Still getting stutter

[osho]

I just encoded SWAT.

The rip is absolutely amazing on PC.

But playing on my Liteon DVD player I get choppiness all over main movie.

I konw it was posted already but it came to conclusion that it has been resolved.

I used latest DVD=RB and Quenc 0.51.

if the question has been generally answered please forgive me .

1, Is there a to fix the rip. running through some other app?

2. what do to so i can get rip playing correctly on standalone players?
I used rejig and that thing worked well only quenc looks to have problems.

3. dont know why but DVDRB creates 200MB smaller rips as it suppose to be.

any help is welcome thx a lot.

[writersblock29]

@osho

Well, let's see... I'll try and help you as much as I can (and what I goof up on, the many very qualified people on this forum can definately catch and set straight for you; particularly the DVDRB's author, Jdobbs).

But I need some info from you! Specifically, what did you initially use to rip the DVD (ripping being defined as bypassing encryption and placing the original files on your hard drive)? If it was DVD Decrypter, was it the most current version?

What version of DVD Rebuilder are you using? .46 (the most current at the time of this writing)? What options did you select in Rebuilder -- Half D1 for extras... Deinterlacing with Decomb (which, perhaps, Rebuilder tried to use on the main movie, which is already progressive -- IF THAT'S TRUE, there's a bug for Jdobbs to check into). Are you using any other utilities to "tweak" your output from DVD Rebuilder? "One Click" enabled, or did you do the three step process? (What I'm looking for is anything you may have done above and beyond what DVD Rebuilder will automatically do.) While encoding, did you recieve any errors that you had to bypass?

What was your burning application (and version number)? Media? Writing speed? And what's the brand of your burner?

Give us as many details as you can: We don't bore easily! Plus, it'll help us narrow it down for you and possibly set you on the right path. If not, then it may give insight to a bug that can be fixed for version .47.

[Joergen]

I think the magic word there is the Lite-on Player (I didnt even know they made players). It's again a rebranded noname OEM player similar to others that have stutter problems (Apex, Cyberhome, Sampo etc).

I bet theres not much to do except try CCE and wait for jdobbs to find a stuttering player.

Also, how many segments did dvd-rb do for SWAT? Although the segment stutter bug also shows on PowerDVD and WinDVD and is not player specific but some sort of encoding fault.

[osho]

Wow i nearly thought noone would reply,

ok here it goes,

I used DVD Decrypter 3.2.2.0
Quenc 0.51
DVD-RB 0.46
and Mpeg2dec...dll from dvd2avidg100.zip
Avisynth 2.5

DVD-RB settings
Quenc Mode
3 click mode
Dynamiclly assign cell bitrates
half D1 and Half space for extras

AVS options
ConvertToYUV2()
audiodub(BlankClip())
NO deinterlance
Quenc option
I did 2 movies
one with Trellis ON(SWAT)
one with Trells OFF(MASTER AND COMMANDER)

No other tweaks were done during whole process
No error occured during encoding

Burning APP newest NERO 6.3.1.6
MEDIA MEMOREX 4x tried both + and -
Burner Pioneer 107D

Movies Master And Commander: The Far Side Of The World (Widescreen)
S.W.A.T. (Widescreen)

BTW my standalone player is LITEON LVD-2002

You know I would really appriciate any help cos the quality is amazing.

I have used DVD-RB on 10 previous movies but using REJIG no problems at all.

thx again

[osho]

Originally posted by Joergen
I think the magic word there is the Lite-on Player (I didnt even know they made players). It's again a rebranded noname OEM player similar to others that have stutter problems (Apex, Cyberhome, Sampo etc).

I bet theres not much to do except try CCE and wait for jdobbs to find a stuttering player.

Also, how many segments did dvd-rb do for SWAT? Although the segment stutter bug also shows on PowerDVD and WinDVD and is not player specific but some sort of encoding fault.

I will look into that but is is not rebranded noname OEM this is Liteon build player. and at the time I bought it, it was 190 dollars.

[Joergen]

But Lite-On is hardly a home theater manufacturer and most probably has not acquired their mpeg2 decoding chips from Sony, NEC, Matsushita (Panasonic), Philips etc but from the same manufacturer as the OTHER comparable brands of non-substantial stature in the (home entertainment) industry.

