new method to squeeze the best out of CCE

[robshot]

www.robshot.com

It's the Tweaked 2 Pass method.


cheers
rob

[gldblade]

The only problem I can think is if you're trying to make SVCDs. The first pass you use would have to have a pretty low quantization, especially since the second pass is going to be larger than the first. Unless you use more CDs...

Filesize predictability is also a little problem, but I'm sure there are ways to compensate.

[SkoalWintergreen]

Hey Roboshot, why are you changing the GOP N/M to 4 and the time code thing to all 0's? I was just curious what effect that has, and if you really like this CCE stuff you should look into making DVD 1:1 using CCE a bunch of mux problems are happening when using IFOEdit.

[kdiddy]

Even though this is really doesnt pertain to VCD or SVCD...is this a rather long method?

[Kedirekin]

You know, at first it seems daunting, but (though I haven't tried it yet) I don't think it's that bad.

A 120 minute movie will have 720 ten-second intervals. Once you've got the global bitrate about right, it's just a matter of moving through all the ten-second intervals to confirm you haven't missed any gray spots. Once you get moving, you can probably quick scan 2 intervals per second, which means all 720 will take you about 6 minutes.

6 minutes of intense concentration might be a little wearing, but it's certainly doable.

One the other hand, I don't know that this approach can be applied to SVCD. We have such stingent requirements on maximum bitrate they we probably can't get rid of the gray altogether, and increasing the global bitrate too high will probably push the encode to too many CDs. I definately need to try it.

[robshot]

I changed the n/m to 4, so the dvd authoring program will accept the m2v stream (if i'm making DVD).
As for SVCD, i use max. 2357, Q 70, and min. bitrate 0

When setting the 2nd pass, the svcd might still have the gray bitches. But I challenge anyone who can make a better quality svcd of the same movie :), also with "smaller" filesize :). Filesize predictability wont be a problem. YOu calculate the size of the audio, then give the rest to video. DUH :)

As for Ifoedit, i'm not making 1:1 DVD. The media is too expensive to just fit 1 movie, when you could do 2.... or 3 (yes I did test with three, keeping all movies with .ac3 5.1)

Cheers :)

[gldblade]

>Filesize predictability wont be a problem. YOu calculate the size of the audio, then give the rest to video. DUH

You'd have to make sure that the average bitrate you get from tweaking doesn't go over the average bitrate you want. But that's what I meant when I said "I'm sure there are ways to compensate".

[robshot]

that's why you set max. bitrate in first pass as 2357. Avg. bitrate won't shoot over that max. bitrate.

If you set avg. bitrate as 1,600, then in difficult scene it becomes 2.300 avg, then it's still safe below the max. of 2357.

And there's +- 2% difference between the estimated bitrate shown on the lower right of the advance menu, where you can see approx. filesize you will get for the entire m2v file you're creating.

Cheers

[GlenC]

Hi Robshot

Nice guide!!!, what do you use to remux the ac3 and m2v back together?, avoiding the syncing issues that some ppl are having?

What do you use for the dvd menus?

Cheers

Glen.

[robshot]

for PAL mpeg2, where there's no 2:3 pulldown involved, I never have syncing problem. Since I use the original audio, and then compared parameters of the original vob and the resulting .mpv from CCE.

When I mux it with dvd authoring software, it sync all the way to the tits :)

For NTSC, i tested several with DVD2AVI doing FORCED FILM. The resulting CCE file will be 24p progressive. Then I use pulldown.exe to add the pulldown flags. the new file is now 29.97fps NTSC with pulldown flags. Mux in autohring, and get also 100% sync file. To the tits!


Cheers

[Deckard]

Hi Robshot,

I read that you've tested backing up three full movies with ac3 on a single 4.3 GB DVD-R with unnoticeable video quality drops. At full resolution and anamorphic? What did that leave you for the average bitrate for the video streams?

Cheers,

Deckard

[tai]

Robshot: what about the other settings for SVCD? What do I need to set for DVD compliant, Zigzag scanning order, Quant scale, DCT precision, Image Quality Priority ?

