Blu-Ray~DVD Backup & Conversion

[laserfan]

He said it does seem faster but not what he was hoping for.Thanks; yeah I guess when we spend precious $$$ on upgrades we tend to look for 2x-or-more improvements. At least I do (or will, expect maybe even 8x speedboost over my P4).

[Capsbackup]

With all this discussion of processors and encoding speeds, I was just wondering if there is any current news on how BD Rebuilder is coming along? There was silence for quite some time, and curiousity, with some patience, is getting the better of me. Maybe "Santa jdobbs" is coming to town?

[Fishman0919]

Thanks; yeah I guess when we spend precious $$$ on upgrades we tend to look for 2x-or-more improvements. At least I do (or will, expect maybe even 8x speedboost over my P4).

You may want to take a look at the new AMD AM3. First reviews and reads look very good.

[Wombler]

A friend at work just built one... Intel Core i7 940... very close to his old Intel Core 2 Quad Q9650 but takes advantage of DDR3 mem.


Damn... I was hoping it would be more clear cut than that.

I'm long overdue an upgrade and was thinking about building a new system with the focus being very much on video processing speed.

I had it in mind to go either for the state of the art in relation to the current generation, or bottom end of the new generation (i7 920) which could be more easily upgraded later.

From what you're saying I'll either have to set the bar higher or content myself with the current technology.

I'm in no immediate hurry though so I think I'll have to have a look at the AM3 and see what the story is with it before I make a decision.

BTW I found this (http://www.tomshardware.co.uk/Intel-Core-i7-Nehalem,review-31404-32.html) for those that are interested in benchmarks.


Wombler

[Sophocles]

From what I've heard the Intel Quads seem to be a better choice than the AMDs... my Phenom is ok -- but I'm hearing the Q6600 overclocked is significantly faster.


I have a water cooled Intel Q9650 (EO stepping) overclocked to 4 GHz that is bringing two pass encodes down to about 4.5 hours. The new i7 chips will provide some progress, but it's probably going to take an 8 core processor with compatible multithreaded software before we are on par with today's DVD backup times. There is also the possibility that GPGPU parallel processing will come along to offset the CPU's weak points and really pick things up. For folding a single decent Video card with CUDA drivers is about 20 times faster than the fastest Quad CPU.

[Wombler]

There is also the possibility that GPGPU parallel processing will come along to offset the CPU's weak points and really pick things up. For folding a single decent Video card with CUDA drivers is about 20 times faster than the fastest Quad CPU.

Yeah I know that and it would make my choices a lot simpler as I could just plough a larger proportion of my money into a new video card but I think I recall Jdobbs stating somewhere that he has no plans to develop this at present. :(


Wombler

[Sharc]

Seems the original scope of this thread has been lost .....

[writersblock29]

@Sharc

Well... yes and no. Sure, this isn't the right place for discussing hardware specs--but on the other hand, it offers insight to how speedy a given project can be expected to complete. I know that if Jdobbs offered a release of BDRebuilder right here and now, the very first thing he'd hear is, "Good golly, this takes forever." From that point on, the only thing people would focus on is the speed at which they could get from point A to point B. It might help to ease their minds (and pocketbooks) to find out that those of us with high-end equipment are just as frustrated as they are, and upgrading really won't ease that much for them. At least, not at this point.

I do understand your point, however. On the other hand, this thread really hasn't remained focused on one aspect for very long: DVD-to-Blu ray, DVD upconversion-to-Blu ray, and now Blu-ray to AVCHD DVD as well as DVDs with true Blu ray content on them. I see it as an examination of possibilities, with an eye on what's available and what people will want to use. It's bound to wander... but there's lots to think about in this area of backup creation, and a lot of good input on this thread.

[DVD Maniac]

No harm in filling the wait for the final release :) Perhaps jdobbs can provide an update on progress?

Im running dual core AMD and only just upgraded so would prefer to wait and see how that performs first. I for one am less stressed about encode times. Remember the good old days of REMPEG and 12 hour encodes you left running overnight :rolleyes:

- its really no big deal providing the end result is a high quality one which works on your set-top players.

