoding or re-encoding) though until recently (DVD-RB) I preferred transcoding for speed and reliability reasons.

I have a project that I've done in latest Nero Recode 2 (basically DVD Shrink) with Deep Analysis, and the same done with 6-pass CCE v2.50 in DVD-RB. While watching the transcoded version on the TV, I didn't really notice anything out of the ordinary, however when I paused the image during high motion shots, I immediately noticed the blockiness in the picture--giant pixelated ugliness. There are no such anomalies in the same frames of the 6-pass CCE encoded version, and in fact I have to look at frame captures magnified on my computer compared to the original DVD in order to even tell a difference.

Prior to DVD-RB I would say using the big three simply isn't worth all the headaches, time, cost (for tools), and projects that failed to convert. For that much effort you may as well buy another copy of the original, unless your time is worthless.

DVD-RB is going to make life so easy for everyone wanting to use CCE to do quality (full) backups.

monomer

7th April 2004, 18:18

Let me start off by saying that I too have also believed that you must really 'stress' a transcoder to begin to see any differences between them. And in a complete and fair comparison, they must also be judge both "in real-time motion" as well as looking at some "stills" for comparison. Obviously your display system will have a great effect on being able to distinguish any of these differences. When "stressed" enough all transcoders will begin to show "pixellation" and "macroblocking" in dark stills, shadow areas and with fast-motion... and "motion artifacts" will show where backgrounds with high detail are in some motion... and all will appear just a bit "softer" (shifting focus) in places.

Now I'm a definite believer in using transcoders for "light" amounts of compression depending upon your display system (...for me that's about 90-85% of the original's size). I believe that really small compressions (say no more than 90% could conceivably deliver a more accurate picture quality than an encoder plus transcoders are just a whole lot faster).

But I looked very carefully at a random four sets of pictures in the above link and something strikes me as very "spooky" about that "comparison study". In each case I could not find one single pixel out-of-place between the originals and the "Intervideo" examples and that's zoomed to 300%... I especially searched the dark places and backgrounds where very slight patterns exist. Not on single pixel is different. This is either the most amazing transcoder ever created that's doing the impossible or the comparison was "rigged". From all my experience making over 600 back-ups (several of them being multiples done for my own personal comparison studies) I will have to admit the skeptical/suspicious side of me is being drawn out after viewing that comparison link... I'm really leaning toward it being a product of devious marketing forces at work or that it was only meant to be a prank.

Of course the only way to be sure is to verify by replicating the comparison "study". I own The Matrix (NTSC), DVD2One(all versions), Instant Copy (both 6 and 7), DVDShrink(all versions), Recode... but unfortunately I do NOT have the Intervideo transcoder... so a verification of the study's results would not be possible for me without my purchasing Intervideo's transcoder, which I won't since my gut tells me the linked study is "bogus". Could some one else try the Matrix with the Intervideo transcoder and post back to this thread with your results??? I am very curious to know if it's by far the World's Very Best Transcoder or is it all just a lie? If it is the World's Champion Transcoder, I will then purchase it just to find out how far it can be "stressed" before ANY change of a single pixel can be detected on my set-up.

peachee

11th April 2004, 05:33

Transcoding is a quality deadend. MPEG encoders are already having a hard enough time pushing it to the limit to keep down macroblocks and mosquito noise. When you strip away compression information from a MPEG stream--a la transcoding, you are always left with a far worse product than from a competant MPEG encoder (CCE, Procoder, TMPGenc). This is obvious in theory. But I witnessed it in fact as well.

There's no need for deep comparisons of pictures between CCE and transcoders. One look on a computer screen and I can tell if a DVDR was transcoded. Transcoded video looks horrible.

I've tried the best transcoders IC7, IC8 and DVDShrink. All of them UNACCEPTABLE. I saw tons of blocks when things (character, background) started moving. It was almost like watching an amateur encoded VCD. And we all know how some VCD people used to claim that VCD looks no different from SVCD and almost as good as DVD. In other words, those are BS statements by amateurs!

Even with 80-85% quality level (15-20%) compression--which is not possible with typical 7-8 GB dvds (you need 40-50%+ compression), I can see blocks in moderate speed motion scenes. Remember, you are throwing away essential compression information when you transcode--there is no way it can look better than the results from a MPEG encoder. Even my bitrate starved CCE encoded SVCDs look better than transcoded dvdrs.

Conclusion: Transcoding fans=amateurs. After all it doesn't take much effort or knowledge to push the "Start" button to transcode. There's no need to learn about framerates, IVTC or about maintaining the same color space or about bitrates and VBR.

Hence, I've stayed away from Transcoders. Sure wouldn't pay for one nor try another since I know it's a deadend in theory and in fact.

Last, sure Transcoding is acceptable if you only watch dvdrs on non-HDTV TVs. But on computer screens it looks real bad ... real amateurish. Also, HDTV will be reality in homes in a few years. All those transcoded dvdrs will look horrible on HDTV and you will have to do it all over again with CCE.

So, if your gonna do the job ... Do it right the first time.

Joergen

11th April 2004, 05:45

peachee: you're being too harsh there. I've used mostly dvdshrink since it came out and never compress under 78% (usually aim for 80%+) and I can open two DVD2AVI windows in RGB mode and examine single frames on my 21" monitor and usually just cant "spot the difference".

Transcoding is a great tool for that 20% extra space. But occationally there ARE DVDS that cant handle even 20% (often disney stuff seems to be this way, where the original is hanging at the edge of quality, jam-packed with language streams and interleaved with extras).

luphy

11th April 2004, 06:20

More than a bit harsh, but to each his/her own. You should note that a lot of folks who use transcoders are quite proficient with the Big3 method. They choose transcoders for their time-factor. Obviously transcoders will never approach re-encoder quality, but it's all a matter of amount gain per time.

True, HDTV will be commonplace soon and those transcoded and even re-encoded movies will show their defects. But hey, dual layer discs are here, and even higher capacity discs are on the horizon. So no need for transcoders or re-encoders.

So everyone's efforts are futile. Life is futile. But hey, I have nothing else to do heh.

insanescape

11th April 2004, 16:13

Also, HDTV will be reality in homes in a few years. All those transcoded dvdrs will look horrible on HDTV and you will have to do it all over again with CCE.

Of course you mean by then we'll be using our dual-layer burning to just do a simple copy? :)

jorel

11th April 2004, 16:49

@ peachee

after read your post,
means that if you watch a movie encoded using CCE
with the right .avs script and filters for the case...
and if you never saw this movie before,
do you know that was encoded? :rolleyes:
--> maybe....
if you saw the movie before...maybe!

for me that work with eletronics for more than 30 years
and see images from lots of labels/models all day long is
--->impossible! :o

and my friends call me "bored for audio & video",
i can't stand "more or less" quality.

waiting your sincere answer,please! ;)

remember that this thread is called:
DVDshrink quality compared to CCE...
thanks in advance !

peachee

11th April 2004, 19:41

I can tell if a dvdr has been transcoded (compressed using IC7, IC8, DVDShrink, etc). It's not difficult if you watch dvdrs with a computer monitor. Transcoded dvds show lots of blocks on a computer screen. I don't have to pause to see these blocks.

Can I tell if a dvdr has been encoded with CCE? Well, that's a bit more difficult without pausing the screen and looking for mosquito noise. But, that's not what I said. I said I could detect transcoded video (IC7, IC8, DVDShrink, etc.). And my conclusion was that CCE is far better because CCE doesn't produce glaringly visible macroblocks very often (instead produces minor mosquito noise which is not as noticeable).

Also, I have rented a silver dvd (looked professionally printed) that I know has been re-encoded by bootleggers. This was a foreign movie rented from a little video shop. Scenes on the dvd were slightly blurry which made me suspicious and when I saw significant blocks in a couple scenes, I knew it had been re-encoded--probably with Tmpgenc (produces more blocks than CCE).

If Phillips and Pioneer doesn't have any big obstacles, when HDTVs are out in abundance, dual layer burners will be out and dual layer media will be more affordable (and hopefully won't be anymore incompatible than current single layer dvdrs). But, all your 100s of transcoded discs will be a waste in current money and time. You wouldn't want to show them off to friends and family on our HDTVs because they looks so bad and you will have to throw them away ... polluting our enviroment and filling up the landfills. Why waste time, money, and environment when you can do it right with CCE?

My SVCDs made from CCE still look good on a computer screen and hence will look good on a HDTV. My 2 transcoded DVDRs (one made by IC8 and one by DVDShrink) look horrible on computer screen and will look horrible on a HDTV.

jorel

11th April 2004, 20:22

:)
peachee wrote:
"Even my bitrate starved CCE encoded SVCDs look better than transcoded dvdrs."
and
"I can tell if a dvdr has been transcoded (compressed using IC7, IC8, DVDShrink, etc)."
more:
"Can I tell if a dvdr has been encoded with CCE?
Well, that's a bit more difficult without pausing the.."

it's right, got the point, i can write the same!

;)

geffroman

11th April 2004, 23:12

Originally posted by peachee
Transcoding is a quality deadend.
So, if your gonna do the job ... Do it right the first time. You miss the point in so many many ways...

First DVDs are not all the same... Some can take massive amounts of compression while others fall apart under very low compression...

Most (99%) of DVD can be transcoded wonderfully, if you make space for the video to be an appropriate output quality... If however, you just stick the disc in and use the one click method, leave all the crap on the disc you will get crap results... If you do it properly you will get amazing results...

If you remove the languages, advertising, studio trailers and still see poor results then you need to decide if you should simple put all extras on a second disc as a bonus disc...

On rare occasions a two disc set splitting the movie is called for. The studios do it on movies like Schindlers list so I don't think it crazy for me to take the same approach... especially with tools like DVDReMake around that can insert clips at the end of disc one that say "Insert Disc Two to Continue".

Bottomline, If you do very little in the way of thinking and testing, then just push the button on the one click solutions, you will get CRAP... Then you can compare it to a very lengthy complicated CCE work flow that can take an entire day to produce one back up and claim quality superiority... You got what you tried to get which is not very impartial.

Or you can learn how to use transcoders to get amazing quality DVD-Rs now and in the future... and do it in about an hour per movie from RIP to STRIP to ENCODE & Burn!

Personally I want to know when and how much I can use and compress with different tools... In my mind it is never ONE SOLUTION fits ALL... Some situations I really want CCE though it is RARE... Other times I use DVDReMake and DVDSHRINK... Some times I spend time in IfoEdit... In all cases I get a great result... because I have learned what it takes to get one... I don't use a Ferrari to drive to the lumber yard and I don't take a dump truck on a date... But you have go to know what works and you only l;earn that by owning DVD+RWs and taking the time to learn...

We all know CCE will ultimately produce a better result than one click encoders if all things are equal although things are rarely equal. Many times CCE lovers comparisons are made with ZERO extras and sometimes MOVIE only versions compared to One Click solutions containing EVERYTHING... How fair is that...?

Each movie is different. Each one requires a different approach. But in EVERY CASE it takes at least 3 to 10 times longer to make that result with CCE based workflow. CCE if done side by side with One Click is better... but sometimes NOT visibly better if at all... But, is it worth the 10 hours of Transcoding time for CCE for most people. If I want a DVD backup and can afford 10 hours, trust me I can afford to simply buy an EXTRA store bought disc...

Originally posted by peachee
Last, sure Transcoding is acceptable if you only watch dvdrs on non-HDTV TVs. But on computer screens it looks real bad ... real amateurish. Also, HDTV will be reality in homes in a few years. All those transcoded dvdrs will look horrible on HDTV and you will have to do it all over again with CCE.

So, if your gonna do the job ... Do it right the first time. HDTV will not make a DVD-R look WORSE than it did on a regular TV... That's just a foolish remark... In fact in most cases the HDTV circuitry won't even be displaying in HDTV mode. If it's plasma it may up sample the source to 720p or 1080i depending on your make, model and quality of HDTV... This will NOT enhance problems... If it's Picture tube HDTV it will most likely display in native 480i or 480p mode... Your DVD player cannot even output an HDTV source...