[robshot]

only change the dvd-compliant (uncheck it)

Auto DCT stays
4:3 or 16:9 depens on the original source.
But I strongly suggested letterboxing an anamorphic source, for making SVCD. It would save much bitrate for actual image (black bars won't eat up much bitrate).

Zig-zag yields better image.

On first pass, Q value is 70, min bitrate 0, max. bitrate 2357
Don't forget to resize your source before input to CCE

As for 3 movies on 1 DVD-R, avg. bitrate ranges from 2,000 to 2,300. But the point is to follow/match Q to not exceed 9 (on 2nd pass).

Cheers

[Airw0lf]

forgive my ignorance if you will, but I'm not 100% sure about how this method will help SVCD makers. I can understand if you're making a DVD this can come in very handy. In such a situation you have a large bitrate range to play with (0 - 9800), so you can carry on increasing avg bitrate a fair bit (killing most of the grey bits in the process) until you run out of space.

But in my case I always thought with SVCD you'd be best off simply setting average bitrate to equal the highest value possible given both the SVCD spec and the number of CDs you're using.

I normally use a bitrate calculator and find my required video bitrate average. Then i set my minimum as 0 , and my maximum as 2600 (i use 112 kbit audio).
So what really is the point of setting the first pass in the range 0 -2357 and subsequently tweaking, because when you play with the settings in advanced you will inevitably end up either hitting the maximum allowable SVCD video bitrate or your maximum possible average bitrate?

Following your method, I would probably do a first pass from 0 - 2600.
Then I would hit advanced, and increase average bitrate until I reached the limit provided by my calculator?
How is this different from selecting multipass vbr from the very beginning and setting min = 0, max = 2600 and the calculated average?

[sokrates7]

robshot : excellent approach, I tested it on various films and it is stunning !

However, I have the problems mentioned in DVD2AVI thread already that prevented me from keeping the resulting files from your method. The aspect ratio is wrong whatever I do. Obviously DVD2AVI does not respect the ratio in 8 of 10 films I have checked (EUROPEAN PAL 25 16:9). It always does resizing to 4:3 without asking even with the 1.83 release as vfapi files do not use the resizing information collected in this release. Any subsequent parameter in CCE does not give me back the original ratio and I really don't fancy those eggheads :))

Do you have any idea how to solve this, maybe by using a different prog as frame server ? Thanks in advance !!

[jdobbs]

sokrates7,

If your original is 16:9 you should also set to 16:9 encoding in CCE.

You might also try frame serving your DVD2AVI output with AVISynth and the MPEG2DEC plugin. You can then use the script to resize as you see fit (see the AVISynth forum). Here's an example that makes a 4:3 copy of a 16:9 original (NTSC/FILM).


# This loads the MPEG2DEC module
LoadPlugin("c:\program files\dvd-rip\avisynth\MPEG2DEC.DLL")
# Read directly from the DVD2AVI file
mpeg2source("d:\image7\Apocalypse.d2v")
# Required for some processors (AMD)
ResampleAudio(44100) # Adjust for AMD Bug
# Changes 16:9 to 4:3 format
BilinearResize(720,360)
# Add borders to get back to 720x480 size
AddBorders(0,60,0,60)

If you really want to fit a lot of movies on a DVD-R change line 8 to read "BilinearResize(352,360)" It gives good quality (not great), creates a file that is DVD compliant, and you can shrink the required bitrate even more dramatically using robshot's method.

[sokrates7]

thanks jdobbs..

the 16:9 setting in cce does not change anything. if the frame served avi does not have this format it will not modify it. if i am informed correctly it does just related to the way the pixels are treated.

i am very thankful for your hint regarding avisynth, i will try to modify the script in order to receive the 16:9 back that dvd2avi had stolen me ;)

is nobody here who would like to modify dvd2avi in order to push a chosen aspect ratio also into the d2v file ??

[robshot]

Airwolf:

forgive my ignorance if you will, but I'm not 100% sure about how this method will help SVCD makers. I can understand if you're making a DVD this can come in very handy. In such a situation you have a large bitrate range to play with (0 - 9800), so you can carry on increasing avg bitrate a fair bit (killing most of the grey bits in the process) until you run out of space.