Anyway, i'm waiting patiently, but I am sure we would like a little Christmas toy to play with

Jdobbs

WE want pressies
Pressies!
Pressies!
Pressies!
Pressies!
Pressies!
Pressies!
Pressies!
.........and more pressies!

[Furiousflea]

The AMD 9950 Black Edition or Reg ver runs at 2.6 norm.... The Black Edition lets you set the multiplexer up to 6.4.

I had it running at 4.0 but was have issues runing x264 and CCE.... backed it down to 3.8 and works fine.

Its called a multiplier not a multiplexer...

Sorry that was annoying me :)

Also, please don't spread false information a 9950 can't hope to compete with a Q6600 for overclocking potential, or even at eqally clocked speeds the Q6600 is a good 20% faster for video encoding - at least!

Most Q6600s do 3.4-3.6Ghz on air, with the majority hitting 3.6Ghz. Some even higher.

On water they nearly ALL do 3.8-4.2Ghz.

A Q6600 at 3.8Ghz would be about as fast as your 9950 clocked at 4.5-4.6Ghz.

No disrespect, and I'm sorry to troll, but if your not speaking from first hand experience (I'm not either but I have done my research) please don't make incorrect assertions.

[Furiousflea]

No harm in filling the wait for the final release :) Perhaps jdobbs can provide an update on progress?

Im running dual core AMD and only just upgraded so would prefer to wait and see how that performs first. I for one am less stressed about encode times. Remember the good old days of REMPEG and 12 hour encodes you left running overnight :rolleyes:

- its really no big deal providing the end result is a high quality one which works on your set-top players.

Anyway, i'm waiting patiently, but I am sure we would like a little Christmas toy to play with

Jdobbs

WE want pressies
Pressies!
Pressies!
Pressies!
Pressies!
Pressies!
Pressies!
Pressies!
.........and more pressies!

It's going to take more like 3-4 days on a dual core AMD, even 3Ghz+ clocked ones.

[Furiousflea]

You may want to take a look at the new AMD AM3. First reviews and reads look very good.

About as good as a Q6600 from my reading, clock for clock. Again not as good at overclocking though.

[jdobbs]

It's going to take more like 3-4 days on a dual core AMD, even 3Ghz+ clocked ones.It gets slow, but nowhere near that kind of time. On my Dual Core AMD clocked at 2.4GHz it takes about 20 hours to do a typical complete Blu-ray backup, and less for a movie-only.

[jdobbs]

No disrespect, and I'm sorry to troll, but if your not speaking from first hand experience (I'm not either but I have done my research) please don't make incorrect assertions.Actually in an earlier post (#542) he said he has both... so I would take that to mean he is speaking from first hand experience.

[Capsbackup]

@jdobbs
I know its been asked several times before, so excuse my impatience, but is there any more recent estimates on a release date for your BD-Rebuilder program? I can't be the only one anxiously awaiting your release!!

[prettyboy85712]

@jdobbs
I know its been asked several times before, so excuse my impatience, but is there any more recent estimates on a release date for your BD-Rebuilder program? I can't be the only one anxiously awaiting your release!!

Same here. I've got my new i7 machine standing by.

[MadMonkey57]

@jdobbs
I think I've read somewhere in this thread that you planned on developping you're own muxing tool within bd-rebuilder. Is that correct?

[jdobbs]

Yes -- but it isn't complete yet. Right now BD Rebuilder is using TSMUXER.

[moviefan]

What is this tool called, created by jdobbs, that fixed some bug in tsMuxer? In what cases is this needed and where can I get it?

[MadMonkey57]

jdobbs created a tool called fixclpi (available in the forum :search:). It fixes an internal table in movie clips generated by tsmuxer. Some players (like PS3) seem to be doing fine without it. Some others (I don't remember what models of Sony's) require it to support proper fast forward / rewind and chapter seeking. Some others (Panasonic's BD30/35) seem to live better without fixclpi (firmware maybe)...

You could then use it in case you encounter problems with your creations (F-FORW/REW, chap. seeking) or if you seek better "BD spec" compliancy, please let us know how it goes if you give it a try.:thanks:

[Fishman0919]

About as good as a Q6600 from my reading, clock for clock. Again not as good at overclocking though.