AFAIK Lite-On manufactures only optical drive mechanics and software (plastic cogs and firmware), they dont even produce the lasers that read and burn discs in their renowned PC hardware.

But, at least there's another brand jdobbs can add to his list of suspect player types. And I like Lite-On, I recommend them to all my friends for cd-burning and dvd-rom drives.

[osho]

all you said is true.

I will test it on Sony when my neighboor comes home.

the thing is rejig with dvd-rb produces correct output. it is Quenc mode

[insanescape]

Sure, you paid 190, but you basically paid a premium for Divx, not superior DVD quality.

[AzraelsCurse]

I still get the same stutter as in the old posts. See the old stutter posts for more info. I don't think it's been resolved yet for anyone still getting stutter after version .40. .40 is the last version that had any affect on stutter, but .40 didn't put a complete stop to it. I think it's largely related to specific DVD players, however I don't consider Sony DVD players noname players. I'm not saying they're the best of the best either, I'm just saying that it's not just his Lite-On with the problem. I've even seen stutter using DVD-RB backed-up DVD's in my Apex 1500, and I have created some way out of spec DVD's that have played perfectly in that thing. I see no stutter using dvd shrink or IC8, but again, more info in the old posts.

[KungFuCow]

I have an LVD-2001. Plays fine on my player.

[Paced]

Originally posted by osho

I have used DVD-RB on 10 previous movies but using REJIG no problems at all.

thx again

I have read in the QuEnc thread (in the Other MPEG 1/2 Encoder forum) that QuEnc has the tendency to go over the maximum/safe bitrate (9800 kbps) that most standalone DVD players can handle - this may be the reason why you're getting the stutter all over the main movie. Here are some suggestions:

1. Before you use DVD-RB, open QuEnc manually, go to the Advanced Options, and set the Max Bitrate to 9000 kbps (and untick 'Auto Max Bitrate' if you haven't done so already). Close QuEnc, then start the DVD-RB process.

or

2. Give CCE a try :)

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

Slightly off the thread topic here, but does anyone know if DVD-RB automatically sets the maximum bitrate in QuEnc to 9000 kbps (like it does in CCE) ? I haven't had the chance to test it yet.

//Update

Looks like DVD-RB doesn't set the maximum bitrate - before running DVD-RB, I opened QuEnc, put 10000 kbps in the Max Bitrate option, closed it, then ran DVD-RB again. 10000 kbps was still set as the Max Bitrate.

[writersblock29]

@Osho

Don't know if there was a bug-fix in this, but there's a newer version of Rebuilder out now (.47). I'd bet that Paced is on the right track, though, with his suggestion; it sounds like you've got most everything configured correctly -- and if Rejig gave you no problems...

As amuzing as the "Ford VS Chevy" debate is (in this case, which brand name sells better DVD players/and/or/Recorders)... if your player's not messing up the playback of any other copies you've made with other software, it ain't your player. Set tops do have a life-span on them -- and when they go down, they go down quickly -- but the chances that your player decided to crash the moment you were playing one of your Rebuilder copies? A little too convenient for my taste!

You might try using Rejig to copy one of the titles that was giving you grief with QuEnc, burn that to a Rewritable, and see how it behaves. (Assuming that Paced's suggestion didn't work, of course!) Since Jdobb's primary concern was to make an easy one-clicker for CCE users, I'd bet that there may just be an overlooked command between Rebuilder and QuEnc at this point. But if you can tweak it within QuEnc, gravy! If the suggestion works, post 'er! Then Jdobbs has a bit of good info for the future, while everyone else with the same problem has a work-around.

Good luck!

[Fr4nz]

Originally posted by osho
I just encoded SWAT.

The rip is absolutely amazing on PC.

But playing on my Liteon DVD player I get choppiness all over main movie.

I konw it was posted already but it came to conclusion that it has been resolved.

I used latest DVD=RB and Quenc 0.51.

if the question has been generally answered please forgive me .

1, Is there a to fix the rip. running through some other app?