This method will surely help SVCD makers.


But in my case I always thought with SVCD you'd be best off simply setting average bitrate to equal the highest value possible given both the SVCD spec and the number of CDs you're using.

It's a simplifying settings. Setting avg. bitrate to equal highest value possible by SVCD spec give you large size. Not all movie were created equal. This method will save CDRs :). One can make 2 SVCDs instead of 3, and can still compete (and WIN) over the 3 SVCDs.

I normally use a bitrate calculator and find my required video bitrate average. Then i set my minimum as 0 , and my maximum as 2600 (i use 112 kbit audio).
So what really is the point of setting the first pass in the range 0 -2357 and subsequently tweaking, because when you play with the settings in advanced you will inevitably end up either hitting the maximum allowable SVCD video bitrate or your maximum possible average bitrate?

Setting 1st pass with 0 - 2357 range is, in other word, to let CCE decides which scenes requires high bitrate, which require low (or VERY LOW) bitrate. This will give you an approx. filesize, which you can tweak in 2nd pass when you hit "advanced" button and then manually decides what avg. bitrate you want, taking into account the approx. bitrate it would give.

Following your method, I would probably do a first pass from 0 - 2600.
Then I would hit advanced, and increase average bitrate until I reached the limit provided by my calculator?
How is this different from selecting multipass vbr from the very beginning and setting min = 0, max = 2600 and the calculated average?

Using the calculator to give you avg. bitrate and # of CDs needed is, again, a generalization way to encode SVCDs. Using the method I wrote, you can set bitrate to a lower avg. bitrate as told by your calculator, thus creating overall smaller size. Perhaps you're wondering about the quality by now. Well, I've got emails telling me that with this method, peeps are making 2 SVCDs out of their usual 3 CDRs and the quality is still _awesome_. The point is, while calculator gives you quick answer to how many CDRs you need and maxed out agv. bitrate, not all movies need such "max out avg. bitrate."

Cheers

ps. I have been trying to answer this post yesterday, but the bbs keep shoing "no message thread" page :/

[robshot]

Sokrates7

It's not the DVD2AVI problem.

It's a mistake when you're doing DVD 2 SVCD aspect ratio calculation.

By default, An anamorphic DVD will still have 4:3 frame size. This is the coneheads :). But mpeg2 stretches when playback. into something like 854x480 or something.

Now, when you're doing SVCD, its 480x480 (or 480x576). This requires correct re-calculation of scaling. I use the long but proven 100% correct using photoshop, and 1 captured .bmp of the movie. With photoshop, i do rescaling until i got the correct IMAGE size. Then I apply this for SVCD framesize.

My take on anamophic movie is to be letterboxed when encoded to SVCD. This will save space as the blackbars requires small size :). Also, not all DVD player can play anamorphic SVCD.

Cheers.

[mb1]

@ robshot
why don´t you try FitCD
http://members.tripod.de/fitcd/
Calculates correct aspect ratio and gives you further improvements to your method, which doesn´t consider tv-overscan and the cce-bug (see fitcds readme too)

tv-overscan area is about 8% which means that considering it you can lower bitrate up to 8% and still get the same picture quality.

[sokrates7]

robshot
thanks, I see more clearly now although I have unterstood need to read more basic information about dvd. some things are still not very well understood....

mb1

this prog looks VERY nice. thanks !

[Evaldas]

hey robshot,

what is that program bitrate viewer? where can we get it?

[Evaldas]

And another question,

as I see from the pic (http://www.semutrangrang.com/robshot/04.jpg) on robshot.com, vfapi reader codec which you have reads *.avs, *.vdr, *.d2v and *.tpr files, while version found on doom9 site reads only last two. Could you point me to version you have.

[Deckard]

@Evaldas
Use Google and you'll find bitrate viewer and search for VFAPI 1.03b

@Robshot
I've noticed that DVD's are encoded with a nonlinear quantizer scale. What I found on the internet was that a nonlinear quantizer scale is used for further bit reduction. I'm not too sure if using a linear scale is really needed for your method. A linear scale might produce good quality but at the cost of a higher bitrate. Maybe a non-linear scale could give you better quality when encoding at the same avg bitrate. I mean, isn't that the reason a non-linear scale is being used to encode DVD's?