My AMD am2 9950 Black Edition is faster then my q6600... much faster... and OC's much better... lookie at my "AMD 9950 OC'ed" below. The AM3's are faster then that.

[Vanderlow]

I wonder how a blu-ray movie will look backed up via .264 on a DVDr 9? It's got to look better then a regular DVD, right? and this is something BD Rebuilder will do (when it's out)?

[Chefkoch_ico]

At least movie only you can do some pretty neat things.

I did this myself with HD-DVD some time ago, however, to not generate artefacts I reduced the resolution to 1440x1080p24 anamorphic. Dont know, if thats in BluRay spec. But with this I could get a 140 minute movie with great quality on a DVD9.

For me is a resize to BD-R25 important, since its more compatible to Standalone BluRay Players and in 99% the quality will be really great, even with menus and extras. And they are not that expensive anymore.

Bye

[jdobbs]

I wonder how a blu-ray movie will look backed up via .264 on a DVDr 9? It's got to look better then a regular DVD, right? and this is something BD Rebuilder will do (when it's out)? If it's movie-only you can't tell the difference between the reencoded movie (1920x1080p) on BD-9 and the original... I've done a lot of them in testing and haven't seen one that has degraded yet. For most you can't tell the difference on a full backup (including extras) either. I have a BFD-25 option in BD Rebuilder and haven't seen a disc yet that really required it. In fact, for typical movies (two hours or so) you can do a movie-only backup to BD-5 with no detectable blocking. I'm telling you, I'm excited about this -- H.264 (via X264) is really efficient.

You may see a beta for Christmas... but don't expect it to be bug-free.

For me is a resize to BD-R25 important, since its more compatible to Standalone BluRay Players BD-9 is Blu-ray compliant... it isn't some made-up standard like mini-DVD.

[Vanderlow]

just so I'm sure we are talking apples to apples when you say BD-9 you mean BD movie shrunk on a 8.5 GB DVDr-9, right? and BD-5 would be a BD backed up .264 on 4.5GB DVDr-5? if that's so, I'm excited too ...

[rack04]

just so I'm sure we are talking apples to apples when you say BD-9 you mean BD movie shrunk on a 8.5 GB DVDr-9, right? and BD-5 would be a BD backed up .264 on 4.5GB DVDr-5? if that's so, I'm excited too ...

http://en.wikipedia.org/wiki/Blu-ray_Disc#BD9.2FBD5_Blu-ray_Disc

[writersblock29]

@Jdobbs

[Quoted] "In fact, for typical movies (two hours or so) you can do a movie-only backup to BD-5 with no detectable blocking."

Hmm... Last night I did a movie-only of Indiana Jones and the Kingdom of the Crystal Skull to DVD5, and there were blocks on pretty much every flat area of the image. The detail of moving objects was good, but the pixelation on inactive areas (the sky, a wall, the ground, for examples) was pure distraction. Doing this movie on DVD9 resulted in a flawless copy. Since I'm using Ripbot264 and not reducing the resolution (both attempts were 1080p), and--while I am keeping the 5.1 (in this case, reducing the HD True track to 448kbs AC3), I'm wondering: Are there any special settings you're using for AVC that Ripbot isn't? Given the unreliablity of dual-layer media these days (it's a headache keeping track of Singapore Verbatims--those India disks are really only good for art projects or trash can liners), I'd love to be able to use single-layer blanks.

[jdobbs]

Are you watching it in motion or looking at a "paused" frame? It's probably a stretch to say "BD-5 with no detectable blocking" (I should have been more careful with my choice of words). But it isn't noticable during normal viewing. For the most part the settings are the same as in the batch I posted earlier in this thread.

[~bT~]

bd5 with 1080p is asking for blocking.
bd9 should solve it.

[writersblock29]

@Jdobbs

This was with regular watching. I had to laugh, though; you reminded me of all the "this encoder is better than that encoder" arguments where it would appear as if some people were so busy nit-picking individual frames, they never had the chance to *cough* actually watch a movie. But back to the Indy Jones film, I decided to test this one because it's only a hair longer than two hours; I wanted to test the boundaries a little bit more. I had done the old 1980 Carpenter movie, The Thing, to DVD5... and it looked very good. What pixelation I did notice I caulked up to just how grainy the original source material is--not to mention that there's a lot of fire (I remember MPEG2 encoders didn't seem to agree with flame, fog, or water... I assumed AVC might be privy to some of the same if even to a lesser degree).

