Still getting stutter

[jdobbs]

I'll take a look at it, but I can't imagine why DTS would cause a stutter. It is left intact from the original. Have you looked at the IFO file? Do these stutters all occur at points that show a discontinuity? (Marked as "Layer Break" by IFOEDIT)

[jdobbs]

We need to keep the subjects straight. Going back through this thread I'm seeing comments about "stutter" -- some of which aren't stutter at all, others that are speaking of audio dropouts and some that are actually discussing video stutter.

Let's define "stutter" for this discussion as meaning picture jumps... not blockiness, pixelations, or anything else... but a point at which it appears to skip or lack smooth movement (like there is 1 or more frames missing in the stream).

Let's also define "audio dropout" as a point at which the sound appears to have gone silent for a short period only to return.

I'd also like to see comments as to how often you see it (every disc you do, rarely, etc.) -- the reason I need to know this is because the vast majority of DVD-RB users are seeing no stutter at all -- so there is something unique in the methods being used or the equipment. I agree I have to fix it, but until I can get my arms around it I'm not sure I will ever be able to do it.

I can tell you that I have done dozens of discs, and the only stutter I have ever seen was related to the layer break.

One think I definitely have to check is the DTS variable. I never keep DTS because I think it is a waste of space -- I will do a couple and see if I start to see any stutter.

[DVD Maniac]

Hi jdobbs and thanks for taking the time to look at this. To help you with isolating this problem here are my answers to your questions / comments -

Have you looked at the IFO file? Do these stutters all occur at points that show a discontinuity? (Marked as "Layer Break" by IFOEDIT)

No, they are frequent and do not cooincide with any Layer break cells / vob points

Let's define "stutter" for this discussion as meaning picture jumps... not blockiness, pixelations, or anything else... but a point at which it appears to skip or lack smooth movement (like there is 1 or more frames missing in the stream).

Let's also define "audio dropout" as a point at which the sound appears to have gone silent for a short period only to return.


You are absolutely right, I am seeing BOTH problem types. Taking my previous examples, the first one was definitely a case of Audio Dropout as you have defined it (no video skipping). Its occuring seemingly randomly and does not coincide consistently with bit rate spikes.

My second example was definitely stutter, with skipped video frames and coincides more directly to bit rate spikes (but as I have said, well below 8000). My Kill Bill example was awful, with stuttering every 10 seconds or so. A DD5 encode works just fine.


One think I definitely have to check is the DTS variable. I never keep DTS because I think it is a waste of space -- I will do a couple and see if I start to see any stutter.

I am pretty sure with the results I have seen so far (6 DTS attempts) that its a DTS issue that does not occur with DD5. I know a lot of forum members share your view that DTS is not worth keeping because of the size. Unfortunately I am a sound junkie and can DEFINITELY hear the difference even with my modest setup:)

Let me know if you want me to send you any more details on my attempted DTS backups so you can hopefully trace the problem

[brikin]

JDobbs programs come a long way....good job.

I am getting an audio stutter with 0.50a. The DVDs are Kill Bill and X2 with just the DD5.1 soundtracks. This audio stutter only happens on my Panasonic DVD player. It plays fine in my Sony.

Need anything else let me know.

Thanks

[brikin]

I had one more thing to mention about this audio stutter. I only get it on my Panasonic DVD player and it only happens once on kill bill and twice on X2. This is something that does not happen a lot.

[dave88]

Some "stutter" is caused by a player not reading media fast enough, try rewinding and playing the section again, if it does not stutter, it is more likely a media/player issue, and not related to DVDRB

[robgarib]

Do you remember my problem with my philips dvd 612 and the dvds encoded with dvdrb?
Until version 0.48 all disks stop at the beginning of the first chapter and jump directly to the second chapter.
I believed that the problem was fixed but now with version 050.b i have found the same error This error now doesn't occurs with all dvds but only with some titles.
I have encoded "The last Samurai" and "LOTR The return of the king" and everything works fine but with "Paycheck" and "In the cut" the dvd jumps.