Originally posted by peachee
Conclusion: Transcoding fans=amateurs. After all it doesn't take much effort or knowledge to push the "Start" button to transcode. There's no need to learn about framerates, IVTC or about maintaining the same color space or about bitrates and VBR. This remark is just ill informed and rude. You clearly don't like one click Transcoding. You appear intelligent yet you are wrong so I can only guess that you have not taken the time to really study what needs to be done to get a good result with a Transcoder. You don't know what decisions need to be made along the way to end up with a great finished DVD-R using a trancoder. Instead you quote old school terms like Frame Rates & Bit Rates and have refused to adopt solid NEW solutions because they are TOO easy for your liking... I don't think the goal was EVER for it to be complicated... It WAS complicated in the early days of DVD back up because it was THEN... This is now... you just appear to be inflexible...

By the way all transcoders are NOT the same. IMHO DVDShrink does the best job of them all. Less bugs than IC7 or IC8 and better quality and easier to use than all the rest. Again, IMHO.

Zhnujm

12th April 2004, 01:02

Im just wondering why this discussion is still around...
Everybody could easily check for himself whats best (for him).

My own conclusion is:
A CCE reencoding will always look better than a transcoding, no matter if its a 8->4GB or a 5->4GB reduction.
A higher reduction will make it only more visible.

Joergen

12th April 2004, 01:05

This is why we dont have "which is better" threads. And the guy above is a good example. Even the experts agree that transcoding yields better results in 10-15% reductions than re-encoding if the source is well mastered studio material (as it usually is).. not to mention how much time and electricity you save.

peachee

12th April 2004, 03:26

I guess there are more non-perfectionists on this board than just one year ago. I hope this board doesn't turn into a massive newbie board that is dvdrhelp.com. :)

I guess you can achieve 10-15% compression if you strip out everything else, extras, languages, commentaries, trailers on some dvds. But, that would suck as it makes the dvd experience non-interactive. I guess if you just want the movie you can do that. But, to me, I'd like to keep the extras and additional soundtracks without making part or all of the dvd look ... blocky. Yes, that's why I think when I hear transcoders ... blockiness. You can see blocks in all the pictures of transcoded video in this thread.

Eheh, the commedian above is funny. You might even convince some that you are serious about transcoded dvds looking good on HDTV comments.

One question to those who still use transcoders and are convinced nothing beats it in quality: Why do you think people still aren't satisfied with transcoders? Why do you think everyone (pros that is) is cheering jdobbs with his achievements? Why do you think people can't wait for dual layer burners?

Answer: Because transcoding looks bad on computer screens and HDTVs. Blocks everywhere with transcoding.

I'm sorry if you're too non-industrious to learn how to use CCE but you are really missing out on quality and you are wasting time and plastic. And if you really can't see the difference between transcoded video and non-block video (original dvd or a proper CCE encoded dvd) then I seriously suggest you see an optomitrist.

ddlooping

12th April 2004, 03:54

Hi all. :)

I might not have an HDTV or plasma screen but I use a 17" Eizo LCD monitor.
It is/was supposed to be the best in its class a couple of years ago (specially in terms of color reproduction and reaction time).
I use it at its native resolution of 1280x1024 - which is better than most HDTV - to test my backups.

I can honestly say I have never seen a difference between originals and their backups when using a ratio of 90% or more with DVD Shrink, using Deep Analysis.
I'd say down to 85%, still using Deep Analysis, you would really hard-pressed to see a degradation in quality.
You just might if you stop watching the movie (which is what it's about) and start pausing, zooming, looking at backgrounds, etc etc.
Below that, it depends highly on the quality of the original title, but between 80% and 70%, CCE (or splitting to two DVDs) is as far as I'm concerned the way to go.

Following the previous rules, about 85% of my backups are made using DVD Shrink (90 minutes inc. deep analysis/burning), 10% with DVD2DVD-R (7-9 hours/overnight or while at work).
The 5% left I don't backup and will buy another copy if my original gets damaged. ;)

jorel

12th April 2004, 04:31

@ peachee
my "old" question still remains.....(and without answer):

"after read your post (again), and the last too,
means that if you watch a movie encoded using CCE
with the right .avs script and filters for the case...
--->and if you never saw this movie before, do you know that was encoded?
--> maybe....and only if you saw the movie before...maybe!"

...but "without pausing" like you wrote cos if you are using pause,
you're searching problems and not watching the movie.

some dvds are very bad and
---> cce encodes from good dvds sources are better than this ugly dvds.
example of bad dvds? life of brian....
this dvd have tons of mosquitos...and some others too (lots)

and again i remember you,
this thread is called:
--> DVDshrink quality compared to CCE...
;)
still waiting your answer for this question!!!
only to remove doubts, ok?

monomer

12th April 2004, 05:00

Originally posted by peachee
...My 2 transcoded DVDRs (one made by IC8 and one by DVDShrink) look horrible on computer screen and will look horrible on a HDTV.

Ahhh, now I understand. You've only tried it twice and evidently tried to include all the extras...

You missed the part about transcoders being very good if you only compress no more than 10%-15%. You need to actually try one of the transcoders with a movie in the smaller compression range and then you can have an informed opinion. BTW, my display is a LCD front projector with a native resolution of 856X480 (that is 'DVD resolution' which means the picture can't get any better even at HDTV resolutions) and my projection screen is 10-FOOT wide... and it's all being fed by my computer, pixel mapped 1:1 ...I can NOT distinguish (even going frame-by-frame) between a 10% reduced transcode and the original... which makes your claims a bit hard for me to swallow.

Oh, and about wanting all that other stuff on your back-up copy... if you actually owned the DVD, you could have the 'interactive' part anytime you wished... BUT many of us make back-ups just so we don't have to put up with the menu, FBI warnings, etc... I want my movies to start the moment I 'pop' one in... no trailers, menus and fumbling with the mouse to get to the movie. Its obvious you're only interested in doing a FULL back-up (probably because you don't own the original) which mean massive compressions on the order of 60% of the original's size... it that case you are right, re-encoding is the only reasonable choice for YOU.

geffroman

12th April 2004, 05:14

Originally posted by Zhnujm
Im just wondering why this discussion is still around...
Everybody could easily check for himself whats best (for him).

My own conclusion is:
A CCE reencoding will always look better than a transcoding, no matter if its a 8->4GB or a 5->4GB reduction.
A higher reduction will make it only more visible. Because some people... in fact a lot of people come here to learn from others experiences rather than create the wheel all over again... Unfortunately, when experienced old timers to DVD backup refuse to learn new software and spew out of date advise without having direct knowledge of tools they choose not to employ it ruins the opportunity for a newbie to get GOOD advise...

jorel

12th April 2004, 05:21

@ monomer
:D
you was faster when you wrote about:
"(probably because you don't own the original)"

like you, i encode cos
"I want my movies to start the moment I 'pop' one in.."
from the dvds that i own, the extras, fbi, menus, etc...
stay in the "original", not in my (our) backups !

i was waiting more 2 or 3 posts from peachee...
in the end he will confess it! (no way out)
:p

peachee

12th April 2004, 10:49

Jorel, what are you asking about? Sorry but your merry Portugese English is confusing.

To the others stuck on Transcoding due to newbieness. Yes, I have backed up a movie with IC. It looked nice on TV. I thought wow. It looks great, why bother with CCE. Then I popped it into my computer. And I saw the ugly blocks in almost every scene. I did strip out the extras and removed the other language and audio tracks in order not to stretch the limits of the IC transcoder ... but it still produced subpar quality. IC 8 Grade C-.

With DVDShrink, I tried to shink a 6 GB foreign movie into a dvdr (I used the still picture option for the extras). It looked even worse than the result from IC 8. Lots of blocks. I then tried reducing the compression and encoded a trailer as a test. It looked bad at 20-30% compression. I tried 15% compression and I still saw unacceptable number of blocks. There is no way to shrink this movie using only 10-15% compression anyway (most of my movies need 30-40% compression). This is why I don't bother with Transcoders. DVDShrink Grade D-.

Also, I do own the dvds (Shaw Brothers classic movies by Celestial Pictures) that I back up. It's not a mere inconvenience of watching FBI warnings that cause me to back up my dvds. The thing is I can't watch the movies on my home theater dvdplayer at all. The reason is I live in Region 1. The movies I buy are Region 3 locked. They have great extras and rare actor interviews that I would like to watch on my home theater. But, my premiere home theater dvdplayer is Region 1. This is why I need to back up my Region 3 movies (into region free) completely without stripping out extras and with good quality.

Last, lots of dvds are coming out as Anamorphic which uses 6-7-8 GBs for the main movie. How do you expect to back up those movies with 10-15% compression?! It's not mathematically possible.

At least with CCE I can handle all movies without reducing the video quality to D- level quality. CCE video is at worst B quality (but usually B+ to A-).

jorel

12th April 2004, 11:57

;)
ok peachee, stay cool!
this is a forum.....here some(few) people teach, some people(lots) learn like me
and all change informations. (and i'm doing friends here too)

now i understand your target doing backups from dvds from another region.
you encode(or backup) the extras cos you want it...
i and most number of people here, remove the extras and remainders,
encoding only the main movie....
and it's good cos without menu you got full "autoplay" ! ;)

if you don't saw my post here, please take a look :
http://forum.doom9.org/showthread.php?s=&threadid=55844&perpage=20&pagenumber=6

i have toy story 1 & 2 with ~1,70Gb and ~2,15Gb using CCE Q10
...amazing quality without blocks, mosquitos,etc...
with ac3-5.1 from source, i still have free space in the dvd-r
and is really better than using dvdshring with 52,2% and 49,7%
compression for the same target (full dvd-r, no free space left)!

i'm thinking in reencode using CCE Q25 to include "bugs life"
in the same dvd with toy story 1 & 2 with ac3-5.1(1 language) from source.
a huge compression is needed to do the same job
using dvdshrink and the result will be useless!

i get better results using CCE than dvdshrink! (obvious)

best regards !
:)

... merry Portugese English ?
lol (kiddin)

ux-3

12th April 2004, 12:24

It happened to me a few times that I made a complex encode with CCE, multipass, Macroblock optimization and all. In the end, I came out about 8 MB above DVD size.:angry:

So I took a quick look to the right and left and loaded a transcoder, and feed the oversized encode through. :D

So what do I want to tell you? ;)

Zhnujm

12th April 2004, 12:37

Originally posted by geffroman
Because some people... in fact a lot of people come here to learn from others experiences rather than create the wheel all over again... Unfortunately, when experienced old timers to DVD backup refuse to learn new software and spew out of date advise without having direct knowledge of tools they choose not to employ it ruins the opportunity for a newbie to get GOOD advise...

No comment.
You can find enough results in this thread, things havent changed until then, not with DVDShrink and not with Instantcopy 8.

dragongodz

12th April 2004, 13:39

jeez guys give it a rest already.

look different people want to use different tools and accept different levels of quality in different cases. so you didnt like the results you got on you tests ? get over it and stop trying to ram your point of view down others throats.

I guess there are more non-perfectionists on this board than just one year ago. I hope this board doesn't turn into a massive newbie board that is dvdrhelp.com. :)
such insults (even when followed with a smily) wont win any freinds.