2. what do to so i can get rip playing correctly on standalone players?
I used rejig and that thing worked well only quenc looks to have problems.

3. dont know why but DVDRB creates 200MB smaller rips as it suppose to be.

any help is welcome thx a lot.

Maybe bad media quality? Try with a better DVD media brand :)

[Sir Didymus]

Originally posted by Paced
I have read in the QuEnc thread (in the Other MPEG 1/2 Encoder forum) that QuEnc has the tendency to go over the maximum/safe bitrate (9800 kbps) that most standalone DVD players can handle - this may be the reason why you're getting the stutter all over the main movie. Here are some suggestions:

1. Before you use DVD-RB, open QuEnc manually, go to the Advanced Options, and set the Max Bitrate to 9000 kbps (and untick 'Auto Max Bitrate' if you haven't done so already). Close QuEnc, then start the DVD-RB process.

or

2. Give CCE a try :)

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

Slightly off the thread topic here, but does anyone know if DVD-RB automatically sets the maximum bitrate in QuEnc to 9000 kbps (like it does in CCE) ? I haven't had the chance to test it yet.

//Update

Looks like DVD-RB doesn't set the maximum bitrate - before running DVD-RB, I opened QuEnc, put 10000 kbps in the Max Bitrate option, closed it, then ran DVD-RB again. 10000 kbps was still set as the Max Bitrate.

@Paced: Great!!!

I totally agree on the quoted post...
It is also relevant the suggestion of limiting the maximum encoding bitrate. It is to confirm, but if it works it should be included by sure in the to-do list of improvements for DVD-RB...

@Osho: Please follow what Paced suggested and let us know if it works...

I have to say that my encodes are mostly based on CCE...
I had relevant stutter troubles up to release 0.40...
All these problems vanished for me with the newer releases of DVD-RB...
So I was quite surprised to see that some people has still having stuttering...
The hypothesis of Paced seems to me it could give a good explaination for these problems...

Cheers,
SD

[Sir Didymus]

@Fr4nz:

IMHO it is hardly a problem of poor media; Osho sayd that using Rejig he had no problems at all...

[osho]

Thank you all for all wise advises.

I believe that it is not media. I believe that memorex is pretty good for DVD-R and +R they definately were not the cheapest ones. burned around 250 with various server backups and movies.

As some suggested that i give it a go with cce well that is what i am doing. currently at 25% encoding phase.

The reason why I was using Quenc was I like to have my soft free. so we shall see what will come up from CCE.

if there still be problem i will try quenc with limited max at 9000kb although I believe it might have no impact since none of the movies in original format reach that high.

I will let you know in morning, it is 3AM here.

Thx all again

[nawo69]

I'v got the same problem but only with one DVD and it only affects my PS2. (I have only done 3 disc's so far)

And i did the same Disc using the Big3 and it plays fine.

I have tried .45 to .47 with CCE and Rejig and it only plays the first 10 seconds of each chapter.


I used DVD Decrypter 3.2.2.0
DVD-RB 0.45 - 0.47
and Mpeg2dec...dll from dvd2avidg100.zip
Avisynth 2.5

DVD-RB settings
CCE SP 2.67 and also tried Rejig
1 click mode
Dynamiclly assign cell bitrates

AVS options
ConvertToYUV2()
audiodub(BlankClip())

Image created with Imgtool and burned with DVD Decryptor

Works fine on PC and my Encore DVD player but it only plays the first 10 seconds of each chapter on my Playstation 2

[osho]

Ok here I am with update.

I have reencoded stuff with CCE.

No stutter at all. on PC and on standalone player.

But I think that Quenc encodes are nicer, more vivid. therefore I would love to keep using it.

the previous quenc encodes have had problems on brand new sony player. therefore I believe it must be a bug in the DVD-RB. even thou the stutter was not so frequent as on my liteon.

[Paced]

Originally posted by osho
Ok here I am with update.

I have reencoded stuff with CCE.

No stutter at all. on PC and on standalone player.

But I think that Quenc encodes are nicer, more vivid. therefore I would love to keep using it.

the previous quenc encodes have had problems on brand new sony player. therefore I believe it must be a bug in the DVD-RB. even thou the stutter was not so frequent as on my liteon.