[jdobbs]

Originally posted by dave88
Some "stutter" is caused by a player not reading media fast enough, try rewinding and playing the section again, if it does not stutter, it is more likely a media/player issue, and not related to DVDRB As much as I'd like to defend DVD-RB... There is a DVD standard for how fast a drive can read, and the peak throughput is determined by that speed and the size of the buffer. Between the two is where the maximum consistent and burst bitrates are calculated.

If you rewind and it plays right the second time. It probably means that DVD-RB has exceeded the standard (probably in bitrate).

A note to everyone on this subject. I just discovered that I had errored in the implementation of the maximum bitrate fix that was included in version 0.50 -- it wasn't being applied to NTSC interlaced sources (which is where it was actually needed). You will see it fixed in version 0.51.


note added: But... if you have a DVD-RB that wasn't written well it is very possible that you might get a "jump" as an error is encountered. Unlike PC drives, standalone players typically do nothing more than standard error correction and if that doesn't work it just ignores the error and plays on. A PC drive will give you an error.

[DVD Maniac]

jdobbs

Have you had a chance to try out some DTS tests with DVD-RB yet? I realise you must be busy with other fixes and improvements but I am hoping you can nail this problem as I have a queue of titles with DTS I want to use DVD-RB for. If you think this is going to be a longish wait for a fix please let me know and I will pursue other backup alternatives when I wish to retain DTS audio

Thanks

[jdobbs]

Originally posted by DVD Maniac
jdobbs

Have you had a chance to try out some DTS tests with DVD-RB yet? I realise you must be busy with other fixes and improvements but I am hoping you can nail this problem as I have a queue of titles with DTS I want to use DVD-RB for. If you think this is going to be a longish wait for a fix please let me know and I will pursue other backup alternatives when I wish to retain DTS audio

Thanks No, sorry I haven't. Just a question for the reading audience: Are there folks our there who are keeping DTS with no problems?

[The_Flash]

I kept DTS only on my Pitch Black backup. Watched the entire disc through without any problems.

[Noah]

Originally posted by jdobbs
No, sorry I haven't. Just a question for the reading audience: Are there folks our there who are keeping DTS with no problems?
In most cases, I have no problems with DTS. I have had issues with audio dropouts at chapter breaks with DTS.

[insertdisk]

I have had no trouble with DTS on two discs so far -

[brikin]

Even when I rewind I am still getting the audio stutter. I am only keeping the DD track.

[cyberbob25]

I am an extreme newb at this, so please forgive me if I should have posted this in the newbies forum, but it is relevant to this thread.

So far, I have only backed up one movie with RB, that being Runaway Jury. I installed Rebuilder .52, CCE 2.67.00.10 Trial, the current version of ECL, and AVISynth 2.5. I configured everything according the guidelines at afterdawn.com, and this forum. When I configured CCE mode, I set it at CCE SP 2.66+, as it would not encode at all when I had CCE Basic 2.67+ selected. I kept all advanced CCE options at default.

Looking through that status window, it shows that the max bitrate is 7038. I removed all audio streams with the exception of AC3 5.1. When I play this in my Sony DVP-CX985V, the Dolby Digital continually drops. (Works fine with all movies from DVD Shrink)

I haven't had the chance yet to test it in any other DVD player, but I'll post an update after I get home today and check. It seemed to play fine in WinDVD, but I had only tested it there to make sure it loaded up OK. I just wanted to post this to let everybody know that people are still having this problem, and not only with DTS.


EDIT1: Forgot to mention that I was burning this to TDK RW at 4x (I know, not the best media, but worked fine at this speed for Shrink). Burned with latest version of Nero (6 something) using DVD video template.

[topsare]

I have the very same setup as cyberbob25 and I also have audio dropouts. I just did Gladiator(PAL), removed all audio but AC3 5.1 and kept two subtitles.

About two minutes from the beginning of the film there are several audio dropouts. The picture quality, however, is perfect.
I tried to burn it to two different disks but the result was same. Original plays ok and it also plays ok in PowerDVD.

I have a Pioneer DV-444.

[brikin]

JDobbs I ran the DVD created by DVD Rebuilder through Nero Recode 2 and I still get the audio dropouts.....Thought I would let you know.