Conclusion: Transcoding fans=amateurs.
Why do you think everyone (pros that is) is cheering jdobbs with his achievements?

people backing up their own dvds are ALL amateurs unless you are doing it for money.
conclusion: people calling themselves "pro" are rather full of themselves and think over highly of their opinion. meaning you are wrong unless you do things their way exclusivly. :D

just use and do what you are happy with and ignore others saying you shouldnt.

now can a mod PLEASE close this as it is getting nowhere and the "stuck in stone" mindset is blatantly obvious ?

jorel

12th April 2004, 15:23

dragongodz wrote:
"now can a mod PLEASE close this.."
i can't see reasons for that.

you post your "conclusion"....
it means that you have the "final point" ?... it's strange!

or can we stay posting our opinions like you?
(it's not a insult, only a question)

:) (it's not a insult too)

we can discuss without fight, right?
this is one of the advantages of the forum (i think)

thanks.

dragongodz

12th April 2004, 15:38

jorel - if you look at the laste few pages you will see this is not going anywhere. the same things are basically being said over and over and nobody is likly to change their mind.

people have already posted their opinions again and again so ye go ahead and post away but if people cant see all the different points of view after all these pages then i seriously doubt they will with any more.

as for my having the last say, doesnt matter to me at all who has it, actually if a mod closed this then usually he would have the last say. :)

constant repetition when peoples opinion isnt going to change is rather pointless though (IMHO) but hey if you guys want to beat each other over the heads with a dead horse(to paraphrase a saying) thats up to you. :)

ChickenMan

12th April 2004, 15:52

Originally posted by dragongodz
now can a mod PLEASE close this as it is getting nowhere and the "stuck in stone" mindset is blatantly obvious ?
Yes, I agree. Everyone has had their say and the whole thread is going no where. So before it gets into a real slagging match, the thread is closed.

jorel

12th April 2004, 15:55

:p
ok my friend...
i "see" your point of view looking througt this direction and you're right!

:)

insanescape

12th April 2004, 17:44

I never quite understood why someone would get all huffy because someone they don't know is copying a dvd and playing it on a TV they'll never see. If they enjoy it, let them be. Just sit at your desk and laugh hysterically to yourself when you realize that you're perfect and there's no possible way people could ever enjoy something you don't agree with. It's so comforting being superior and right, I'm surprised you even have the desire to post at the end of the day.

peachee

12th April 2004, 20:14

This thread asked for a comparison between DVDShrink and CCE.

I gave my take which has been confirmed by those who posted screenshots on this thread.

Conclusion: All transcoders create blocky output (visible without pausing) including the 2 best transcoders, IC and DVDShrink. CCE produces far better output because CCE keeps blocks at a bare minimum (not visible with pausing).

All these people expressing hurt feelings and claiming that video quality between CCE and DVDShrink is purely opinion based are not facing reality. The reality is: CCE is better in quality, much better. This is a fact proven over and over again (look at the screenshots in this thread). Where are the screenshots that show DVDShrink looks better than CCE? Locked up in heaven? Sorry, but we are on Earth.

Before you claim again that DVDShrink produces better quality, post the evidence. So far, all the evidence (screenshots) posted in this thread show that CCE is better.

Sorry if feelings are still hurt but this isn't emotional support forum. This forum is about quality video encoding which is supposed to be about facts, not about promoting lies to make newbies feel happy about their beginner efforts. Besides, how else will newbies ever improve if we are telling them that they are doing just as good as the pros and that there's no need to improve?

geffroman

12th April 2004, 20:31

Originally posted by dragongodz
now can a mod PLEASE close this as it is getting nowhere and the "stuck in stone" mindset is blatantly obvious ? That is not right... Future readers will stumble onto this thread and may want to engage in this discussion again... When peachtree first made his point it was believable to the uninitiated and could cause newbies to be affected by his remarks... Enough has now been said (in this OPEN forum) and with peachies following posts that a reasonable person would assume that there are more than one point of view and they will test these things for themselves. All sides have been heard... a few times... Any reasonable person got the point... the conversation has come to a conclusion... censorship by a moderator is not the answer as NO ONE has attacked or used profanity... some ignorant and slightly rude things were said... but don't you see, that too helps newbies decide who to trust...

I say leave the thread alone... it has served this crowd well in the past few days and could help others again in the future who themselves feel the need to discuss this subject... who are WE to tell a few newbies 2 months from now what they can and cannot discuss...

insanescape

12th April 2004, 20:37

peachee: This thread asks the question of how shrink compared to CCE. Answering is one thing, but throwing your ideology around and basically telling people they're worthless if they don't do it your preferred way is going a bit beyond the scope of the question.

Your evidence I'm sure is appreciated by anyone willing to back up their DVDs. You should probably respect their ability to take facts and dtermine what would be best for THEM opposed to belittling their obvious hobbyist efforts and treating it like if they went with what you see as inferior that they're somehow ignorant and undeserving of a dvd burner.

Joergen

12th April 2004, 20:43

Well said insanescape. This ignorant fool bought the A03 when it cost over 600 dollars a couple years ago, and a DVD-R cost $10. I did a few movies by means of dvd2avi, vfapi, cce and reauthoring them in dvd maestro, subtitles and all.

And now I happily (and what a wonderful tool it is) use shrink whenever it yields proper results, but am very excited about dvd-rb because it finally gives CCE's superior quality (at maximum compression) without wasting hours of my time doing it.

ddlooping

12th April 2004, 21:45

Originally posted by peachee
Before you claim again that DVDShrink produces better quality, post the evidence.
I couldn't be bothered re-reading the whole thread, but I don't actually think anyone claimed DVD Shrink produces better quality.

What the general concensus seem to be is that at low levels of compression and using Deep Analysis, the DVD Shrink backups are indistinguishable from the originals.
Why then spend hours with CCE when the job can be done in less than one and to obtain the same result? ;)
Originally posted by peachee
Last, lots of dvds are coming out as Anamorphic which uses 6-7-8 GBs for the main movie. How do you expect to back up those movies with 10-15% compression?! It's not mathematically possible.

Pardon my french, but this is a lot of b*ll*cks. :)
I'd say that most DVDs are and have always been anamorphic (compatible with 4:3 screens).
The bitrate at which they were encoded is what decides their overal size.

I do understand your need to use CCE if your backups are "full-disc", as most of them will require a fairly high level of compression.
DVD Shrink as most transcoders will introduce noticeable degradation.
How noticeable being still dependent on the quality/bitrate of the original DVD and the viewing equipment.

Zhnujm

12th April 2004, 23:51

Picture where DVDShrink is better than CCE: :)
ftp://zhnujm.dyndns.org/Image4s.png

5.0->4.4GB reduction
Top left - Original
Top right - DVDShrink
Bottom left - CCE
Bottom right - Intervideo DVDCopy

Dont judge the DVDCopy result, it is a 1:1 copy of the original.
Thats also an important point in comaparisons, most transcoders do not change all frames, at least not for a low compression. So if you compare such frames you will obviously get VERY good results...

Also i want to quote this:
Originally posted by Zhnujm
You should think more positive, at least its already on page 5 WITHOUT ending in a flamewar. :D

We are now on page 8 - maybe some more can follow.

dragongodz

13th April 2004, 06:56

who are WE to tell a few newbies 2 months from now
if a newbie cant follow all thats already been said(and testing for themselves) then they could easily open a new thread asking for clarification on any pionts they dont understand couldnt they ? there is no 1 thread for CCE or 1 for DVDShrink etc is there ? :)

peachee - you still do not get it. people are NOT claiming that dvdshrink output is better than cce(except maybe the odd anomoly such as Zhnujm posted). what they are saying is dvdshrink output for when you have to only reduce the final size by a relativly small amount is indestiguishable from the original playing. yes playing, i dont know anyone that watches a movie at 1 frame per second.

Yes, I have backed up a movie with IC. It looked nice on TV. I thought wow. It looks great, why bother with CCE. Then I popped it into my computer. And I saw the ugly blocks in almost every scene.
since you didnt get blocks in every scene on your tv ever considered it could be the mpeg2 playback program or codec that may be a problem ? i know lsx decoder on my system produced crappy playback. have you tried MediaPlayerClassic(by Gabest) which has its own mpeg2 internal decoding ?

Before you claim again that DVDShrink produces better quality, post the evidence. So far, all the evidence (screenshots) posted in this thread show that CCE is better.
so lets see we have no doubt thousands of people happily using dvdshrink with no complaints about its quality for low compression against a few saying it produces crap output no matter what amount. hmmmm so all the people saying they have had fine output from dvdshrink are lieing unless they post screenshots are they ? :devil:
now you will notice that dvdshrink has had bugs(guess what so has cce) and some dvds will no doubt look/work better than others but your sort of blanket statement just does not hold.

Sorry if feelings are still hurt but this isn't emotional support forum.
no its also not a forum for.......better not say it. :sly:

This forum is about quality video encoding which is supposed to be about facts, not about promoting lies to make newbies feel happy about their beginner efforts. Besides, how else will newbies ever improve if we are telling them that they are doing just as good as the pros and that there's no need to improve?
and again its calling people liers because they dont agree with you because they have not had the same results you have.
and again its the "pro" bull. do you do dvd backups as a business or are you just backing up some of your own ? if its just your own then guess what, you are an amateur matey so please drop the pro/expert/elite style crap as it may impress some newbs but thats all.
and finally who the hell do you think you are to tell people whether their backups are good enough ? if someone is not happy with the quality they get then they ask for pointers to try and get better results. you can also do things like guides to show people how you think better quality can be done etc. however when you start trying to tell people their backups, they are going to watch in their homes, are not good enough or not "pro" you cross the line from being helpfull to being a....(sorry i tried a few different descriptions but i would probably get a strike for most of them).

Zhnujm -i am thinking positive, i am thinking positive, i am thinking positive. i am positive this is not going to get any better. :D

geffroman

13th April 2004, 07:04

Originally posted by dragongodz
Zhnujm -i am thinking positive, i am thinking positive, i am thinking positive. i am positive this is not going to get any better. :D Not with posts like that

dragongodz

13th April 2004, 07:24

hmm so i mention that peachee should possibly test to make sure his pc isnt the cause of bad viewed dvdshrink output ,that he shouldnt call people liars because they dont get the same results and he has bugger all right to tell someone what THEY must find acceptable and i am in the wrong ? ye, right, whatever.

or was it the comment that a newbie can start a new thread later i wonder ? :)

wmansir

13th April 2004, 08:59

For the benefit of all I am going to remind everyone forum rule #4 is:

4) Be nice to each other ...

That starts with not disrespecting other members. It also includes not labeling those whom you disagree with (amateur, stupid, lazy, perfectionist). In other words, respond to the issue, not the person.

The purpose of this forum is to communicate ideas and information. That goal is severely hurt when people are insulted, agitated, or intimidated. If you feel someone has crossed the line, PM a mod. Do not respond in kind.

Earlier today another Mod told me he has only given out 2 strikes in 3 years. I would hate to tie him in my first 3 days.

jorel

13th April 2004, 09:08

wmansir,
thanks for wise words and recomendations!
that will take all back to the topic.
:)

Rassilon

13th April 2004, 09:12

Originally posted by mpucoder
@Fmazzanti I'll bet you work with PAL. It's a totally different story with NTSC.

How is that mpucoder?

So does that mean that If I'm converting PAL I use one program, and for NTSC another??

jorel

13th April 2004, 09:18

i have doubts if pictures can show the general quality.
pictures only show the details in the frame.

like Zhnujm posted:
5.0->4.4GB reduction
Top left - Original
Top right - DVDShrink
Bottom left - CCE
Bottom right - Intervideo DVDCopy

if i right understood means that
original = 5Gb
DVDShrink = 4.4Gb
and about the size from CCE and what parameters was used to encode?

idea:
can you post 1000 frames from the original and the same from dvdshrink?
i will use the frames from the original to encode in CCE
to see the quality and the size!
then, see the real general quality in the movie and not in only one picture!
;)

can you post it please?
thanks in advance!
:)

YaoMing11

13th April 2004, 10:44

I just read this thread through and the only thing I can add is DVDshrink looks blurry and CCE looks sharp to me. It's as simple as that when viewing a movie made from both with my own eyes. I think thats enough right there to turn me away from dvdshrink. This is just my own opinion I wanted to add.

ux-3

13th April 2004, 11:46

Originally posted by Zhnujm
Picture where DVDShrink is better than CCE: :)
ftp://zhnujm.dyndns.org/Image4s.png

I just had a look at them. A 1+ minute look. I was almost about to give up, and declare them all equal, when I decided the CCE picture is less colorful. Then I shifted it more into centerscreen, when suddenly the original seemed slightly less colorful. So I was really judging the variation of the brilliance of my display and not the quality of the encode. I am not saying that there is no real difference, but without a clou, I am not able to find it. And I don't spend so much time on each frame of a movie. So I guess I do as most people here do, I use the tool I consider best for the job. And I let myself be guided by other factors than just image quality. If I suspect alien subtitles, I much rather transcode than reencode, simply because it takes ME to long to get the subtitles to come out the way I want them to.
As the tools change constantly, so are my choices. Until recently IC7/8 was a resonable choice for me to do some episodes, unless I wanted to spend my life on them. Now I would rather let DVD2DVD-R handle that job, or perhaps Rebuilder.