You still can use QuEnc, if CCE gives you no stutters, then it's obvious (I think) that QuEnc was going over the safe maximum bitrate (9800kbps); when you have the time, give what I said above a try (set QuEnc's Max Bitrate to 9000 kbps before you start DVD-RB).

[writersblock29]

@Osho

I figured you might find this interesting!

Version 0.48

Below are a summary of the changes/updates associated with this version:

- Corrected an error related to fast forwarding and rewinding through chapter points. The first entry in any Cell's first DSI was not correctly flagging time offsets that had no prior VOBU. My thanks to Sir Didymus for finding and pointing out this error. Extraordinary beta testing!

- Added support for the "-maxbitrate" option to all QUEnc encode command lines. The value is set to the DVD-RB default of "9000" unless changed via the "max_bitrate=" INI option. This might fix some of the stuttering errors seen when using QuEnc.

- Enabled checking that disallows mismatching of modes between the PREPARE and ENCODE phases.

- Fixed an error in which "IDCT7" was not being properly applied when selected. Thanks to djan for discovering this error.

http://forum.doom9.org/attachment.php?s=&postid=491235

(I'm not sure if the link will work... but it's availible in this forum! Sounds like you were heard! Thanks, Jdobbs!)

[JDay]

The new version of DVD-RB does not fix this. This might be a bug in the way DVD-RB handles the pulldown for progressive 23.976fps sources. The original m2vs work fine with both CCE and QuEnc. Is it possible that DVD-RB handles the pulldown differently for CCE and QuEnc? Pulldown works fine for CCE and Quenc (interlaced sources) or when I do the pulldown manually with pulldown.exe on progressive files from QuEnc.

Edit: I did a short clip with CCE and it came out fine. I then replaced the m2v with one encoded with QuEnc and clicked rebuild (all in CCE mode). The output had the stutter. Looks like a problem with QuEnc, rather than DVD-RB.

Update: The above test was with the second cell only, which was entirely progressive. I tried it with the first cell, which is interlaced during the logos and credits, progressive after that (I encoded as interlaced, I think RB flagged it as progressive during rebuild, not exactly sure how that works). That whole cell turned out fine. I'll try running the whole movie through next.

[jdobbs]

When DVD-RB completes it's rebuild every individual frame's TFF, RFF, and progressive flags are set exactly as they were in the original.

[RB]

Originally posted by jdobbs
When DVD-RB completes it's rebuild every individual frame's TFF, RFF, and progressive flags are set exactly as they were in the original.
jdobbs, are you really modifying the per-frame PROGRESSIVE_FRAME flag during rebuild? I don't think you can do that as it at least tells the player how chroma is stored (there is a difference between progressive and interlaced YV12).

[JDay]

I posted this in the "Other MPEG1/2" forum, but I'm not sure if this is a problem with QuEnc or DVD-RB, so I'm posting it here, as well.
I tested the same sequence with QuEnc and CCE. The m2vs, of course, played fine. I then ran them through pulldown and both came out fine when only "mark each frame as progressive" was active. If "mark entire sequence as progressive" was activated, both "stuttered". I'm pretty sure this is a QuEnc issue, rather than a rebuilder issue since even when I "sneak" Quenc segments into CCE projects or CCE clips into QuEnc encodes, the QuEnc segments "stutter" and the CCE clips do not. Is is possible that QuEnc is putting info in the flags or the headers on its progressive streams (different than CCE) that might be causing this?
Also:
Hmm... I've found a little hack to "fix" this issue. You can use pulldown.exe to set the prog_seq flags to interlaced, not changing anything else (-nopulldown -framerate 23.976 -prog_seq i). So, it looks like it's related to the prog_seq flag. DVD-RB's pulldown works fine on material encoded as interlaced, and on progressive material, if its flagged as interlaced. Is this a bug with QuEnc or DVD-RB?

[jdobbs]

Originally posted by RB
jdobbs, are you really modifying the per-frame PROGRESSIVE_FRAME flag during rebuild? I don't think you can do that as it at least tells the player how chroma is stored (there is a difference between progressive and interlaced YV12). Actually the decision between interlaced and progressive is made before encode. But it matches the original.