[Joergen]

Are you all (who get audio dropouts) using digital outputs?

[topsare]

Yes, at least I am using digital output.
The dropouts are short but still clearly noticeable.

[Noah]

Yep, digital out -> receiver.

[brikin]

I am using digital outs also.

[cyberbob25]

I've tried it in two of my DVD players now (Sony, and Pioneer multi-region). The Sony goes through optical out to my receiver, the Pioneer is only composite out. Both experienced the audio dropouts in the same places (seems that it happens when the camera changes, such as when two people are talking, and it switches between the two).

I loaded the encoded files into WinDVD and they played fine (I know that software players are more forgiving). I also played the burned disc on my computer in WinDVD and it played fine.

Alas, I have figured out what was causing it on my end. I was using DVDReMake to strip out extras, trailers, warnings, credits, and production logos. I did a full backup of the disc last night, and this worked fine in all of my DVD players.

Can anybody suggest a program that can be used to strip out all of that stuff safely without causing audio problems in ReBuilder?

[Joergen]

So its digital output AND/OR preprocessing with remake. The first being a legitimate bug and the latter being what a 36$ patch might not even fix ;)

[topsare]

Not so hasty.

I did also some preprocessing with DVDRemake(2.0.1). But after problems I did full backup without any preprocessing and the audio dropouts were still there in the same places.
And the disk that I had problems with played ok with my friend's standalone (ProCaster, 50€).

So the problem, in my opinion, is similar to the problem with the subtitles: most of the players are not so picky and they play the DVD even if it is not exactly according to the DVD standard.

I wonder if the problem could be related to the max bitrate. I will try it again tomorrow with the following settings and let you know it it had any effects to my problem.
[Options]
min_bitrate=300
max_bitrate=8000
[CCEOptions]
VBR_bias=33

Topsare

PS. RB is still the best software around

[cyberbob25]

I too was wondering if my problem was due to bitrate spikes (especially since it seemed to only happen when the camera switches from one view to another). However, during encoding, RB reported that max bitrate was only going to be 7038 (not sure if this takes into account bitrate spikes or not).

[jdobbs]

Originally posted by Joergen
So its digital output AND/OR preprocessing with remake. The first being a legitimate bug and the latter being what a 36$ patch might not even fix ;) I don't see how digital output could be related. The data is stored on the disc the same way no matter which output you use (digital or analog). If there were a difference between digital and analog it would almost exclusively have to be the player that was at fault.

I do think there is a problem with audio dropouts, and I think it is related to how DVD-RB makes the decision as to where in the stream (GOP) an audo packet gets added. If you took one of the offending VTSs and ran it through IFOEdit, then reintegrated it (using IFOUpdate) I think you'd see the dropouts go away. I'm working this issue.

[Noah]

Glad to hear you're working on the audio dropouts, jdobbs. Have you been able to reproduce this issue yourself?

[mrslacker]

I use digital output, but I don't see why that would matter either. A dvd player must read the same data even if it is down sampling first to use the analog jacks.

I have been fighting with a video freeze on The Legend of the Drunken Master for a few days now. 56 seconds in on my Sony progressive scan, it will freeze for 10-15 seconds. It is near a scene change and immediately after a bitrate spike according to Bitrate Viewer. I've tried the suggested bitrate options, as well as a spetrum of bias settings (20-45). The bitrate has several peaks above 9000 no mater what the encode settings (above 10000 with bias 20). The DD audio is 448 kbps, which seems high for most AC3 streams. For the sake of extra info, it is a pretty high average bitrate encode, being in the neighborhood of 6Mbps according to the DVD-RB prepare.

Oh, I'm using CCE CP 2.67.0.27 and RB .52 and I've done a couple of dozen movies since v.45 before having this problem. I have also tried this same encode with QuEnc .51 and it freezes in the same spot! The bitrate profile looks similar after all. The only other time I've had freezing video on my stand-alone was with Deer Hunter using vbr bias 20 (its a long movie), and it worked fine at bias 25 (and RB v .49). Can I give any more info to help?

One newbie question: does Bitrate Viewer report video+audio bitrate for vob files or just video?