Sure I like to use CCE for longer movies, because the viewing impression is better. I recall having argued perhaps 1.5 years ago, that picture comparisons are only a part of the story. The transcoder problems become aparent when you compare sequential pictures, where you get fluctuations in unicolor areas. This is not aparent in a single frame, yet very distractive. Still, I have absolutely no problem to throw my recent CCE encode into a transcoder to lose 20 MB, rather than running it again.

geffroman

13th April 2004, 12:34

Originally posted by dragongodz
hmm so i mention that peachee should possibly test to make sure his pc isnt the cause of bad viewed dvdshrink output ,that he shouldnt call people liars because they dont get the same results and he has bugger all right to tell someone what THEY must find acceptable and i am in the wrong ? ye, right, whatever.

or was it the comment that a newbie can start a new thread later i wonder ? :) The conversation was winding down... you spewed a bunch of REPEAT stuff and started saying negative things and distorting what was said by the offender (he never CALLED anyone a liar)... You didn't add anything new and turned this thread negative again... You are on the right side of MY argument but your tactics are not helping the cause...

Stux

13th April 2004, 17:05

The DVDShrink version was slightly blurrier and a little blockier

jorel

13th April 2004, 19:47

Originally posted by Stux
The DVDShrink version was slightly blurrier and a little blockier

right Stux!
;)

.... more "salt" :

CCE picture have more details in the water (bottom left),
more sharpness throught the trees on the right,
more details in the grass/bush and general vivid colors are better.
---> ...no need zoom to see that!
;)

ux-3

13th April 2004, 20:45

Originally posted by jorel
right Stux!
;)

.... more "salt" :

CCE picture have more details in the water (bottom left),
more sharpness throught the trees on the right,
more details in the grass/bush and general vivid colors are better.
---> ...no need zoom to see that!
;)

;) You know, you sound a bit like the people in those really expensive hifi-magazines. I am not saying that you are wrong, I am just saying that I wouldn't see it. I suppose the pictures are chosen as an extreme, since usually, I can make out the difference between CCE and Transcoders when they are side to side.

jorel

13th April 2004, 22:34

ok ux-3!
and you give a good idea:

if Zhnujm could change some positions of the pictures like this:

Top left - Original....Top right - Intervideo DVDCopy
Bottom left - CCE......Bottom right - DVDShrink

then everybody will see the differences easily!
;)

Zhnujm

14th April 2004, 00:10

Unfortunatly i dont have time today and tomorrow,
but for me the CCE pictures show some artefacts around the clouds at the top.

Over wide parts the picture quality at this low reduction is equal, but there are enough scenes that look blocky with transcoding, especially scenes that are a bit blocky in the original.
The trancoders seem to magnify this effect were CCE sometimes even look better than the original. I think i can post some scenes.
I dont say i can see this on my TV (surely not), but the difference is there if you look closely.

The CCE encoding just uses standard settings, no filtering, 2 pass vbr.
And is of course the same size as the transcoded files.

jorel

14th April 2004, 01:58

Originally posted by Zhnujm
Unfortunatly i dont have time today and tomorrow,

to change the positions of the pictures ?
ok, i (we) can wait!

Originally posted by Zhnujm
The CCE encoding just uses standard settings, no filtering, 2 pass vbr.
And is of course the same size as the transcoded files.

no matter what adjust, source, filters, resizes....etc,
used in CCE, i never got the same size of the source or transcoded file!
removing doubts: how you did it ?

please, again i ask you:
can you post 1000 frames from the original and the same from dvdshrink?
i will use the frames from the original to encode in CCE
to see the quality and the size!

thanks in advance!
;)

peachee

14th April 2004, 05:51

Nice pictures Zhnujm.

CCE does produce mosquito artifacts when you look at freeze frame shots at times. However, it rarely ever turns into blocks. And when the picture is moving (not paused), you will not notice mosquito artifacts with CCE. CCE looks great when the video moves.

With Transcoders (IC 7/8 and DVDShrink), I noticed very big and clear macroblocks. This isn't a codec issue, it's a high resolution issue. Regular TVs don't have high enough resolution to show all the details and so blocks aren't as noticeable. With HDTVs and computer monitor, you will see blocks. I viewed it with Intervideos' Windvd and with Media Center Classic. Blocks were very noticeable with Transcoders.

Last Note:
I think the issue isn't about which product/method is better (even though the question asked that and the answer is quite clear given the overwhelming evidence). The issue is about catering to the masses and average Joes and fluffing their self esteem (after they purchase consumer products). I've always been more of a perfectionist with little patience for 3rd rate products so that might be my curse in this imperfect world.

U.S. politics (and probably all politics) is similar. Lots of the middleclass and poor people believe in the top politicians and sweet tongued businessmen who woo them with sweet words ("God Bless America ... Home of the Brave ... Freedom is once again protected from those who hate us more than they love life ... No money down, really etc. etc.") on the tele and in papers/mags. Smart (alecky?) people who point out the sweet lies and expose the ugly problems (by showing the hard stark facts) are condemned and demonized by the propanganda machine (choice of weapon of the elitists who want to continue to take advantage of the poor masses). I guess this is the challenge of humanity is all about: that is, trying to help poor people see the truth (before it is too late) even though they ignorantly hate you due to the powerful propanganda of opportunistic politicians/businessmen.

A las, I can't change my nature in pointing out the truths of the world even though it often gets me on the bad side of the opportunistic wranglers of the masses. So go ahead, make me a martyr.

blueray

14th April 2004, 08:25

Instant Copy 8 vs CCE
After doing several large movies - Scarface, Eyes wide shut, etc...(movie only mode, requiring compression down to 60-70%) using both Instant Copy 8 (with hidden settings editor) and CCE (thru Dvd2svcd in dvd to dvd mode) I've found that each time I prefer the IC8 results to the CCE method. While neither measures up to the original exactly, IC8 produces sharper, more defined video than CCE (smoother, blurier). With Dvd2svcd, I use Image Quality Priority of 17, Bias of 20, and no filters. Am I the only one who prefers IC to CCE?

geffroman

14th April 2004, 09:25

Originally posted by peachee
A las, I can't change my nature in pointing out the truths of the world even though it often gets me on the bad side of the opportunistic wranglers of the masses. So go ahead, make me a martyr. You are no martyr... You have but 9posts at this momment... Most if not all have been condescending and rude... It is your attitude that sucks... not your opinion... It is your rude behavior that has sparked this discussion to negativity... I believe you like the conflict and attention not the truth... You made your point but you continue just to bate an argument... If it was truth you truly seek, you would have to look in the mirror and see your smugness, smallness and unpopularity... There would be real reckoning... No you are no martyr, no smarter, no fun, not entertaining, not educational, and no help... For if you had a valid point you make it moot with your tone... this has been clear as day to everyone but you... in the end you are only effective at self defeation... welcome to your charm... failure to communicate... IMHO

ux-3

14th April 2004, 11:10

Everybody please, remember

...rule #4 is: 4) Be nice to each other ...

Why don't we just ignore personal matters and stick to the issues?

pippocalo

14th April 2004, 11:53

When the compression is low (under 85-90%) DVDshrink is better for quality and speed.
Transcoders is better, when the compressin is low, because you don't have original of film but DVD that is already compressed. Therefore With CCE you must have decompression and re-compression. Transcoder may be better at low level compression because only transcoding process is applied and modification are very few.

At middle level of compression ( from 85% to 50%) you can choose between speed (DVDShrink) and quality/speed optimization (IC);
but if quality is your first choice CCE is mandatory.

Over 50% of compression there is only CCE.

geffroman

14th April 2004, 12:24

Originally posted by ux-3
Everybody please, remember

...rule #4 is: 4) Be nice to each other ...

Why don't we just ignore personal matters and stick to the issues? Tried that...

ux-3

14th April 2004, 12:26

Originally posted by blueray
With Dvd2svcd, I use Image Quality Priority of 17

IQP 17 to 24 I used for SVCD, with DVD I prefer 10 to 14, depending on material and bitrate. That should make the image look sharper. Just try it on a good source.

ux-3

14th April 2004, 12:31

Originally posted by geffroman
Tried that...

Well, judging from your previous post, you didn't try hard enough.:D

Ignore the crusade he is on, look at the bottom line he has to offer. If it isn't wrapped apropriately, consider the difference in # of posts and the style in other forums and let him learn this without resorting to his methods.* Just my recomendation...

* By that I mean to criticise only the content, but not the person.

monomer

14th April 2004, 17:50

ux-3, you are right in what you say but you've got to admit peachee's elitist attitude is hard to ignore and besides I did enjoy the humor in geffroman's 'Re: The Passion of Peachee' response to peachee's perceived self-styled martyrdom.

Anyway, let me ask this... Can one actually see (visually) every single 'bit' of video information provided on a DVD? ...for instance, every single shade of black? ...if one pixel is one shade towards blue could you pick it out of a 'sea' of black pixels? If more video information exists than the eye can see, couldn't you effectively remove it and still have exactly the same picture quality? I believe that's what should be the concern when discussing the abilities of a transcoder. I will reiterate... transcoders can select and remove video information that cannot be detected by the human eye when compressions are on the orders of 10% or less (in reduction). Of course to be entirely accurate, I must add that it would also depend upon the original bitrate use and how much action and shadows are in the particular movie to begin with. As a general rule the higher the bitrate is the more information can be removed without affecting the visual quality (ie. what the human eye can actually detect). Also the more shadows (dark scenes) a movie contains the more opportunities there are to remove minute differences in shades of black and again be undetectable to the human eye. However even with a bright, fast-action, low bitrate movie I would venture to say one could reduce the size by 10% using a transcoder and still not be able to detect any visual degradation at all (even using a computer monitor at HDTV resolutions).

Just for fun I think it would be interesting if Zhnujm would post another series of comparsion stills but refrain for telling us which is which, other than identifying the original... and let's all of us then take a guess which are the transcodes and which is the CCE... wouldn't that be fun? ...plus it just might be educational for some of us.

(Zhnujm... maybe say 6 stills... 2 originals (only identify one), 2 transcodes, and 2 CCEs. Using multiples (exact copies) of each should eliminate the pure guess factor (you know the 50/50 chance thing?).

What da ya say?

jorel

14th April 2004, 18:08

"wouldn't that be fun?"
great idea monomer.
;)

but excuse me, i'm trying to be sincere, not to fight:
it's not personal....
can i trust in the pictures that will be posted if he wrote:

"The CCE encoding just uses standard settings, no filtering, 2 pass vbr.
--->And is of course the same size as the transcoded files."

sorry ....for this reason i ask for 1000 frames from the source.
i never get the same size of any source encoding with CCE!!!
i need to use the source....
in another thread about audio when i was the only one with right results,
in the end i live alone the thread cos was posted that i "had luck"
after all the tests that i did.

i will that my sincere opinion don't give problems to the thread or for me.
if any moderator is against my opinion,please,delete my post.
thanks!
:)

monomer

14th April 2004, 21:18

Jorel... I'm not looking to definitively prove anything, there are simply too many variables involved. One of them being motion artifacts... and I'm not suggesting we download a movie either (I only have a 56K modem and it took me over 12 minutes! to download Zhnujm's 4 stills yesterday). Also since CCE reallocates bits and transcoders selectively remove information, if you look hard enough 'frame-by-frame' its conceivably possible to find a particular set of 'stills' that will 'prove' whatever point of view you wish. But having someone claim to be able to distinguish, with 100% accuracy, a transcode from a CCE encode or the original has me intrigued because I sure know even on my set-up I couldn't do it with lightly compressed (<10%) movie. I'd just like to see if anyone else can do it... besides our resident quality expert, 'peachee'.