Deckard

[Ferrit]

Hey everyone, i see that you all trying to go further and optimize this method.
Please do post on what you achived. it interesting me coz i did most of the test, and the results are on robshot.com

[mb1]

Ok, since I feel not really understood I´ll post a short description of my way to encode svcds (which is established in advanced german encoding area) ...

Here´s a short explained example (movie "episode 1", 2,35:1; PAL):

-ripped with smartripper
-d2v with dvd2avi (no cropping)
-loaded into avisynth (with mpeg2dec.dll)
resolution 720x576; coded film pixel 718x438;

FitCD-Script:
LoadPlugin("*YOUR PATH*\mpeg2dec.dll")
mpeg2source("*.d2v")
BicubicResize(448,318,0,0.6,4,69,712,438)
#TemporalSmoother(3,5)
AddBorders(16,129,16,129)
#Trim(0,0).FadeOut(150)
#ResampleAudio(44100) # CCE 2.5 'fix' for Athlons

Crops to 712x438, resizes to 448x318 (CCE-Bug !) and adds borders to 480x576 for SVCD.

Now loaded the avs into CCE 2.50 and encoded with this matrix (tsunami cce-patcher !):
Intra
8 16 19 22 26 27 29 34
16 16 22 24 27 29 34 37
19 22 26 27 29 34 34 38
22 22 26 27 29 34 37 40
22 26 27 29 32 35 40 48
26 27 29 32 35 40 48 58
26 27 29 34 38 46 56 69
27 29 35 38 46 56 69 83

Non intra
16 17 18 19 20 21 22 99
17 18 19 20 21 22 23 99
18 19 20 21 22 23 24 99
19 20 21 22 23 24 26 99
20 21 22 23 25 26 99 99
21 22 23 24 26 99 99 99
22 23 24 26 99 99 99 99
23 24 25 99 99 99 99 99


or this one

Intra
8,13,15,19,19,23,23,29,
13,13,17,19,22,23,27,29,
15,17,19,22,23,27,29,35,
17,21,22,23,27,29,35,37,
21,21,23,27,31,33,37,99,
21,25,27,31,33,37,99,99,
25,25,31,33,37,99,99,99,
25,31,33,37,99,99,99,99

Non intra
8,11,13,15,15,19,19,24,
11,11,15,15,19,21,24,25,
13,15,17,19,21,24,25,34,
13,17,19,21,23,26,34,99,
17,19,21,23,26,32,99,99,
17,21,23,26,32,99,99,99,
22,22,28,30,99,99,99,99,
22,28,30,99,99,99,99,99

Then you have to mux the encoded videostream and
audiostream with bbMPEG using the above used intra
and non intra matrices (settings - in and output files),
´cause patched cce 2.50 could not write those matrices
into the header which is recommended for correct decoding.

-if you want use robshots optimizations within cce
-author normally and burn
thats it !

regards
mb1

[shh]

just my 2cents :D:

When encoding interlaced, one should prefer the Anternate-scan mode
For progressive, the ZigZag method is the best.

I can't believe that reducing the GOP-length to 12 leads to better quality. The extra I-pictures just cut bitrate, which could easyly be saved with P-pictures.

An 'Image Quality Priority'-value of 5 is too high. Even for the good DVD-sources. This would lead to more mosquito-noise at the edges.
I'd say use 10-15 for excellent sources. Otherwise 15-20.

Also, using the 'Anti Noise Filter' (which seems to be a quanisation-filter) always helped enhancing quality.
Using a value of 2 is *absolutely invisible* on the PC-monitor, but enhances the quantisation-level (and therefore file-size and encoding-efficiency)

I personally prefer a 'Non-linear' quantisation.
Yes, 'linear' looks a little bit less noisy, but non-linear-scan shows a lower quantisation. And if you let CCE decide the bitrate itself (VBR-1pass, you enter a Q-value...), it produces smaller file-sizes with alternate-scan.
And, even if you go to a 10cm distance from your TV-set, this additional color-noise is absolutely not visible.