[jdobbs]

Originally posted by Noah
Glad to hear you're working on the audio dropouts, jdobbs. Have you been able to reproduce this issue yourself? I've seen it a couple of times -- so yes, but very rarely.

[onesoul]

I am thninking about the max bitrate being the guilty with the dvd-rb audio droput but I am not sure. I did a backup of an avi file (progressive 23.976 ntsc) (not using dvd-rb) with cce 2.67.00.27 and although I set the max bitrate to 9000, with bitrate viewer it would reach 9300 b/s, no problems there but the weird part comes now, after applying pulldown on the encoded video, bitrate viewer shows a peak of 10390 b/s. (Nominal bitrate information doesn't change to what I set in cce, in this example 9000 b/s)

My questions, is bitrate viewer reliable and why max bitrate change after pulldown?

[Joergen]

Digital output is NOT always the same as analog. There are copy protection schemes for digital output/input and if the decoder thinks that the input signal is copy protected momentarily it will cut playback. I read something similar relating to DVD Shrink. At least Sony amplifiers report about the copy protection on the front display.

I think its mainly to prevent recording from a store-bought MiniDisc album.

[Joergen]

Originally posted by onesoul
My questions, is bitrate viewer reliable and why max bitrate change after pulldown?

Doesnt bitrate viewer think it is processing 29 or 25fps material? Hence the 20% increase with 23fps material.

[cyberbob25]

After my problems the last few days with the Dolby Digital dropping out in Runaway Jury (as per previous posts, it works if I backup entire movie), I decided to try a different movie.

This time, I edited out all extra content (with DVDReMake once again)in Catch Me If You Can (all warnings, 2 of the 3 redundant menus, and the longest opening credits of any movie I have ever seen) leaving about 7.3 GB of movie (Runaway Jury was just over 5 after stripping).

Using rebuilder, I stripped out all soundtracks but DTS, and this disc works fine (at least no dropouts in the first 15 minutes).

I just wanted to add this so that we know that there is no definitive problem with DTS as others have previously posted, and that DVDReMake isn't the definitive problem either.

[DVD Maniac]

I am still experiencing Audio Drop outs on my latest attempts (but the video freeze ups and stuttering appears to have been fixed). Originally I was only finding this with DTS audio but now I am definitely getting audio drop out on DD5 as well

Examples (PAL - Interlaced)

Master and Commander - DTS
Intolerable Creulty - DTS / DD5
Paycheck - DD5

Random audio drop out at seemingly unimportant points in the film (ie not chapter points or high video bit rate / complex video scences).

All episode based backups with DD5/2 appear to work fine - just movies that experience the problem - hope this helps with the problem solving jdobbs!

[Joergen]

If I remember correctly, even the ORIGINAL of master and commander PAL has slight audio dropouts.. have you checked? Another one that has odd digital audio squirk/compacting is Love Actually PAL.

[topsare]

I did the test I mentioned earlier (max_bitrate=8000...) and it really had some effect. There is a point in the very beginning of the film before the big fight where before there were two dropouts; one longer and one shorter. Now there is only one short.

I hope this helps tracking the cause.

[onesoul]

Originally posted by Joergen
Doesnt bitrate viewer think it is processing 29 or 25fps material? Hence the 20% increase with 23fps material.

I think it is a consequence of the pulldown flags that duplicate certain frames (2:3) so it would explain the increased bitrate. I did the encode with the following settings: 1 pass vbr, Q:21, min:0 max:8000 b/s, iqp: 16. It is ntsc 23,976 with pulldown flags with ac3 448 kb/s. I then watched the result at my set top dvd player which displays current bit rate and at many points it reaches 10 mbps. So maybe it would be better to use 8000 b/s as max setting with cce to be inside dvd standards. What do you think jdobbs?