Besides, it may take the discussion a little further than just pure speculation and opinion which, though interesting, doesn't seem to really go anywhere ...and besides it may also add a little to this thread's entertainment value.

ux-3

14th April 2004, 22:53

Well, Monomer, so you do propose a double blind test. That sounds like a lot of fun. Remembering my Hifi days, you could fool most people into believing anythig, by just making one source just a little louder. You could redicule the audio purists - but I still rather have CD than Tape, so the point is limited.

Kedirekin

14th April 2004, 23:04

Time for a thought experiment...

Here is a scenario: a movie only transcode small enough that it can be done with no compression.

Hopefully we could all agree that, in that scenario, a re-encode with CCE cannot have better quality than the transcode (if you'll allow me to call it a transcode). In fact, I'd hazard that the output from the transcoder would be superior; re-encoding will introduce generational degradation no matter how good it is. [Perhaps some people won't agree ... oh well, I can't help if you insist on being difficult].

Now that we've agreed on that :sly: , how about compressing by just 0.01%? Will the output from the transcoder be better, worse, or the same as re-encoding with CCE? I suspect fewer people will agree on this point, but at 0.01% I think the transcoder will win. At 0.01% the degradation of re-quantizing will be less than the generational degradation of re-encoding.

And, IMO that is all the further you need to go. If you agree that transcoding at 0.01% compression produces better results than re-encoding, then you perforce agree there is some point where transcoding and re-encoding trade places from a quality stand point. The exact value of that point is very much in the eye of the beholder.

For example, I am very intolerant of mosquito noise, so I would set the point quite low (perhaps 85%). Someone who is very intolerant of macroblock noise will set the point much higher (maybe 98%). But in any case, there is a point where transcoding will be better than re-encoding.

Bottom line: a professional realizes when a tool is useful and uses the correct tool for the circumstances. I don't know what you call a person who insists that an obviously useful tool is never useful and should never be used; amateur surely doesn't fit.

DDogg

15th April 2004, 00:35

I started to post this before, but it seemed a mine field. I'll try it again to get Kedirekin's views. Consider it another thought experiment ;-)

There seems to be many generalizations made about what percentage of reduction is acceptable in a transcoder. I'm fairly certain that this is dictated for a transcoder by the complexity of the source just like it is for an encoder. I don't use transcoders much so I had not really thought about it much, but I thought it might be that one source might only allow 15% reduction without serious degradation and another might allow 50%.

This is not real scientific as I didn't record my numbers, but I roughly tested this by demuxing a movie stream of a DVD I had handy and derived the average bitrate from the filesize of the original movie. IIRC, it was about 4100. I then ran a OPV 1% sample of the same m2v at a Q32 and derived the ABR from that. The sample derived-ABR showed an ABR of around 2000 was needed to hold the average quantization of the Q32 sample. This seemed to show about 50 % 'headroom' between the Q32 D-ABR and the ABR of the original.

That seemed to indicate that this particular source might allow something like 50% and still be of appreciable quality to an CCE OPV Q32 encode which is pretty good quality.

The I used CloneDVD2 to transcode the movie at 50% and then demuxed the movie stream. Bitrate viewer showed a similar average quant between both the sample and the transcode which seems to indicate source complexity is the primary consideration, and one cannot generalize about transcoder reduction levels. They may be very specific to the source, and every source will allow only a certain level of reduction before "breaking".

To get serious about it I/we/whoever would have to do the tests properly and graph the data. Maybe I will get around to that one day if it seems to have merit.

Kedirekin, what's your general reaction to this? Am I smoking crack again?

If anybody wants to play around with this you can find a derived bitrate spreadsheet here (http://forum.doom9.org/showthread.php?s=&threadid=74141) that I was using for another project. It makes it a little less of a grind.

wmansir

15th April 2004, 00:35

@Kedirekin

Your thought experiment remindes me of this anecdote, which I have not been able to verify as true:Winston Churchill was at a large house party and was chatting up some Duchess or another and was already quite, shall we say toasty.

He looked at her and asked, "Madam, would you sleep with me for, say, a million pounds."

She paused and responded, "Well, Sir Winston, I believe I would."

"How about for ten pounds?"

The aristocratic lady was horrified.

"Sir Winston!" she gasped. "What kind of woman do you think I am?"

To which, Churchill smiled and replied, "Madam, we have already settled that question. Now all we're doing is haggling over price."

jorel

15th April 2004, 00:38

@ monomer
you wrote:
"I'm not looking to definitively prove anything, ..."
no? why you propose:
"Just for fun I think it would be interesting if Zhnujm would post another series of comparsion stills but refrain for telling us which is which, other than identifying the original... and let's all of us then take a guess which are the transcodes and which is the CCE... wouldn't that be fun? ...plus it just might be educational for some of us."
:rolleyes:
my english is horrible but i understood exactly what you posted.
and:
"and I'm not suggesting we download a movie either"
not you but i did this suggestion!
..i'm not against you but i feel something....strange! :eek:
why you're angry!?!?!
seems faster to fight but slow to think that i'm not against you!...sorry!
i only posted your own words here ! ;)

@ Kedirekin
my sincere respect for your knowledge but i have some differents opinions...
you wrote:
"Now that we've agreed on that..."
this not means that you're right and not means that i'm right!
and:
"I am very intolerant of mosquito noise.."
just like me...and for noises in audio i'm intolerant too.
i have some old vinils that i load in wave editor
t-racks,cooledit... and using some filters,
i got better result than listen the vinyls
with scratchs,rumbles,noises,etc.

i use CCE with some filters for any case with .avs script
that give me better result than the source used.
(not the same script with the same parameters for "everything" )
then,
here you are absolutely correct and i think the same:
"Bottom line: a professional realizes when a tool is useful and uses the correct tool for the circumstances."
i encode only musicals for me and cartoon for my kids.
musicals need very good treatment cos most of them have
fast lights ,smokes , tons of colors, fast moviments and clean sound.
musicals i see "all the time", is different for me if i encode matrix as sample.
i can't see matrix 20 times but i can see led zep 2000 times!
i don't encode movies!

for this reason i ask for 1000 frames from the source!
and i don't post anywhere that dvdshrink is bad,
i only can get better results using encoders
with the right filters in avisynth script!
....and with short final size too!
:)
this is not my opinion only, is my work too!
eletronic,image and audio quality is work from 31 years,
audio is my hobby at 43 years and image at 30 years.
i know what is quality and this is my target!
thanks!

Kedirekin

15th April 2004, 02:03

@DDogg,
I completely agree - the success you'll see in a transcode is strongly dependent on the source. If the source is already heavily quantized, quantizing it further is going to have unpleasant results. If the source is not heavily (barely?) quantized, you can trancode it with fairly heavy compression and probably get pleasing results.

Beyond that, I think I need time to digest what you've said and the contents of the other post. At a glance, what you're trying to do (if I understand what you're trying to do) seems valid, but I don't know if it's universally applicable. I believe there are dangers in trying to boil an MPG down to a few numbers (and I think that is what you're trying to do to drive a GO/NO-GO decision).

Anyway, let me think on it.

Kedirekin

15th April 2004, 02:04

@wmansir,
LOL. That's a great quote (if indeed it ever happened). It's hard to say which historical figure has the best anecdotes, but Churchill and Twain are certainly in the running.

Kedirekin

15th April 2004, 02:07

@jorel

you wrote:
"Now that we've agreed on that..."
this not means that you're right and not means that i'm right!

Yep, but it was meant to be a little tongue-in-cheek. That's why I put the sly-smiley after it. But seriously, if we can't even agree that a no-compression transcode (and I use the term "transcode" loosely) is better than a CCE re-encode, than we have no common basis for discussion. We'd basically have to agree that we fundamentally disagree.

I notice that you say you use filters to make your re-encodes better than the original. I would say we're using different definitions of "better" here. Your definition of better is a result that is more visually pleasing to you. My definition of better is a result that is most like the original source, including things that you might consider flaws.

I do hope you can agree that what you find more visually appealing is a personal thing, and that others might dislike changes that you like. Softening versus sharpening is an obvious example of this - some people love to soften a video to "improve" the quality, while others hate it, and visa-versa for sharpening.

I can understand your comments about musicals. I assume you're talking about concert footage. That's a situation where the source has already maxed out the quantization on the original DVD. As DDogg alluded to, attempting to transcode a DVD that is already heavily quantized will probably have unpleasant results. In that situation, I'd almost always split the disk to two DVDRs. If you get results that please you by re-encoding, then more power to you.

BTW: I used the term mosquito noise here to refer to a type of video noise, not audio noise. It's also sometimes called ringing. It refers to fine grained bits of lightness and darkness that seems to float around sudden transitions in contrast - it quite literally looks like mosquitoes are flying around the edges of objects (for example, a person with curly dark hair with a daylight sky in the background).

jorel

15th April 2004, 03:44

ok!
;)

little comment about mosquitos noise:
i know,knew and understood what you wrote.
you posted that was intolerant of mosquito/ringing noises...
and i wrote:
just like me...
--->and
for noises in audio i'm intolerant too.
means that i'm intolerant for this 2 problems....
you understood different? (my poor english of course)

i post about audio only to do comparisons
that using filters we can get better results for
---> each situation....depend of the problem.
when somebody write "noises" in audio,
we can understand easily but
if this "noises" is wroten about video,
we need big details for that definitions, right?
i can't find better words....
i'm writing the same that was wrote!

"basically have to agree that we fundamentally are"
---> searching for good results!

thanks for wise and friendly answers !
best regards.
:)

DDogg

15th April 2004, 03:44

@Kedirekin,
Beyond that, I think I need time to digest what you've said and the contents of the other post. Always the gentleman :-) Ah, er, I got on a bit of a roll talking to myself :-o
... dangers in trying to boil an MPG down to a few numbers Sure, I hear what you are saying and I realize it will take time to make much sense from the ramblings, although I think the condensation should hold up to critical analysis (X'ing fingers) ...

It's probably more like: custom matching source complexity to a bitrate of sufficient robustness to 'hold' it securely and keep it from falling out the sides. Metaphorically, maybe like putting the right size pair of pants on a fatman, 'cause it is sure ugly when you don't. You got to measure his waist before buying the pants. :-)

@All - I think it would be fairly easy to design a methodology so that each of you could compare transcoder and encoder final results in a non subjective way, but which method is best may flip back and forth on you depending what the source complexity is like.

Example(from memory: About Schmidt is 2 hours and takes about 1950 Kbps to hold a Q32 (just saying that in shorthand). Good Fences, also 2 hours takes 3850 Kbps to hold the same level of Q.

A transcoder could probably do 45-50% on the first and deliver excellent quality. On the second example, I'm sure 15% would be evident cause its already banging the ceiling bitrate wise. Nearly any reduction of bitrate (which is what you are doing) would likely show.

However, for you folks to do a comparison between yourselves is much more difficult. You would all have to have the same exact resolution of the same DVD title and create final encode sets that were the exact same size. Difficult logistics.