Both my television (50" 1080p hooked up via HDMI) and my computer monitor (1440 X 900 19" LCD via DVI) displayed the same behavior. Both of these projects I also committed to DVD9, and those are the ones I'm keeping; those looked absolutely indistinguishable from the original sources.

Now I'm eyeballing Gangs of New York. That sucker's close to three hours long. If I don't die of old age before the encoding's done, it'll be interesting to see how that looks. But I didn't want to commit to it before I knew if there's some other ST-friendly settings I could be using to maximize the efficientcy of X264. I trust your eyes even if I understand how subjective the issue of quality is, so I'd figured I'd ask about your DVD5 encoding settings.

[jdobbs]

Of course there's always the option of going 1280x720p for DVD-5 too. I haven't added that to BD Rebuilder yet, but I'm eventually going to.

[G_M_C]

[...]

Now I'm eyeballing Gangs of New York. That sucker's close to three hours long. If I don't die of old age before the encoding's done, it'll be interesting to see how that looks.

[...]


That's never gonna look good, whatever encoder or settings you use. Not even buying a complete SGI encoder-farm will help you.

Why you ask ? Cause it's rated as the "UGLIEST VIDEO TRANSFER" of 2008, and one of the worst quality transfers ever. ( http://www.highdefdigest.com/news/show/Worst_Blu-rays_of_2008/Kenneth_Brown/Little_Lumps_of_Coal:_The_Worst_Blu-rays_of_2008/2344)

[MadMonkey57]

I can share some of my experiences in this debate over H.264 / quality / BD5 / BD9 / 720p / 1080p stuff.

In addition to my own personal feelings about the overall quality of my encodings, I like the objectivity of metrics. If you are interested, I suggest you start here : MSU Video Quality Measure (http://compression.ru/video/quality_measure/index_en.html). Basically, you can use tools to compare an encoding with the original movie (and even compare 2 different encodings) and thus give you an objective point of view on your encoding.

I made something like 30 encodings and ran all of them through the tools. I have come to the following arbitrary conclusion on authoring BD5s :

- Movie length < 1h45mn AND 1 AC3 audio stream 640 kbits/s : i keep 1080p
- Otherwise, I downsample to 720p.

Please again consider that this is arbitrary based on a few tests and comparisons, there are other parameters that should be considered (i.e. animation movies, fast paced action movies, top/bottom black borders, TV size, ...), but i wanted a simple rule for deciding on downsampling or not.

So far, "I" am quite happy with what I authored and "I" like what I watch on my 52" Sony LCD. The important thing is that "you" enjoy your own BD5/9s.

[Fishman0919]

bd5 with 1080p is asking for blocking.
bd9 should solve it.

Well the 30 or so movies only I've reencoder at 1080p onto a BD-5 (single layer DVD) look indistinguishable from the org.

Edit: The only movie I was really forced to reencoder to BD-9 or resize to 720p so far was "The Dark Knight"

For anyone interested, here is the command line I use...
takes about 7-10 hours for me... depending on the movie.

--level 4.0 --sar 1:1 --aud --vbv-bufsize 18000 --vbv-maxrate 18000 --filter 0,0 --ref 3 --mixed-refs --bframes 3 --b-pyramid --keyint 24 --min-keyint 2 --direct auto --b-adapt 1 --subme 3 --trellis 2 --partitions all --8x8dct --me hex --merange 16 --no-fast-pskip --threads auto --thread-input --progress --no-psnr --no-ssim --output

--level 4.0 --sar 1:1 --aud --vbv-bufsize 18000 --vbv-maxrate 18000 --filter 0,0 --ref 3 --mixed-refs --bframes 3 --b-pyramid --keyint 24 --min-keyint 2 --weightb --direct auto --subme 7 --trellis 2 --partitions all --8x8dct --me umh --merange 24 --no-fast-pskip --threads auto --thread-input --progress --no-psnr --no-ssim --output

[MadMonkey57]

What do you mean exactly by "indistinguishable" (objectivity / subjectivity) ?
What size is your TV screen ?
What length are your movies ?
How many audio streams ? What bitrates ?