Additionally one should use an 8-bit DCT-resolution.
The additional precision of 2bits isn't visible on the TVs and, why should I allow the encoder to waste some bits for DCT-precision (wich isn't visible) where it could use those bits to encode *more DCT-coefficients* instead of dropping them at quantisation...

Cheers
shh

[Ferrit]

nice view, but i did the tests for 2 months as mentioned on the guide.
and any way, i always do progressive, even if the source is interlaced, and settings dc to 10 bits gives better results (i have Progressive TV).
besides, u dont need noise filter if u do backup of dvd, and for quality is better use 5 and not higher, it gives better results on complex frames.
any way, the second pass is the most important thing, just try, then come back and tell me if i'm wrong...

Cheers

[Twizllur]

What would be the best ways to add a step for resizing into that formula?..
avisynth
virtualdub
aviutil

or they all just about the same ?
-thnx

[mb1]

Partially agree to shh and partially to Ferrit.

But deinterlacing for mpeg2-encoding is nearly always a big mistake.
Look at these pictures:
http://members.tripod.de/mb1xsvcd/deinterlacing.htm
Sorry, but site is in german.
But I think you can recognize various deinterlacing-methods.

[shh]

> but i did the tests for 2 months

Well, we have been testing for over 1 year now. :)
These aren't only my personal results...

> DCT-auto/10, i have Progressive TV...

Ok, that's possible. I can't proof this.

> u dont need noise filter if u do backup of dvd.

Many DVD-sources are quite bad and really need a filter for better quality.(For that ones I would recommend a denoise of 4-6). DVD-sources also come with compressions of 1:30 and therefore already have artifacts and color-noise. Whether from compressing or from noisy film-source.
The problem is, that the encoder detects this noise as detail and wants to retain it. (And with your low image-quality-priority-value it even wants to retain more noise). We are not talking about real filtering... the value of 2 just cleans up the encoding-artifacts of the source a little bit. As I said, this de-noising is invisible, but it enhances the picture so that the encoder can concentrate on real details to retain...

I also must say that deinterlacing or coding interlaced pictures in progressive-mode is a bad thing. Fist, you don't have the special (field-based) motion-compensation.
Second, the interlaced-lines (we call them combs here) are (again) detected as detail. But the fields are just weaved together to one progressive picture and you have high-frequency-changes at the edges. High-frequency-changes are dropped/reduced at quantisation. Result is more stair-case effect at the edges and even more mosquito-noise.
This all isn't necessary... Why do you think they have made an extra interlaced-mode? ;)
Verify the interlaced mode yourself, you'll be surprised.

> for quality is better use 5 and not higher, it gives better results on complex frames

That is true if you have enough bitrate. But for SVCDs (what I make) the bitrates often are too low to 'describe the picture in the complexity' you want the CCE to. Result is mosquito-noise at the edges (better visible, if you not only look at one frame, but switch between them). Also the bitrates for general SVCD-playing on standalones are very restricted and going down to 0 or up to 9800kbps is not possible. (you know, only a few can play miniDVD or SVCDs @ 4500kbps...)
But, I will try your method and look if it enhances quality (for me)
Can you give me an example of which bitrates for what filmsize you've used, (also the black-area) please?
BTW, if your method really enhances the quality revolutionary (still to proof for me ;)), one could write a program to do the settings automatically on the .ecl-project-file...

Cheers
shh

http://members.tripod.de/fitcd/MPEG/1996-03.pdf
http://www-mtl.mit.edu/Courses/6.095/spring-00/unit3/dct.html
http://www.bbc.co.uk/rd/pubs/papers/paper_14/paper_14.html

[The_Flash]

I'm fairly happy with my encoding output from CCE, but I'm still confused about how to reproduce a 16:9 output file. If the source is 16:9, I set to 16:9 in CCE, but the resulting file is still detected as 4:3 when played back in WinDVD after being authored. Changing my .avs script to resize to 720,360 and then adding borders fixes the 16:9 for 4:3 output, but I'm wondering what I'm doing wrong or what settings I need to keep the 16:9 format of the source. Any help is appreciated, thanks.