All this is IMO and no I didn't really think it all out. I rarely do :-p

dragongodz

15th April 2004, 06:41

Also the more shadows (dark scenes) a movie contains the more opportunities there are to remove minute differences in shades of black and again be undetectable to the human eye. However even with a bright, fast-action, low bitrate movie
just a couple of corrections monomer if you dont mind. dark scenes usually have low bitrate because not as much detail is needed to be kept or is in fact there. thats in the original encode, unless the encoder uses a low quant to try and get to the target bitrate. so in general terms dark scenes will show less true but will also have (again in general) less that can be done to it.
bright and high action scenes however uaually have a lower quant(except when the encoder is trying to limit the bitrate for final size etc etc) so can have more done to them. they will show it much more aswell though so a bit of catch22.
there are exceptions to every rule of course. :)

I'm fairly certain that this is dictated for a transcoder by the complexity of the source just like it is for an encoder.
100% correct. in 1 of the Rejig threads this was breifly talked about aswell.

now somewhere it was also mentioned that screenshots can be made to prove any point just by picking the right ones. yep thats true which is why i said before that i dont know anyone that watches a film at 1fps. they can be useful to point out an effect, such as saying "this is what i see while playing" but to compare a frame and say that it alone proves anything is wrong. look at Doom9's codec tests, he uses some frames to show effects he sees but if you read the comments he bases his opinion on the actual playing footage. need i say more ?
:D

monomer

15th April 2004, 19:01

Originally posted by jorel
@ monomer
you wrote:
"I'm not looking to definitively prove anything, ..."
no? why you propose:
"Just for fun I think it would be interesting if Zhnujm would post another series of comparsion stills but refrain for telling us which is which, other than identifying the original... and let's all of us then take a guess which are the transcodes and which is the CCE... wouldn't that be fun? ...plus it just might be educational for some of us."
:rolleyes:
my english is horrible but i understood exactly what you posted.
and:
"and I'm not suggesting we download a movie either"
not you but i did this suggestion!
..i'm not against you but i feel something....strange! :eek:
why you're angry!?!?!
seems faster to fight but slow to think that i'm not against you!...sorry!
i only posted your own words here ! ;)

Jorel... I not angry at you at all. If you got that impression from my posting, believe me I am truely sorry and I do apologize. I was just suggesting we all have a little fun if Zhnujm was willing, that's all. I'd love to do the movie thing ("1000 frames") but that would be an impossibility for me... that's all I was trying to say there. Sometimes in English it can be hard to interpert how things written by someone are meant to be taken by the reader... and since I never use 'smileys' I can understand your confusion. By the way your English is infinitely better than my Portuguese.

And did you know (if your name is actually Jorel) that's the name of Superman's real father on planet Krytonite, except its spelled Jor-el?

jorel

15th April 2004, 22:11

all right monomer,
sorry if i misunderstand you! :o
the "guilty" is my bad english and emoticons help me a lot!

you wrote:
"Sometimes in English it can be hard to interpert how things written by someone.."
oh yes and really hard is think and write in another language (my case)

and if my english is better than your portuguese,
means that you don't know portuguese?
:p (kiddin)...my english is a "trash" !

about Jor-el
yeah..i knew and my username is intentional.
i "mix" the 3 first letters of my name
with the 2 first letter of my wife's name!

thanks for your good education and for don't use any Krytonite to answer me!
:p :)

DDogg wrote about Kedirekin:
" Always the gentleman :-) "
true....very nice person !

@ all
:o
sorry....it's off topic but
i needed to post a "public" answers
and thanks for monomer and not using pms!

ookzDVD

16th April 2004, 06:55

@Forum,

How about the DVD-Rebuild output ?
anyone here compare it with DVDShrink output ?

Thank you.

wmansir

16th April 2004, 09:06

Originally posted by ookzDVD
@Forum,

How about the DVD-Rebuild output ?
anyone here compare it with DVDShrink output ?

Thank you.

DVD-RB is more of a front end for different encoders (as far as quality of encoding goes), CCE, QuEnc or Rejig.

CCE is what this thread is about.

QuEnc is an open source encoder currently in development by Nic here on the boards. Check the Other MPEG1/2 Encoders subforum.

Rejig (also by Nic) is a Transcoder, like DVDShrink. Check the thread in this subforum for info on how it stacks up against other transcoders.

jorel

16th April 2004, 18:53

DDogg wrote:
"Am I smoking crack again?"
:eek: (lol)

but...why did you stop? :confused:

you will see better colors with this " plugin " !
:p
(kiddin my friend, don't shoot) ;)

Zhnujm

16th April 2004, 22:29

Hi !
As my original statement was that cce looks also better with a low reduction and not the opposite (and thats only my personal opinion) i want to post some more pictures to show why i think so.

Unfortunatly i cannot use the exact same frame as in the first test as DVDShrink did not compress this frame, so i post the frame before and after (left original,middle cce,right dvdshrink):
ftp://zhnujm.dyndns.org/Image12s.png
ftp://zhnujm.dyndns.org/Image13s.png
DVDShrink is more blocky IMHO.

Also i came across some really ugly parts like these:
ftp://zhnujm.dyndns.org/Image10s.png
ftp://zhnujm.dyndns.org/Image11s.png
Here you can clearly see that DVDShrink enlarges the macroblocks, and cce seem to smooth them. I like the smoothing more. (i wonder how much people would vote for the cce picture as original)

I think the quality of DVDShrink could be better if the reduction would be more constant and in every frame. At higher ratios there is still this ugly pulsation effect (for example in the other movie i posted, Conan the Barbarian). Instantcopy did not have this effect.

So my own result is that there are parts in that DVDShrink looks better (but hardly noticable for me), but there are also parts where cce looks better (and these are easily visible for me).
I think i leave out the second audio track to do a 4.7GB to 4.4GB reduction and see if the results change.

Originally posted by jorel
no matter what adjust, source, filters, resizes....etc,
used in CCE, i never got the same size of the source or transcoded file!
removing doubts: how you did it ?

I think its a matter of how you define "same size" :D
The DVDShrink result is 4.35GB, the CCE result 4.3GB.
Most times the cce encoding is even more smaller for me, if you want the perfect size you can still redo it with a few more bits. But i like the idea of saying "and the cce encoding is even a bit smaller" afterwards :D

Heres also the rearranged picture:
ftp://zhnujm.dyndns.org/Image14s.png

I can easily post you a 1000 frame part, but that would be around 20MB each. Theres absolutly no problem for me to upload it but you may have problems downloading it - the server is not very fast as you may have already noticed :rolleyes:
Just tell me wich part of the movie and you can try.

ookzDVD

17th April 2004, 03:11

Originally posted by wmansir
DVD-RB is more of a front end for different encoders (as far as quality of encoding goes), CCE, QuEnc or Rejig.

CCE is what this thread is about.

QuEnc is an open source encoder currently in development by Nic here on the boards. Check the Other MPEG1/2 Encoders subforum.

Rejig (also by Nic) is a Transcoder, like DVDShrink. Check the thread in this subforum for info on how it stacks up against other transcoders.

Sorry then, the CCE result from DVD-RB I meant, with the default setting (using CCE 2.67) how about the result ?

cable2

17th April 2004, 05:52

Article Link (http://www.emedialive.com/Articles/ReadArticle.aspx?ArticleID=8421)

Technology may make these kind of debates moot, starting by the end of this year. I guess it will depend on when the burners and media reach the right price point. The link above should get you to the full article, you might have to cut and paste it though. I though it was fairly well written, I hope some of you may find it so too.

dragongodz

17th April 2004, 06:10

Instantcopy did not have this effect.
while all the transcoders work in a similar way they do not do everything exactly the same. just as CCE does not do the same as Tmpgenc or MainConcept etc etc etc.

Technology may make these kind of debates moot, starting by the end of this year. I guess it will depend on when the burners and media reach the right price point.
i doubt it. not everyone will have the money or desire to go and buy a new dvd burner just for that. many are quite happy with movie only backups they can use and abuse even if they are not 100% the quality of the original(which cuts for both encoders and transcoders).
also when dual layer dvdr's have dropped to similar prices to current single layer dvdr then you can reasonably expect single layer dvdr's to have also dropped, probably closer to cdr prices. :)
also you can see from the divx and xvid sections of this very forum that people still make movie to cd conversion quite happily so why not dvd to single layer ? :D

jorel

18th April 2004, 00:22

@ Zhnujm
you wrote:
" I think its a matter of how you define "same size" "
ok i understand...but you was wrote in the first post from page 9 that:
"The CCE encoding just uses standard settings, no filtering, 2 pass vbr.
---> And is of course the same size as the transcoded files."
quoted your own words....it's your definition !
;)

and more:
"i can easily post you a 1000 frame part, but that would be around 20MB each. Theres absolutly no problem for me to upload it but you may have problems downloading it - the server is not very fast as you may have already noticed
Just tell me wich part of the movie and you can try."

right...i don't know that movie.
please, choose the scenes and
Mb size (more or less than 20Mb, no problems) and post the link!

thanks in advance Zhnujm !
:)

jorel

18th April 2004, 01:12

i have some reasons to use Tmpgenc and CCE more than dvdshrink!
i encode Kvcds mpeg1 using Tmpgenc.
CCE for mpeg2 and to reauthor vobs.
dvdhrink for cartoons, no more than 20% !

i was reading that another thread here in the forum:
http://forum.doom9.org/showthread.php?threadid=74561
and in the first post found that link with some comparisons:
please,take a look....

http://nerds.palmdrive.net/dvds_tmpgenc2.html

i'm not trying to show what encoder is best for this or that situations,
this is not the idea and
isn't important in this case about CCE/DVDshrink thread,
and here we don't ask or answer "what is best"...

i only agree with the results.
using encoders i -->ever<-- get better results using the right scripts/filters!
...and short final size of course!

please,post your comments! thanks.
;)

ps to all:
we all are posting here trying to show good results and
what arguments we're using for the target needed.
that is my objetive and (i think) yours too!
i'm not trying to show what's best or worse,
i learn a little more reading all posts with all opinions!
:)
(hey,...this "ps" is too big)

Zhnujm

18th April 2004, 14:12

@jorel:
You can try to download it here:
Original - ftp://zhnujm.dyndns.org/original/ 29.8MB
DVDShrink - ftp://zhnujm.dyndns.org/shrink/ 25.1MB

I splitted it into 2MB parts, hope that helps.

btw, i extracted also the CCE part, it is even bigger than the original - 32.9MB

jorel

19th April 2004, 13:12

Originally posted by Zhnujm
@jorel:
You can try to download it here:
Original - ftp://zhnujm.dyndns.org/original/ 29.8MB
DVDShrink - ftp://zhnujm.dyndns.org/shrink/ 25.1MB

I splitted it into 2MB parts, hope that helps.

btw, i extracted also the CCE part, it is even bigger than the original - 32.9MB

:)
OK Zhnujm !

all files are here and was not so slow to download. (round 7.5Kbps)
(well, sunday is slow to download everything here)

thank you very much!

Joergen

20th April 2004, 01:40

I just did a disc that is the BEST POSSIBLE SCENARIO for a transcoder like DVDShrink. The movie is 4.66GB and I just load it up in shrink, it shows 95% for all titles. Select no compression for the menu part (50MB) and fire away. Perfect copy in notime with quality better than an encoder could (at least in theory) do. It's especially times like that when I thank DVDShrink for his great software.

Then I did another, where the overall bitrate needed to be 68% (3900kbit) of the original. And I thank Jdobbs for his DVD-RB that did this one with stellar quality keeping menus and extras.

Best of both worlds PEOPLE!

jorel

20th April 2004, 02:58

@ Joergen
i agree with your point of view too!
;)
the best is "what you need and use" for each situation (target/size)!