You see, I'm not saying that I don't believe you (on the contrary, I'm a 100% sure you fully enjoy your movies), I'm simply trying to outline that it depends on so many technical parameters and personal preference that it's almost impossible to find the golden rule... And by the way, I believe everyone has his/her own golden rule... I believe you have found yours...

[jdobbs]

Hmmm... to me it means "I can't tell the difference." Which has been my experience as well. If someone walked into the room and saw it playing they wouldn't be able to tell if it was the original or the copy playing. Does that mean it's perfect? Probably not -- but damn close. But of course the original isn't perfect either. If you stop most original BDs playing you will see blocking as well.

I use close to the same thinking you do in deciding which way to go. But I usually decide to go BD-9 if it is over about 2.25 - 2.5 hrs, except for high-action films. Those would probably start showing at about 2 hrs.

Lately though I've been doing most of mine as a full backups (with extras, menus etc.) and I put those all (with a few small source exceptions) on BD-9s.

[MadMonkey57]

Well to me it means the video quality measurement tools reach a high score comparing original and backup movies. It's just a matter of "personal preference"... Again, I'm not saying I am right and you are wrong. I'm just saying it's pretty hard to decide on going BD5/BD9 or 1080p/720p...

[Fishman0919]

What do you mean exactly by "indistinguishable" (objectivity / subjectivity) ?
What size is your TV screen ?
What length are your movies ?
How many audio streams ? What bitrates ?

You see, I'm not saying that I don't believe you (on the contrary, I'm a 100% sure you fully enjoy your movies), I'm simply trying to outline that it depends on so many technical parameters and personal preference that it's almost impossible to find the golden rule... And by the way, I believe everyone has his/her own golden rule... I believe you have found yours...

I mean... like jdobbs said... anyone watching the movie... including myself... can't tell the difference while watching it. Which is what matters to me.

Screen sizes are Sony KDL-52XBR2 52" and a Samsung HL61A750 61"

Over 30 different movies... all different lenghts.... my rule of thumb is pretty much the same as yours but I go with 2hr's plus... but not to say it still doesn't look good, just for me... got to cut off somewhere.

Just the one main audio... mostly 640k DD... sometimes with a longer movie.. I will convert the audio to 448k

[MadMonkey57]

Ok. Thanks for reporting back.

As I said before, I certainly respect your point of views (both yours and jdobbs'), it's just that I personally rely more on subjective metrics than on being able to tell a difference (probably because I am not good at such a task...). Anyway, it's good to know your experiences. They give me opportunities to refine my golden rule (and quit relying on maths all the time...:p)

[DVD Maniac]

You may see a beta for Christmas... but don't expect it to be bug-free.



I don't care about bugs, lived in London for much of my life and its full of nasty creatures, oh and the insects are a nightmare too.

I want a pressie to play with!
I want a pressie to play with!
I want a pressie to play with!
I want a pressie to play with!
I want a pressie to play with!

[jdobbs]

Well to me it means the video quality measurement tools reach a high score comparing original and backup movies. It's just a matter of "personal preference"... Again, I'm not saying I am right and you are wrong. I'm just saying it's pretty hard to decide on going BD5/BD9 or 1080p/720p...With that I disagree. I've been throught the entire suite of automated "quality measuring" tools and algorithms over the past few years -- and none of them are even close to right. In fact -- when measured against human perception they don't even seem to have any reliable correlation. I've seen examples where PSNR gives good scores to video that is so terrible that it is intuitively obvious to the casual observer. Eyeballs and double-blind tests are the only reliable measurements.

As you said earlier -- it's all opinion and perception.Just the one main audio... mostly 640k DD... sometimes with a longer movie.. I will convert the audio to 448k.I convert all my discs' audio to 448K if it is 6 channel and 192K if it is 2 channel. I can't hear the difference that 640K gives...

One other thing you have to watch too: On BD sometimes the subtitles are huge. I made a backup of "The Matrix" the other day -- and the combined subtitles for just the feature came to more than a gigabyte.