[sokrates7]

The_Flash

Please look at robshots post earlier in this thread. Also http://www.doom9.org/aspectratios.htm
might help you.

I had the same problem ;)

[jse]

Originally posted by shh
When encoding interlaced, one should prefer the Anternate-scan mode
For progressive, the ZigZag method is the best.

I can't believe that reducing the GOP-length to 12 leads to better quality. The extra I-pictures just cut bitrate, which could easyly be saved with P-pictures.

An 'Image Quality Priority'-value of 5 is too high. Even for the good DVD-sources. This would lead to more mosquito-noise at the edges.
I'd say use 10-15 for excellent sources. Otherwise 15-20.

Also, using the 'Anti Noise Filter' (which seems to be a quanisation-filter) always helped enhancing quality.
Using a value of 2 is *absolutely invisible* on the PC-monitor, but enhances the quantisation-level (and therefore file-size and encoding-efficiency)


THanks for the tips. I tried robshots method with material from my DV camcorder. Using the initial settings (except for progressive mode), I got too many gray areas under advanced settings. I wasn't able to remove then even if I set average bitrate to 9000. I then tried to encode first pass with GOP=15, Image Quality Priority = 20, AntiNoise=6, alternate scanning order, and non-linear quantisation. Going through the advanced settings again, I still had a rather high Q value. There were just a few grey areas to deal with though, and average bitrate could be set lower.

Obviously, the material from the DV tape wasn't easy to compress, and I've later tried with other clips. I always tend to end up with more compresasbility and lower Q values when using the latter settings, mentioned above. I've not tried with MPEGII files from a DVD though, perhaps that would change the result.

[d121]

I want to try robshot's method to encode avi file to mpeg2. My CCE2.64 is a plugin version to Premiere 6.0 where I could not find the "Advance" button to fine tune the bitrate. Could someone advise me how to do it?

[Evaldas]

Which intra and non-intra matrices (the first one or the second one, posted by mb1 few posts above) you would recommend for SVCD encoding with CCE 2.64 SP?

[Arky]

Hi, just a bit of info to contirbute to the linear-vs-non-linear quantization debate. Although most of us don't care to read it, because we cannot afford the hardware, it is still amazing what one can learn from apparently irrelevant sources!

...In the spirit of this, the following is taken directly from the Spruce MPX-3000 (hardware) encoder PDF manual:


**********************************************************************

"There are two types of quantization available in MPEG-2 video encoding: linear and non-linear (which is the default). MPEG-1 useslinear quantization only. MPEG-2 has both types of quantization available - though non-linear is used almost exclusively, as it provides greater dynamic range.

Note: When MPEG-2 was defined, the non-linear quantization scale was introduced as an across-the-board improvement on the previously used linear scale. The option of linear scale in MPEG-2 was defined only to accommodate full backward compatibility with MPEG-1. **There are no conditions known under which a linear quantization type would be preferable to Non-Linear type in MPEG-2"

**********************************************************************

Let me make it clear that I do not offer this contribution as a criticism of Robshot et al's method - quite to the contrary, I congratulate them on a job well-done - a decent guide to cinemacraft, for us enthusiasts, has been needed for a long time and i was thrilled to see the excellent Robshot et al guide. It is the very nature of our little community that we advance our hobby through discussion and experimentation, and it is from this standpoint that I would be eager to see the Robshot et al Cinemacraft method/guide developed/honed further. Make of it what you will.

Thanks again for the excellent work.


Arky ;o)

[Arky]

I appreciate that this is the CCE forum, but I nonetheless feel that this link is pertinent to the discussion above:

www.flexion.org/cgi-bin/UltraBoard/UltraBoard.cgi?action=Read&BID=33&TID=6034&P=1&SID=48582#ID6

I refer, specifically, to Ugly's remarks about a new hardware encoder which is allegedly very good value and excellent for DVD conversion...


Arky ;o)

[mpeggato]

HI Robshot

I am newbie, i tried CCE 264 crash
with CCE 262 works well

I tried to encode a DVD film of 105 min. PAL with your method, i see this method in this forum and if it is exact explain how encode or better reencode in mpeg2 to save space in Kilobytes without touching the DVD compliant specifics ............ ok .