@ Zhnujm and all
i take 10 frames for each sample posted in vdub.
was choosed from 651 to 660 frames...scenes with the river!
differences that i saw between the original from the dvdshrink was:

in the frames: 652, 653, 655, 656, 658, 659....the original is really better.
in the frames: 651, 654, 657, 660.....is seamless,can't see differences.

seems that dvdshrink works in 2 from 3 frames,
2 frames get compressions, the next frame don't have compression....
:o ...i was clear?(poor english)
don't know if work in the same way using more compression !
...comments are welcome (as always),please.

for me, scenes with fire, water(in moviment) and smoke
have the form and dimentional aspect change all the time and
this type of scenes are the worse to encode or transcode.

now i'm using the original source to encode withh CCE to compare!
:)

ddlooping

20th April 2004, 04:27

Originally posted by jorel

... seems that dvdshrink works in 2 from 3 frames,
2 frames get compressions, the next frame don't have compression....
:o ...i was clear?(poor english)
don't know if work in the same way using more compression !
...comments are welcome (as always),please.

jorel, DVD Shrink works with B-frames when little compression is required, with B an P frames when more compression is required, and with I, P and B frames when a fairly high level of compression is required.
I'm not sure about the accuracy of my frame naming though, you might have to switch B and P. ;)

jorel

20th April 2004, 04:44

very cool and clear ddlooping (your explanations and how dvdshrink works)

thanks!
:)

dragongodz

20th April 2004, 08:35

ddlooping - you got it right. B frames first. :)

jorel - as for other transcoders well that can vary. example is Rejig starts off just doing B frames but adjusts as it goes to try and make sure it hits the target size/percentage. that is if the currently done frames output size is falling behind the target then some P frames will start to be done etc.
as for the test you did, well dont forget what i said about frame comparisons. that is being able to see a difference in a still frame is not the same as seeing anything in real playback. so, just as Doom9 does with his codec tests, single frames should be used as an example of what a person sees during playback and not the sum total to base any opinion on.

i noticed aswell that in MediaPlayerClassic i turned output to VMR7(renderless) and mpeg playback looked crap. turned it back to VRM(windowed) and mpeg quality much better(you have to quit and restart MPC after each change to make it active). this is a gfx card/driver issue i have no doubt. just something for people to also consider when playing files from their pc. that is a pc is not a tv so dont asume what you see on your pc is what you will see on your tv. :eek:

jorel

20th April 2004, 10:28

you're right!

i did that frame by frame comparisons only trying to see
"if and/or where" the image change from the original source
and it was used as example and not to as "base to opinion"!
when i post some sequences of samples, i always recommend :
"burn in rw and play in standalone - tv, not in pc monitor or video out".

the pc monitor is our reference with lots of details,
but the tv show (or better,don't show) the real results,
loosing details with his limitations.
the result is very different... i posted the same somewhere here!

all your observations posted here are importants dragongodz,
and with louvor cos "everybody" forget the differences between tvs and monitors,
thanks for that!
:)

Zhnujm

20th April 2004, 21:43

I wonder if this compression of only a part of frames is a feature or a limitation.
I think it should be better to reduce the size of every frame.
Or is it not possible for a transcoder to do a very very small amount of reduction ?

begu

19th August 2004, 21:34

little offtopic, but I decided to add:

well, I did indy 1 (lost arc PAL R2) with 59 % of the original size (movie). I did it with CCE basic 2pass defaults using DVD-RB and shrink versions: 3.17 and 3.2. I did the all AEC settings in 3.2. Also the IC8 is finishing right now.

But to add some perspective I did Xvid too with SAME bitrate (eg. same size for xvid too) (note: I did have to do it in 2 files, because the 2 gb file limit for .AVI). I used: no bframes, mpeg quant, adaptive quant, 2pass and maximum VHQ, chromamotion, maximum motion search, no trellis.

And guess what. The CCE and shrink 3.2 are very close. The shrink had pumping effect in quality (maybe due to PBB frame types and different reduction?) but it was not so bad. The CCE was pretty good with no so much pumping. BUT: the xvid did have the most best CONSTANT quality above those all others (didn't comapare with ic8 yet). Maybe the image is slighlty blurred and the dynamics are not so deep than in transcoders, but I really liked the quality quite much.

I have x-card (sigma's hardware decoder for mpeg2 and mpeg4) and high quality sony TV. The hardware decoder has not any postprocessing and has very sharp picture quality. Using good contrast and the TV own sharpness at maximum, all compression artifacts are clearly visible. And now, the xvid still looks the best. There really is compression artefacts in both shrink and CCE results. I only see a few in xvid. And this is NOT a troll, really. Yes, the othres than xvid is a bit sharper, but with lots of artefacts too. And the shrink pupms in the quality (as does the CCE little) compared to constant fluid quality of xvid.

Well, try it too, just make the same bitrate for the xvid, sound overkill, but it is not. (xvid used quantizer 2 almost all the time for low motion scenes, bitrate varying between 4400 and 5200 kbps). I'm just looking the best quality for my backups (I have a projector too, sanyo Z1, and the artefacts are really noticeable at big screen)

Now I have to test the IC8 quality too. I hope the best. :) As far as I remember I did the hannibal using 60 % with old shrink 317 and ic8. The ic8 was way better than shrink in the big screen. Feel free to comment. Sorry I have no examples about the quality, You may just read my personal findings. But, I was surprised about the xvid, really :)

ddlooping

19th August 2004, 21:39

Hi begu. :)

Very interesting findings. A few questions:

- Did you use DVD Shrink Deep Analysis?
- Did you use its default "Sharp" option.
- Did you also try its "Smooth" setting?

TheSeeker

19th August 2004, 22:21

Im not really very surprised. Thats why divx and xvid have gotten so popular is that the compression algorithms are bordering on modern day miracles. The only reason I have really started to move away from divx compression is that i wanted to be able to watch the movies on my tv in the living room without hauling my computer out from my room. I know I could get a dvd player with divx compatibility but I decided to go this route instead? Was it the correct choice? I think it was considering that now I can bring my movies over to a friends house who doesnt have a divx compatible dvd player and watch all the movies I like. And if I have to cut out extras and unneeded language soundtracks to get decent quality dvd's oh well I say. But your right. Of all the compression algorithms divx/xvid (or mpge 4) can definately provide the best quality video at the bitrates we are talking about.

jorel

20th August 2004, 00:15

so good..great topic is back!
DVDShrink is better in this new version but only can compare with CCE if no more than 30% of compression is used. for more than this,CCE stills champion....i'm using one dvd with 7.41GB from source and with DVDShrink give me ~53% of compression,the image is pulsing,sometimes the audio is out of sincro and in scenes without moviments the image and subs are pulsing like posted here: http://forum.digital-digest.com/showthread.php?postid=188621#post188621

ddlooping

20th August 2004, 01:35

So jeo = jorel!? :)

If "pulsing" is so extreme, I'd suggest you use "Smooth" or "Maximum Smoothness". ;)

jorel

20th August 2004, 04:04

ddlooping,
the "pulsing" was not so extreme, i re-did using "default" sharp and is really better.....the source is one documentary(witness of history)from 1900-2000 and lots of scenes in B&W, only the "last years" have better images with colors. was not needed to use "Smooth" or "Maximum Smoothness". thank you (here and there)! :)
oh yeah...jeo = jorel.

ddlooping

20th August 2004, 04:08

jorel, would you mind telling us how the "Sharp" version now compares to the CCE one? :)

jorel

20th August 2004, 04:22

Originally posted by ddlooping
jorel, would you mind telling us how the "Sharp" version now compares to the CCE one? :)

of course:...i did using main movie without subs, only one AC3-5.1 from source for DVD-5(4.7Gb) target:
with DVDShrink3.1 or older versions,for my taste no more than 80% but with the new 3.2 version i got the same with 70% comparing with CCE.
using cartoons sources, if compressions settings stay in 70% is hard to find the differences with default sharp or maximum sharpness but more than 70%,or better,if the source have more than 6GB,CCE (dvd2dvd with roba) using undot,deen and asharp(1,2) in .avs script is really, really better!

ddlooping

20th August 2004, 04:30

Thanks for the reply, jorel. :)

It'd be really interesting if you could do a comparison between CCE and DVD Shrink using the "Smooth" setting.

Here is a comparison test I made with LOTR-ROTK, 60.9%:
http://www.dvdshrink.info/temp/beta/comparative_tests_smooth.html

I'd appreciate it if you could let me know what you think. ;)
The following compilation is the most revealing (1.5MB):
http://www.dvdshrink.info/temp/beta/32b3_CCEvbr2_3.zip

jorel

20th August 2004, 06:31

ddlooping,
the tests with LOTR-ROTK at 60.9% are cool but very small for comparisons. i download the 32b3_CCEvbr2_3.zip and found 2 little vobs. from the first vob(vts_01_1.vob), i get the screenshot from frame 45: http://img44.exs.cx/my.php?loc=img44&image=f45-vts01.jpg
and from the second vob(vts_02_1.vob), i get the screenshot from frame 56: http://img44.exs.cx/my.php?loc=img44&image=f56-vts02.jpg
see that is the same scene and if i understood,the first vob was did using DVDShrink and the second vob was did using CCE. for me the second screenshot( from CCE? ) have more sharpness and details(like in his hair on the right).......---> correct me if i'm wrong about the first and second vobs...and they have differents sizes, the first vob have 762Kb and the second have 906Kb
a few months i did toy story 1&2 without subs and only one ac3-5.1 from source using dvd2dvd and it can be used for comparisons test. i'm thinking to use this 2 cartoons and do the same using DVDShrink3.2 using "smooth" (or other quality settings) to compare with that encoded dvd...same sources for same target(sizes)...what do you think about that? :)

ddlooping

20th August 2004, 06:45

jorel, I'm sorry but you totally missed the point. :(

You're comparing frame-captures which have probably been compressed at different ratios.

Please compare the clips by playing the compilation in your software player, at normal or slow-motion speed (Full-Frame would also emphasize the quality issues).

One of the two clips has a very distinct "crawling noise" aberration, most noticeable on the sides of Gandalf's nose.
It can only be seen while playing the clip, as when paused it does look like the frame is sharper when it's actually "paused-crawling-noise". :)

Please do the Toy Story test, that'd be very interesting. ;)

jorel

20th August 2004, 07:01

ok....the frames are differentes but they are the same pictures,the first vob have 57 frames(00:02.280) and the second vob have 71 frames(00:02.840),than,--->the same scene is in different frame of each vob...
but i got the idea: comparing in powerdvd the first sequence(vob) have less "mosquitos" and seems (more) clean in normal or slow motion! ;) is very good and easy to see.

ddlooping

20th August 2004, 07:06

Thanks for the confirmation, I appreciate it (*sigh of relief*). :)

You can also observe the original does not have this crawling noise.

http://www.dvdshrink.info/temp/beta/original_CCEvbr2_3.zip

http://www.dvdshrink.info/temp/beta/original_32smooth_3.zip

begu

20th August 2004, 07:51

Originally posted by ddlooping
Hi begu. :)

Very interesting findings. A few questions:

- Did you use DVD Shrink Deep Analysis?
- Did you use its default "Sharp" option.
- Did you also try its "Smooth" setting?

yes, in all the clips (it did remember the analysis after the firs one)
I did the all settings that the shrink offers. Even did the 3.17 (but I leave it out due to lower quality compared to 3.2.

The xvid did not have such deep and rich colors on my monitor than the other (mpeg2, transcoded or encoded [shrink, ic8 or CCE basic]), but really it had the most less compression artefacts, and the sharpness was almost the same. If not counting the slight lack of color and contrast richness (compared to mpeg2 variants) the picture, or should I say motion picture was the most 'natural' looking.

I'd say, maybe the mpeg2 wins at the original bitrates (6000-9000) but when in reduced bitrate the xvid (using mpeg quantizer, no bframes, look earlier post for details) produces the most less compression artefacts and the most natural looking motion picture. Only problem is the reduced dynamics in the compression process. Also the CCE do have some reduced dynamics. The transcoders ofcourse follow the dynamics of the original. I will have to do tests on my TV with xcard (not my comp. monitor) It will clearly show any artefacts better than monitor (because of the HW decoder with no postprocessing and the maxed sharpness of my TV, sony trinitron). Then I will look, if the xvid looks better (I used most recent ffdshow for xvid decoding) than in the monitor.