[moviefan]

If I may throw in my opinion on the topic of deciding whether to encode to BD5 or BD9, 1080p or 720, and how "good quality" is measured, I personally think that good quality with 2 AC3 tracks and 1080p is generally only achieved with BD9 encodes. Concerning the quality: I don't measure the quality on basis of a comparison between the original and an encode but on the subjective look of the encode, especially because my aim is not to be transparent but to improve the look for my taste. This means, I want to degrain a little and stabilize the picture because it really bothers me when there are dancing pixels in flat areas where there is no real movement at all... I like rather clean, stable and sharp picture and not this grainy stuff... Apart from the better look, in my opinion, degraining and stabilizing makes a video more compressible, so with movies shorter than 2h, a BD5 encode might look very good as well.

[Wombler]

Eyeballs and double-blind tests are the only reliable measurements.

That is most definitely the case and is the ultimate test in my mind too.

If your eye can't perceive any loss in quality then to all practical intents and purposes it's as good as the original.

Equally if one version looks better than the other to the eye it's pointless to say that the quality measurement software says otherwise.

One other thing you have to watch too: On BD sometimes the subtitles are huge. I made a backup of "The Matrix" the other day -- and the combined subtitles for just the feature came to more than a gigabyte.

Wow that's unbelievable. How come they're so phenomenally huge compared to DVD?


Wombler

[rendez2k]

I may have missed this somewhere.... I currently use RipBot264 for BR to BD9 which appears to work a treat (although it takes ages of course). What features will this have which are different or better?

[Sharc]

To me it seems that the biggest challenge for the x.264 encoder are fading flat scenes, like the blue sky at the beginning of "The bucket list" where I noticed color banding in my movie-only 720p backup with 2 AC3 tracks on a BD5

[Fishman0919]

I may have missed this somewhere.... I currently use RipBot264 for BR to BD9 which appears to work a treat (although it takes ages of course). What features will this have which are different or better?

It seems that jdobbs is making a BD Rebuilder... much the same as his DVD Rebuilder. It backs up the whole disc. Menu, main movie, extras... etc.

RipBot264, which is great, backs up only the main movie.

[Fishman0919]

To me it seems that the biggest challenge for the x.264 encoder are fading flat scenes, like the blue sky at the beginning of "The bucket list" where I noticed color banding in my movie-only 720p backup with 2 AC3 tracks on a BD5

aq-mode and aq-strength can help with that.

--aq-mode 2 --aq-strength 1.0

[Sharc]

aq-mode and aq-strength can help with that.

--aq-mode 2 --aq-strength 1.0

Hmmm...., according to x264 --longhelp the --aq-mode 2 does no longer exist, only 0 or 1 are possible. Default is 1 (=Variance AQ)

Isn't --aq-mode 1 --aq-strength 1.0 set by default since a while? I thought that it needs no longer be explicitly specified, unless for values other than the default 1.0. Am I wrong here?

[MadMonkey57]

With that I disagree. I've been throught the entire suite of automated "quality measuring" tools and algorithms over the past few years -- and none of them are even close to right. ... Eyeballs and double-blind tests are the only reliable measurements...

...Concerning the quality: I don't measure the quality on basis of a comparison between the original and an encode but on the subjective look of the encode...

...If your eye can't perceive any loss in quality then to all practical intents and purposes it's as good as the original... Equally if one version looks better than the other to the eye it's pointless to say that the quality measurement software says otherwise...

All right, i get the message...:)
And I understand, I really do.
Nevertheless, the latest techniques and tools in terms of quality measurement deserve some consideration IMO...

Anyway, the whole point in disagreeing (in forums) is to give us a chance to widen our "field of vision" and thus to improve. I'm glad I improved.

[jdobbs]

Well I posted BD Rebuilder v0.17.9 (beta) for a few minutes this morning -- then realized I'd changed something overnight that made it completely stop working. So if anyone downloaded it -- don't waste your time, its broken. I'll post again after I make some changes.

[Update] Well... it looks like it may have been just a bad disc that got me all spun up. I'm still testing, but will post again if all looks well.