I maked a file from 3,5 Gb to 2 Gb at 720x576 25 frame/sec without audio.

What is the PROBLEM?

Is this: The quality is not special or similar of the original but not good and dark too, the second and real problem is the motion during the play, the playing is like "little stop and go" (scuse me for my english i am italian); What's or Where I mistake.

Please explain me the method right way.

Mpeggato :(

[talman]

I assume you use an authoring program with the resulting mpv file and wav file. Questions: is the wav file a true dolby digital 5.1 file (assuming it was on the dvd) or is it a decoded 2 channel dolby digital file? Second question: are there any decent authoring programs out there that will work with a 5.1 file? Thanks for the help.

talmania23@yahoo.com

[SiC]

You don't end up with a wav file in the first place. Also there is no such thing as 5.1 channel wav file. When you load the *.vob file into DVD2AVI, you'll end up with an *.ac3 file (Dolby Digital) and it could be 2 channel (192kbps) or 5.1 channel (448kbps) depending on which stream you choose. Low end authoring softwares only accept the 2channel *.ac3 files but higher end programs like DVD Maestro & Scenarist will accpet the 5.1 channel *.ac3 file.

[talman]

Thanks SiC!! I didn't think you could have a 5.1 .wav file---DVD2AVI must have grabbed the wrong audio track. Back to the drawing board...

[D0Hbert]

I've tried encoding with CCE with half-d1 resolution. What I basically did was make a project file off the m2v with dvd2avi. Then I wrote a AVS script loading the project file, then frameserved the avs using avisynth. I loaded the avi in CCE and encoded my movie. My question is how come the colors I got from the encoded movie all looked oversaturated? The reds are so much redder, etc. When I loaded the frameserved avi with Tmpeg, encoded a few mins of the film and compared the results, the color output from Tmpeg looked natural. Then I tried my old tool for doing half-d1 outputs, Rempeg. Again, color I got off Rempeg matched those of Tmpeg's. CCE's output still looked unnatural. Is there anything wrong with what I did? I'm using 2.64 of CCE. For setup, I basically followed Rob's tutorial, except for the bitrate, half-d1 output doesnt need that high a bitrate setting. Any input is greatly appreciated, thanks.

[Milkman Dan]

How did you get 2.64 to accept avs files? 2.62 and 2.64 just screamed at me when I tried to use them. Had to go back to 2.50 and use the resampleAudio trick...?

[Evaldas]

I'm using CCE 2.64, but i frameserve to it with vfapi 1.04. I create avs script and then make fake avi with vfapi. Since vfapi don't accept avs files, only d2v and tpr, you need to install tmpeg avs reader (http://www.math.berkeley.edu/~benrg/avisynth/tmpgenc-readavs.zip) and it works like a charm, theres no more need for using tmpeg and tpr files.

[Deckard]

Did you use a luminance level of 0 to 255 as mentioned in the Robshot guide? It seems that using a luminance level of 16 to 235 in CCE produces a result that's closer to the original.
Link: http://www.uni-kassel.de/~eckhardm/hq.htm (in German but with charts)

Deckard

[D0Hbert]

Thanks for the info. Yes I used 0 to 255, but aren't we supposed to actually use that setting to get all possible colors available from the source? 0 being all black, etc?

[arcman]

How I have understood this is that if your original source was for TV use(DVD for example) then you use 16-235 color area to get the colors right. This limitation comes from the TV standards? Or was this actually so that the destination actually counts. I.e. you want to show your work in TV then you use this 16-235 but if you want to show it on computer then you use 0-255.... Uh now I got confused... :confused:

Arc

[Evaldas]

I still don't understand how usefull can be robshot's method for SVCDers. Since I can't use higher bitrate to get rid of grey bitches, cause I want to fit to 2 cds, not more.
And how helpfull can be setting min and max local bitrate at the maximum (2357 kbps)? If the GOP gets max bitrate at the first pass, so why it should't get the same on the second pass.

Sorry for terrible English, I hope you understand what I mean.