I understand that if You don't have HW decoder player for xvid, the transcoded/re-encoded DVD will always look better on TV. Maybe some matrox/ati high end cards could deliver similar quality than a high end dvd player on TV screen, but I doub it. I will test the projector picture too.

edit:
I did the encoding with virtual dub and decodefix 110. No filters, just fast recompress in virtual dub and the cropping in avisynth, no resizing. Also did the full frame, eg. just cropped out the black borders in up and down part of the frame, no crop at the sides.

edit2:
because there are so many tools out there, nowdays I'll have to do all of them and then just decide, wich looks the best. But I know it is very hard to judge in some cases. :)

ddlooping

20th August 2004, 14:19

begu, the reason I asked if you'd tried the "Smooth" setting is you said "Shrink had pumping effect in quality".
If you tried all the AEC options, which one did yeld the best results, and did it have this pumping effect?
Was there one that did haven't this effect, and if yes, how did it compare to CCE?

Thanks. ;)

ddlooping

20th August 2004, 18:36

A big apology to all who have downloaded the "Smooth" comparison tests.
The links were actually pointing to the v3.2 beta 3 test files. :o

This error has now been rectified. :)
A "Maximum Sharpness" comparison page has also been added. ;)

http://www.dvdshrink.info/temp/beta/comparative_tests_smooth.html
http://www.dvdshrink.info/temp/beta/comparative_tests_maxsharp.html

MackemX

21st August 2004, 12:16

these comparisions are better than stills but are these still fair comparisions?

some have different sizes and different frames and who's to say that if you sampled the next 3 seconds that the results aren't going to be the same?

I've done a few tests and I know I could take selective parts of a DVD and show mixed results. If you could slow time down and still watch the film but have time to analyse each frame (basically process information much faster than humans can now). You could then do a frame by frame comparision, marking which is the best in your eyes and going through the whole DVD. Basically you are just comparing each and every still but still viewing it in 'Bullet Time' :)

then you could say something like DVDShrink is closer to the original for 30% of the movie (or more pleasing on the eye), CCE 45%, and 25% looked the same within reasonable viewing distance

I did notice in one of the DVDShrink files, the nominal bitrate was something like 78000000(how many 0's is it again? :)), whereas CCE is 98000000. would this not have an overall effect on quality?

I know you also get these noise to ratio (or whatever) values that people use to do tests but you don't watch those as you watch the images. It's whatever is pleasing on the viewer's eye and some like it sharp, some like it smooth blah blah. That's why TV's etc have settings such as contrast, colour, sharpness, brightness as most likes to change them. My mother for example has probably left her's the same as the day she got the TV, whereas I change mine quite often depending on what I'm viewing

if anyone is serious enough about getting the best quality then they would just do their own tests and not rely on others or others opinions ;). If people think something is better quality than what I think then I accept and respect that even if I don't agree as

I agree DVDShrink has improved (big pat on everyone's back involved :D) but that's about as far as I would go now. I wouldn't say any particular program is better if comparing the quality of the output as long as I'm happy with the output as my perception of quality may be completely different to the next person :p

'To Each His Own' :cool:

ddlooping

21st August 2004, 12:56

MackaDude, in a nutshell:

- These comparisons are as fair as I could make them to be.
My choice of clips was based on video content that, in my opinion, caused problems to previous version of DVD Shrink.

- The DVD2DVD-R/CCE clips have different cut points, due to the fact that DVD Shrink "Start/End" works at the GOP level, and that CCE re-encoded the original stream (different GOP structure?).

- The files have different sizes due to the way each transcoder "spreads" compression over the whole title, and in the case of CCE, the way the clips were cut (see above).

- In theory, the higher the nominal bitrate, the better.
But as you said, what matters is how the backup compares to the original, and fullfil the viewer's need for quality.

These tests were mainly done to show v3.2 improvememt over previous versions, and hopefully, so others would put their prejudices aside and decide to make their own tests. :)

MackemX

21st August 2004, 15:35

Originally posted by ddlooping These tests were mainly done to show v3.2 improvememt over previous versions, and hopefully, so others would put their prejudices aside and decide to make their own tests. :)

I think the quality from DVDShrink is tremendous if you compare it to the rest. I also like the way the user has control over the end quality with 8 permutations :D

job well done mate ;)

- The DVD2DVD-R/CCE clips have different cut points, due to the fact that DVD Shrink "Start/End" works at the GOP level, and that CCE re-encoded the original stream (different GOP structure?).

- The files have different sizes due to the way each transcoder "spreads" compression over the whole title, and in the case of CCE, the way the clips were cut (see above).

I did some quick tests with segments I cut using DVDShrink's start/end feature so I could ensure that the programs got the same video to work with. I used 5 of the settings with DVDShrink and all 4 with Recode. The original bitrate was around 4000 (trying to make them work :). The results were around 2200-2400 so I set this into CCE also. CCE was better in my eyes but this is under close examination and under normal playback I doubt you would really you notice it. I was going to upload the video samples but they are too big but I had already done a bitrate graph HERE (http://www.deano.dsl.pipex.com/dvdshrink/quality.htm)

the average bitrate is different but look at the Peak, especially Recode Basic & both DVDShrink's Smooth's. Recode is more constant whereas the smooth keeps more to the original bitrate distribution. Most of Recode's settings seem to flatten the bitrate for some reason. These are the same 3 results but shown at a different scale HERE (http://www.deano.dsl.pipex.com/dvdshrink/quality2.htm) which shows how Max Smooth keeps the original bitrate distribution but on a smaller scale

as you say, over the whole movie then the distribution may be completely different but overall DVDShrink is closer to CCE than it ever was :D

p.s. notice the times are 10 or 11 seconds, this is just the location of the cursor in Bitrate Viewer

ddlooping

21st August 2004, 18:42

Interesting test, MackemX, thanks for sharing your results. :)

jorel

21st August 2004, 21:26

@ ddlooping and all

comparing quality results using same sources for CCE(dvd2dvd-Roba) and DVDShrink
sources(toy story 1&2) was full ripped with dvddecrypter and was selected that parts to get a big file with dvdshrink without compression:
1- toy story1-main movie 01:20.32 3,414Mb (audio 5.1 portuguese 224Mb)
2- toy story2-main movie 01:32.10 4,887Mb (audio 5.1 portuguese 257Mb)
3- tin toy (from ts1)- 00:05.10 218Mb (audio 2 ch english 7mb)
4- outtakes(from ts2)- 00:05.29 296Mb (audio 5.1 portuguese 15Mb)
4- luxo jr.(from ts2)- 00:02.22 88MB (audio 5.1 english 6Mb)

full size ~ 8,906Mb

* the audio sizes are included in the video sizes,
they wasn't encoded with d2d/Roba or Shrink(of course)..same as source!

DVDShrink percentage using automatic compression settings:
dvd structure (all files together) 46,8%
1- 1,788Mb 48,7%
2- 2,397Mb 46%
3- 98Mb 42,9%
4- 135Mb 42,3%
5- 43Mb 45%
target: DVD-5(4.7Gb) (default)
performing deep analysis and choosing "smooth"(ddlopping recomendation)in adaptive quality error compensation

DVD2DVD/Roba target using the same source using undot and deen in the script: 4,480 cd size!
ps: each part was encoded separate... i don't remember the exact values and time needed to encode cos i did this compilation last month and author in tmpgencdvdauthor to build menus and chapters.

result of quality:
need to say?.....CCE is better for that big source (~ 8,906Mb)
as i wrote before, DVDShrink3.2 is really better but can't do a good job for that big source(more than 6Gb for my taste).

you (all) can choose what scene(frame from source)you want to compare....
---> just choose and i post (same scene from source,from DVDShrink and from CCE(D2D?Roba))!

ddlooping

21st August 2004, 21:50

Thanks for doing this test, jorel. :)

At a compression > 50%, the outcome was never in doubt though. :D
DVD Shrink did get better, but not that better. ;)

ddlooping

22nd August 2004, 03:05

MackemX, I just had a thought (yes, it happens sometimes, lol).
Would you mind including the CCE bitrate graphs in your comparisons?

geffroman

22nd August 2004, 06:14

Viewing with a Sony XBR 40 inch in progressive 480P mode I have now compared countless movies with CCE in 2 pass mode and Shrink 3.2 with all new features on.

It is pretty consistent... I don't like ANYTHING compressed more than 40% (That would be 60% compression in the program) But if I have to go below 60%, it is a toss up until about 55% where CCE starts to be a little better (by my eye), but it has to be some really damn good original data to suffer this kind of compression and still be clean enough (by my eye).

It is to the point that it almost doesn't matter which program (using my basic settings) so the speed and ease of use and completeness of Shrink as a program means I use it most often.

I'll still split a movie to achieve best quality if I must... and sometimes CCE and DVD-RB are the perfect tool. I am happy to have all of them in my tool box.

Just one man's experience and opinion. :D

I mispoke a little... My personal cut off in most cases using Shrink is about 60%... If the original is really good I might go to 57% or 58%...

ddlooping

22nd August 2004, 12:23

Very nice TV, geffroman. It must be quite revealing. :)

luphy

23rd August 2004, 19:30

@geffroman,

When you say those movies that Shrink did at around 55% that still looked good....what were the average bitrates for those movies AFTER transcoding?

What have you found to be a good bitrate cutoff for Shrink before CCE starts to look better personally on your equipment?

And have you ever looked to see where along the Custom Ratio compression bar these movies fall (the ones that still look good at 55%) when using Shrink?

Thanks.

ddlooping

23rd August 2004, 20:44

Originally posted by luphy
And have you ever looked to see where along the Custom Ratio compression bar these movies fall (the ones that still look good at 55%) when using Shrink?
Good call, luphy. :)

The degree to which the ratio can be lowered manually (using "Custom Ratio"), compared to the "Automatic" ratio, is usually a pretty good indicator of what the DVD Shrink output quality will be. ;)

ddlooping

23rd August 2004, 21:21

- "LOTR-ROTK" test: Automatic >> 60.9% - Minimum Custom Ratio >> 48.8%.
A margin of 12.1% before maximum compressability (as found by DVD Shrink "Deep Analysis") is reached.

In my opinion, the DVD Shrink backup was of very good quality.
I personally prefered it to the DVD-RB/CCE ones (quality_prec=16; quality_prec=20).

--------------------------------------------

- "Terminator 2" test: Automatic >> 59.6% - Minimum Custom Ratio >> 56.4%.
The automatic setting is only 3.2% away from maximum compressability (as found by DVD Shrink "Deep Analysis").

Even though watchable, the DVD Shrink backup exhibits a fair amount of "macroblocking".
In some scenes, even the least revealing viewing equipment would show it.
The DVD-RB/CCE backup might be slightly grainy, but it also shows less "macroblocking" in some key scenes.
In my opinion, an overal better backup.
However, both gave results that would warrant splitting the backup to two discs, or using dual-layer media, IMHO.

--------------------------------------------
This is obviously not a very extensive test, but the gap between the automatic ratio and the minimum custom ratio (maximum compressability) seems to be a good indication of what output quality to expect from DVD Shrink. :)

DDogg

27th August 2004, 16:51

ddlooping, I think most would agree that the maximum reduction possible is dependent on the source complexity. In one of my more manic phases I once calculated the Average BR of the original source and then did a 1% sample via CCE OPV at Q28. Depending on the source complexity, the derived ABR of the sample might show a reduction from 15% (think SPR) to 60% (think Matrix 1). When these reduction rates were then used in DVDShrink as a maximum they seemed to provide somewhat of a parallel result. That would seem to correlate with your findings, or at least I think it does. Also, just wanted to mention not everybody agrees with my thinking and methods, but there may be some merit in it.

I did not spend much more time on it as the process was way too cumbersome and required advanced use of CCE which seemed to be contrary to what the average DVDShrink user would want, but I did wonder at the time if the author might be able to do something similar to predict the maximum level of compression a source might allow as part of the pre analysis step. I'm curious how the program arrives at the maximum compression value that you mentioned?