[jdobbs]

Originally posted by onesoul
I think it is a consequence of the pulldown flags that duplicate certain frames (2:3) so it would explain the increased bitrate. I did the encode with the following settings: 1 pass vbr, Q:21, min:0 max:8000 b/s, iqp: 16. It is ntsc 23,976 with pulldown flags with ac3 448 kb/s. I then watched the result at my set top dvd player which displays current bit rate and at many points it reaches 10 mbps. So maybe it would be better to use 8000 b/s as max setting with cce to be inside dvd standards. What do you think jdobbs? But the DVD standard allows 9800Kbs for video and can legally burst even higher (over 10Mbs)... that's why the max is set to 9000, to be safe. I would have to make the assumption that CCE can be off on its max bitrate by over 8% in order to come to the conclusion that 9000Kbs was too high. Based on the accuracy of CCE's other settings I have to believe that isn't the problem.

I'm a realist and always first assume that it is my code causing the problem until I prove otherwise. I'm still inclined to believe it is the way I'm interleaving the audio and video streams that is causing this.

[jdobbs]

Originally posted by topsare
I did the test I mentioned earlier (max_bitrate=8000...) and it really had some effect. There is a point in the very beginning of the film before the big fight where before there were two dropouts; one longer and one shorter. Now there is only one short.

I hope this helps tracking the cause. This is more probably because the video is now a different size and therefore is interleaved differently. Not "positively", but "probably"

[onesoul]

jdobbs you missed the part that after pulldown it would achieve higher bitrate, this last encode I did was at (cce) max: 8000 b/s and it would reach 9447 b/s, video alone, with bitrate viewer. I did the same encode before at max 9000 b/s and after pulldown it would reach 10390 b/s.

So the the 10 (audio 448kbps included) mbps bitrate I saw at my stand alone player was the one which I encoded with cce max of 8000 b/s. If I had used the encoding that would reach 10390 b/s it would be very close to the 10.08 mbps standard limits of DVD. Don't forget to add audio bitrate of 448kbps to the bitrates video alone and adding another audio track would definitely pass DVD standards.

I did these encodes 1 pass vbr out of dvd-rb but I think it applies here too.
I would only suggest a lower max bitrate encoding (like 8000b/s which btw is the default used by dvd2dvd) to ntsc backup that requires pulldown flags. Pal isn't affected.

Hope I was clear to what I tried to say.

Cheers

[Noah]

I'm not sure Bitrate Viewer is always entirely accurate. I've been using it to check my encodes ever since the stutter issue cropped up, and sometimes it appears to be an accurate indicator of bitrate peaks that translate into stutter, and other times not. I.E., BV may show peaks of 10600k that result in stutter on one disc and peaks of 10600k that don't stutter on another.

Nevertheless, it has been my experience that CCE will produce bitrate peaks (according to Bitrate Viewer) that exceed the max bitrate specified with progressive/24fps as well as interlaced/30fps sources.

Still, its simple enough to lower the limit to compensate (when I remember ;) ), and I've eliminated bitrate as a potential factor in the audio dropout problem, which I consider to be the most serious issue I have with Rebuilder at the moment.

[brikin]

I agree with Noah. I just re-did my problem DVD and set the max bitrate to 5000. Using Bitrate Viewer I do not get any spikes over 5600 and I still get the audio stutter when I play the DVD. JDobbs sounds like he is on the right track.

[DDogg]

Noah, the beta version of DVDLav Pro available here (http://www.mediachance.com/dvdlab/historypro.html) has what seems to be a very accurate bitrate viewer tool avaialable in the tools menu. Also, it is very fast.

It has been a real eye opener for me. The amount of spikes far above the max bitrate set in CCE is very unsettling. My old Apex has a problem with these spikes and the glitches do show exactly where the spikes are. You can see where to compare as this tool allows you to see the frame/s that corresponds to the spike. Very handy. Whether the actual number of the spike shown by the tool is accurate I don't know, but where it shows a high spike is where my Apex glitches.

It seems like I have to drop my MAX down to 6000-7000 to avoid these problems on the Apex. My newer players do not have this problem so maybe I'll just junk it. Seems like my son's PS2 also has a few problems with spikes, but I have not pinned it down enough to really say.

[jdobbs]

Originally posted by DDogg
Noah, the beta version of DVDLav Pro available here (http://www.mediachance.com/dvdlab/historypro.html) has what seems to be a very accurate bitrate viewer tool avaialable in the tools menu. Also, it is very fast.