[DVD Maniac]

Until recently I have been fortunate enough not to experience the stutter problems. The one thing that was happening was on some episode DVD's where the chapter points were not working properly - a problem I believe has been resolved with v48.

Now I just tried Kill Bill Vol1. Because the main movie VTS is a relatively small % reduction required (I deleted most of the crappy extras), I decide to keep the DTS sound track. The resulting output played fine with my pc player, but on both my set-tops I got choppy playback. Wierd thing was that the louder the sound at the point in the scene - the more choppy it got.

I have now encoded just the .ac3 stream with Shrink and all is well. I will run a check with Rebuilder tonight, this time keeping ac3. I suspect this will work also.

Could it be that Rebuilder has problems with DTS sound tracks? Anyone had problems with DTS and Rebuilder?

[jdobbs]

Originally posted by DVD Maniac
Until recently I have been fortunate enough not to experience the stutter problems. The one thing that was happening was on some episode DVD's where the chapter points were not working properly - a problem I believe has been resolved with v48.

Now I just tried Kill Bill Vol1. Because the main movie VTS is a relatively small % reduction required (I deleted most of the crappy extras), I decide to keep the DTS sound track. The resulting output played fine with my pc player, but on both my set-tops I got choppy playback. Wierd thing was that the louder the sound at the point in the scene - the more choppy it got.

I have now encoded just the .ac3 stream with Shrink and all is well. I will run a check with Rebuilder tonight, this time keeping ac3. I suspect this will work also.

Could it be that Rebuilder has problems with DTS sound tracks? Anyone had problems with DTS and Rebuilder? Not likely. As far as Rebuilder is concerned all audio tracks are treated equally -- and they are not modified (except clock references).

Recommendation -- check your maximum rate. DTS can be very large compared to AC3. Its possible your combined video and audio are exceeding 9,800Kbs

[writersblock29]

@DVD Maniac

You may still have that problem. I backed up KILL BILL using .47 Rebuilder, but had stuttering in key points: The moment Uma awakens from her coma, for example -- or the coffee table crash at the beginning fight scene. I use CCE as my encoder, and had made copies both with "half D1" and without (written to rewritables), but wound up doing a "movie only" copy using shrink -- which required no compression. I kinda shrugged, reminding myself that I own the original disk; those special features aren't going anywhere if I want to watch them later. *I suppose I should mention that in all cases, I only kept the Dolby Surround track rather than the DTS.*

NOW, I also want to say that I haven't run into any other movies that have displayed this behavior with Rebuilder -- and I've use RB for a few projects so far. I'm half tempted to try KILL BILL again, and check that bitrate setting during the stuttering scenes. Who knows?

[TheBigDave]

I'm still having stuttering problems also. I'm using the latest DVDRB with CCE 2.50. It doesn't happen on all discs. But it's been most noticable on Asian DVDs. The stutter is almost like a layer-break pause or an extra 2-3 still frames inserted into the movie. Strangely it doesn't seem to effect the audio and the movie stays in sync.

I haven't been able to figure out a pattern to the stuttering. Sometimes it happens just once or twice a movie. Another time it happened every 3-4 minutes. And on my latest back-up it happened every 2-3 seconds, but just during a few scenes.

The stuttering only seems to occur on my Toshiba SD-2805. It doesn't happen on my Toshiba SD-V291 or my Liteon lvd-2001. I use Ritek G04 discs which I have never had problems with on these players.

I've tried re-doing it and the stuttering occurs in the exact same places. If I use the Big 3 method it doesn't happen.

I don't have any problems with stuff like time codes, chapters or FF/RW. Other than the stuttering, it turns out perfect.

Anyways, I just wanted to bring this to your attention. If there's any other details I can give you, let me know. Thanks for all your great work.

[TheBigDave]

Originally posted by TheBigDave
And on my latest back-up it happened every 2-3 seconds, but just during a few scenes.