[Deckard]

@Evaldas

The guide is helpful when you have a DVD player that can handle SVCD's with a high bitrate. My Amoisonic player can handle SVCD bitrates up to 5000 kb/s. This means I can still choose a low average bitrate when doing multipass encoding but with a high max bitrate. So for difficult scenes the bitrate will go up to the max bitrate and you will end up with less gray areas. In this way I get good encodes that fit on 2 cd's.
For two cd encodes I use a luminance setting of 16-235, DCT - 8, and the default non-linear quantizer scale. Use the Fitcd program advised by mb1 in a previous post and use a matrix for low bitrates like the ones mb1 mentions.

Cheers,

Deckard

[Evaldas]

@Deckard

Thanks for your explanation. Now I got the point. Cause I still get the grey areas in the areas where tha maximum bitrate is used and they still look blocky.
I use the same settings as you and I use Andreas matrix. I did some tests at marginaly low bitrate of 1200 kbps and Andreas and CCE standart matrices looked better than others.

[lgcbmb]

@arcman
regarding luminance levels..

0-255 luma is standard for dv-cams, while 16-235 luma is standard for devices that conform to itu-r 601 (dvd?). you can see this in action by applying the levels filter in vdub, and setting it to; input 16.1.235. the resulting output will look better then 0-255, and will display less block noise in a final encode (my experience). i dont have a dv-cam to see how its luma works first hand. i guess this means you should always use 16-235 luma while encoding dvd source, and 0-255 for dv-cam source.

[Mr_Grimm]

First of all, Great Guide robshot.

THe only problem I'm having is that the time displayed final product is less than half of the original DVD running time. When I play the new DVD files with PowerDVD4, it's REALLY choppy and the sound is out of sync (the TIMECODE was set to 00:00:00:00). Everything seems ok before I start my second pass (the length of the movie is displyed correctly when I'm adjust the final bitrate). I thought it might be the way I frameserved the movie, but I tried it the other way & got the same results (I used both the VFAPI reader codec and the .avs). The only thing I can think of is maybe IFOEDIT89 doesn't like the .m2v CCE-SP250 creates.

Any suggestions would be greatly appreciated.

[Doom9]

@mr_grimm: you already found the solution and even posted it... sort of. You cannot remux the CCE generated m2v file with ifoedit.. you'll get the effects you described. Currently the only way to use a CCE generated m2v for your DVD projects and still retain some extras and menus is described in my CCE & IfoEdit guide.

[raistlin]

where can i dl tsunami cce-patcher ?

[Mr_Grimm]

I've been trying to backup Jay and Silent Bob for 2 days now. This movie is half FILM xx% and half NTSC xx%. I did some research and decided on this process:

1) ifoedit 0.9
Ripped only the movie and the Dolby Digital track to vob files

2) DVD2AVI v1.76
Set Field Oper.=NONE, Demuxed Track 1, and Save D2V project.

3) Create .AVS File
Here is the content of my AVS file
LoadPlugin("C:\LOCATION OF.....\mpeg2dec.dll")
LoadPlugin("C:\LOCATION OF.....\Decomb.dll")
mpeg2source("C:\LOCATION OF....\*.d2v")
Telecide()
Decimate(mode=1,threshold=50)
ResampleAudio(44100)

4) Processed AVS file with CCE 2.50 using robshot's method
(Speed was cut in half from 1.9 to .9)

After the 2nd pass with CCE, I imported the mpeg file into Maestro (just to see what the framerate was). It shows the framerate as "29.97 FPS DROP" and it's approx. 7 seconds longer than the demuxed audio file from DVD2AVI (Video time = 01:44:24 and Audio time = 01:44:17). I also got a I-Frames error when I tried to import the chapter timecodes.

So now I'm lost and don't know what to do (HELP!!!).


BTW, I've got a couple of Questions

1- Is there a problem with my AVS file? Do I need to make adjustments or am I heading in the wrong direction?

2- In CCE (the VIDEO settings) , which box should I check "Upper field first" or "Progressive frames"? I chose Progressive frames because I assumed the Decomb process would remove all the interlaced frames.

Thanks in advance

[DVDHack]

What has happened to the RobShot web site, It was very useful and seems to have disappeared. Isthere a new link?