I just ran this DVD through Bitrate Viewer. Here's a picture of BV's reading during a 30-second scene, which had stutters every 2-3 seconds. As you can see the bitrate quickly jumped up very high. I'm guessing this is what caused it?

http://images.andale.com/f2/115/106/1203959/1083462694925_DVDRB_Stutter.jpg

Here's a picture of Bitrate Viewer from the same scene on the original DVD which doesn't stutter.

http://images.andale.com/f2/115/106/1203959/1082803715761_DVDRB_NoStutter.jpg

The bitrate for that scene appears to be higher on the DVDRB DVD. I thought the maximum bitrate was 9800? But Bitrate Viewer shows that it's 10861 on the DVDRB DVD for that scene. Is that too high?

UPDATE

Here's a Bitrate Viewer picture from another disc that had some stuttering. This is just a short 1 second stutter. As you can see the bitrate at that moment is again very high (10609).

http://images.andale.com/f2/115/106/1203959/1082877805298_DVDRB_Stutter2.jpg

[jdobbs]

@TheBigDave

What is your setting for VBR_Bias? Has it changed from the default? If it is set too low it can create some pretty significant bitrate swings. If that isn't it you might want to try lowering the maximum bitrate to 8500 (instead of default value of 9000, it is referenced in the Rebuilder.txt file).

[TheBigDave]

Originally posted by jdobbs
What is your setting for VBR_Bias? Has it changed from the default? If it is set too low it can create some pretty significant bitrate swings. If that isn't it you might want to try lowering the maximum bitrate to 8500 (instead of default value of 9000, it is referenced in the Rebuilder.txt file).

Thanks for the quick response. I have VBR_Bias at 25 (and Quality_Prec at 16). I think I read in this forum that these were the recommended settings. Do you think it would help if I raised it?

To lower the maximum bitrate do I need to open the Rebuilder.ECL file and change it for each cell? Is there a way to do it automatically?

If the maximum bitrate is set to 9000, do you know why I'm getting readings of 10600+? Is that normal?

Thanks for the help.

[DDogg]

TheBigDave, try pasting/replacing the values below into you ini file. They are way safe values. If it stops the stutter you will know what it was.

[Options]
min_bitrate=300
max_bitrate=8000

[CCEOptions]
VBR_bias=33

[TheBigDave]

Originally posted by DDogg
TheBigDave, try pasting/replacing the values below into you ini file. They are way safe values. If it stops the stutter you will know what it was.

Thanks for the input. I'll try that out later tonight. I'll post tomorrow to let you know if it works.

[jdobbs]

Originally posted by TheBigDave
Thanks for the quick response. I have VBR_Bias at 25 (and Quality_Prec at 16). I think I read in this forum that these were the recommended settings. Do you think it would help if I raised it?

To lower the maximum bitrate do I need to open the Rebuilder.ECL file and change it for each cell? Is there a way to do it automatically?

If the maximum bitrate is set to 9000, do you know why I'm getting readings of 10600+? Is that normal?

Thanks for the help. No it's not normal. But there are circumstances in which the encoder will momentarily go higher than it should. You also have to add in all the audio bitrates when figuring the high end. Normally 9,000 is safe and gives you the best high-end -- but you seem to have hit a couple of movies that have some anomaly that causes the spike. I'd recommend you try DDogg's suggestion and see how it works.

[Trahald]

ive wondered something (it may be something you already accounted for.. if so.. ignore this :) )

this would apply only to ntsc - On hybrid material.. you encode at 23.976 fps.. you output all real frames. rff's are added (amoung other things) to reach '29.97fps' in the film areas (which is fine.. thats what currently is done to film) the video areas just have the tff/interlace flag fixed to match origional.. the video is stamped 29.97 fps.
for the film areas the encoded bitrate is fine since the rffs only appear to make it seam 29.97 fps. but the video parts actually will have to be read faster to keep up since its truely 29.97 fps. this would cause a spike in bitrate.

this would also apply to any non hybrid video(29.97) if its encoded at 23.976 fps and the fps setting is changed after

[DVD Maniac]

You may still have that problem. I backed up KILL BILL using .47 Rebuilder, but had stuttering in key points: The moment Uma awakens from her coma, for example -- or the coffee table crash at the beginning fight scene

writersblock29 - Thanks for the confirmation, I am getting the stutter at exactly the same points and so it must be a bit rate problem. I will try a test with Ddogg's suggestion -



[Options]
min_bitrate=300
max_bitrate=8000

[CCEOptions]
VBR_bias=33

Not sure if this will work with DTS though due to its higher bit rates?

[jdobbs]

Originally posted by Trahald
ive wondered something (it may be something you already accounted for.. if so.. ignore this :) )

this would apply only to ntsc - On hybrid material.. you encode at 23.976 fps.. you output all real frames. rff's are added (amoung other things) to reach '29.97fps' in the film areas (which is fine.. thats what currently is done to film) the video areas just have the tff/interlace flag fixed to match origional.. the video is stamped 29.97 fps.
for the film areas the encoded bitrate is fine since the rffs only appear to make it seam 29.97 fps. but the video parts actually will have to be read faster to keep up since its truely 29.97 fps. this would cause a spike in bitrate.

this would also apply to any non hybrid video(29.97) if its encoded at 23.976 fps and the fps setting is changed after Unfortunately, if you look at TheBigDave's example, he is seeing this bitrate surge on a progressive source -- so at least in this case it would appear not to be the problem. I did give some thought to this when I was developing this method -- but now that you bring it up I think I may need to go back and review it.

[wmansir]

@Trahald

I'm not sure if DVD-RB accounts for Video, Film, and Hybrid states when determining bitrate, but if not I think of it as a feature. I think it does with Dynamic assign, just because of the nature of that method. For fixed/equal bitrate distribution, of course it doesn't, since each job gets the same bitrate.

I'm not sure I follow your description, but I think you may have it backwards. The VIDEO is getting a 25% bitrate increase, not the Film parts.

The Film parts are keept at ~24fps, so there bitrate is the 'true bitrate'. But when encoding the Video parts are slowed down to 24fps. That means 1 second of Video (30 frames) is counted as 1.25 seconds when it is run at 24fps (as it is fed to the encoder). So if your using a bitrate of 1000kbps that 1 second of video will get 1250kb, instead of 1000kb.

Of course CCE is pretty bad (IMO) at encoding interlaced material, so I think a 25% bitrate increase should help keep the quality balanced, if not the actual bits.

[jdobbs]

You also have to factor in the fact that the demand for bitrate goes down during encode -- because the same 1 second of complexity is spread over a 1.25 second window.

[jdobbs]

Originally posted by Trahald
ive wondered something (it may be something you already accounted for.. if so.. ignore this :) )

this would apply only to ntsc - On hybrid material.. you encode at 23.976 fps.. you output all real frames. rff's are added (amoung other things) to reach '29.97fps' in the film areas (which is fine.. thats what currently is done to film) the video areas just have the tff/interlace flag fixed to match origional.. the video is stamped 29.97 fps.
for the film areas the encoded bitrate is fine since the rffs only appear to make it seam 29.97 fps. but the video parts actually will have to be read faster to keep up since its truely 29.97 fps. this would cause a spike in bitrate.

this would also apply to any non hybrid video(29.97) if its encoded at 23.976 fps and the fps setting is changed after After checking and doing a few tests I think you have identified a valid error (although probably not related to TheBigDave's specific problem). When interlaced sources were being processed the maximum vbr bitrate should have been decreased to compensate for the 1.25 multiplier. I've fixed it and it will be a part of version 0.50.

I think QuEnc may have been undeservedly getting the blame for some of this when it was being used also. Shame on me.

Thanks.

[Trahald]

Originally posted by wmansir
@Trahald
I'm not sure I follow your description, but I think you may have it backwards. The VIDEO is getting a 25% bitrate increase, not the Film parts.

The Film parts are keept at ~24fps, so there bitrate is the 'true bitrate'. But when encoding the Video parts are slowed down to 24fps. That means 1 second of Video (30 frames) is counted as 1.25 seconds when it is run at 24fps (as it is fed to the encoder). So if your using a bitrate of 1000kbps that 1 second of video will get 1250kb, instead of 1000kb.

Of course CCE is pretty bad (IMO) at encoding interlaced material, so I think a 25% bitrate increase should help keep the quality balanced, if not the actual bits.

actually that was exactly what i was saying. 24 for film areas is fine. pulldown does essentially what jdobbs does.. just a different way. that has never been an issue

but the max of 9000 will actually be 11250 during VIDEO parts since cce has bitrate set for 24 frames during a second going into the decoder, but now 30 real frames must be read during the same second.

to address the issue thebigdave brought up.. bitrate viewer has always screwed up bitrate calc on pulldown material.. i think its internal calculator counts for 30 fps. (which would be wrong for film) This happens using pulldown.exe as well. you would have to test the m2v (pre-pulldown) file.. not the VTS_XX_X.vob file

[Trahald]

Originally posted by jdobbs
After checking and doing a few tests I think you have identified a valid error (although probably not related to TheBigDave's specific problem). When interlaced sources were being processed the maximum vbr bitrate should have been decreased to compensate for the 1.25 multiplier. I've fixed it and it will be a part of version 0.50.

I think QuEnc may have been undeservedly getting the blame for some of this when it was being used also. Shame on me.

Thanks.

i type too slow.. hehe.. ok cool.

[TheBigDave]

Originally posted by DDogg
[Options]
min_bitrate=300
max_bitrate=8000
[CCEOptions]
VBR_bias=33

I tried this with DVDRB v0.49 and it worked great! Here's a Bitrate Viewer pic of the stuttering scene which no longer stutters:

http://images.andale.com/f2/115/106/1203959/1083000936634_DVDRB_FixedStutter.jpg

Someone mentioned earlier in this thread that we should look at BV pictures of the m2v files. So here's a BV pic of the original unaltered m2v file:

http://images.andale.com/f2/115/106/1203959/1083712038564_DVDRB_Originalm2v.jpg

Here's a pic of the stuttering m2v file:

http://images.andale.com/f2/115/106/1203959/1086489073158_DVDRB_Stutterm2v.jpg

And here's the non-stuttering m2v file:
http://images.andale.com/f2/115/106/1203959/1086163199779_DVDRB_NoStutterm2v.jpg

If I can provide any additional info, please let me know. Thanks for all the help.

[DDogg]

TheBigDave, I'm pleased that helped you. You taking the time to do a great post with those bitrate graphs is what really helped us better understand the problem.

@Jdobbs, personal note - Quack, Quack :D

[Joergen]

This bitrate stutter is limited to NTSC, or are there possibly other scenarios where a PAL encoding could go beyond the 9000kbit?

Not sure why the "hundreds of segments" movies stutter (had only one though). I havent had any stutter problems with normal 30-50 segment movies.

[KungFuCow]

Could be related to how the player handles out of range bit rates. Ive noticed on my Sony player that sometimes I get a split second glitch sometimes on playback. Could be the same thing, just this particular player may recover more gracefully than some others.

[jdobbs]

Originally posted by DDogg
TheBigDave, I'm pleased that helped you. You taking the time to do a great post with those bitrate graphs is what really helped us better understand the problem.

@Jdobbs, personal note - Quack, Quack :D :confused:

[DDogg]

jdobbs, I was referring to when you and I had a friendly 'discussion' about whether bitrate peaks could be contributing to 'stuttering'. I had seen the same thing with the intro to Matrix Revolutions which had some very difficult encoding scenes that caused a near instantaneous bitrate surge. You disagreed I think that dropping the MAX might help. I then suggested to you if something walked like a duck, and quacked like a duck, it probably was a duck, thus my statement above. I very clearly hear a duck :) Just kidding you, and it was meant in the friendliest way.

[jdobbs]

I think the discussion was along the lines of whether 9000 was too high and that was why the bitrate went over... and as I said then and still say the answer is "No." I don't see how it is related to this discussion -- as Trahald made the valid point not that 9000 is too high (it isn't) but that the method of converting can result in a value that was over 9000.

If you had made that same point I would have agreed with you then as well. So I guess what I'm saying is the if it moos like a cow -- and walks like a horse, then you can't come to the conclusion that it is a duck, no matter how much you wish it was. :) This is also meant in the friendliest way.