It has been a real eye opener for me. The amount of spikes far above the max bitrate set in CCE is very unsettling. My old Apex has a problem with these spikes and the glitches do show exactly where the spikes are. You can see where to compare as this tool allows you to see the frame/s that corresponds to the spike. Very handy. Whether the actual number of the spike shown by the tool is accurate I don't know, but where it shows a high spike is where my Apex glitches.

It seems like I have to drop my MAX down to 6000-7000 to avoid these problems on the Apex. My newer players do not have this problem so maybe I'll just junk it. Seems like my son's PS2 also has a few problems with spikes, but I have not pinned it down enough to really say. I wonder if this is happening on all versions of CCE or only a couple???? If worse comes to worst I can always set the default maximum lower -- but it may have a negative affect on picture quality.

[onesoul]

If bitrate viewer is accurate or not I don't know for sure but those kind of bitrates (plus audio) are reached by actual bitrate display of my stand alone (toshiba sd210e).

[DDogg]

http://img31.imageshack.us/img31/3581/Q19-100-7000.jpg

These are very prelimary comments, but it seems as if the spikes are always no more than 2k over the Max. I did encodes at Max 6000,7000,8000 and 9000. Each time the bitrate viewer in DVDLab showed spikes about 2k more than the max. So, if this is true, then setting 7000 as Max effectively gives you a 9000 Max. Sounds nuts to me though.

The image posted above is from CCE 2.67.00.27 OPV 19 100Min-7000Max.

DvdLabs BR viewer shows peak at 7915 but when the cursor is moved to the single line peaks, it shows 8961. This problem seems worse on OPV, but multipass stills shows the problem to a lesser degree.

I think I'll shift to 7000 as Max for a while and see how that works on the old Apex. Btw, my tests were done using straight CCE. DVD-Rb was not involved. It seems to be a CCE specific problem.

[onesoul]

I tried the bitrate viewer of dvd lab pro (nice program btw) with the encoded video I mentioned earlier (opv max:8000 min:0) and although it appears a peak of 8594 kbps, with the cursor I find 10020kbps (same thing that happened to DDogg), my toshiba corroborates this as I mentioned before.

I'd like to emphasize that the increased bitrate happens after pulldown so cce isn't really doing a bad job. With pal this problem doesn't occur. So imo there's only need to reduce max bitrate with progressive video that needs pulldown (23,976fps film -> 29,97fps ntsc destination). (With interlaced sources dvd-rb already reduces the max bitrate for other reasons.)

[DVD Maniac]

If I remember correctly, even the ORIGINAL of master and commander PAL has slight audio dropouts.. have you checked? Another one that has odd digital audio squirk/compacting is Love Actually PAL.

Thanks for the tip Joergen, although I did not find audio drop out on the original and believe me, my Phillips set-top will find even the slightest mastering errors! Shrink did the backup with DTS with no audio drop probs so I still think this is a problem related to the way Rebuilder is re-muxing the audio and video streams (as jdobbs has mentioned).

I have run all my failed (audio drop out) Rebuilder attempts through bitrate viewer and I can see no definitive correlation between bit rate spikes and the audio drop outs. They appear to occur randomly and do not coincide witn Chapter points or complex video scenes.

As posted previously, episode based sources with DD2 and DD5 audio tracks have performed consistently well with Rebuilder - I just don't understand why?????

[jdobbs]

Originally posted by onesoul
I tried the bitrate viewer of dvd lab pro (nice program btw) with the encoded video I mentioned earlier (opv max:8000 min:0) and although it appears a peak of 8594 kbps, with the cursor I find 10020kbps (same thing that happened to DDogg), my toshiba corroborates this as I mentioned before.

I'd like to emphasize that the increased bitrate happens after pulldown so cce isn't really doing a bad job. With pal this problem doesn't occur. So imo there's only need to reduce max bitrate with progressive video that needs pulldown (23,976fps film -> 29,97fps ntsc destination). (With interlaced sources dvd-rb already reduces the max bitrate for other reasons.) If what you are saying is true, than this may not even really be a problem. The DVD standard (9.8Mbs for video and 10.08Mbs total) is at the level it is because of the 1x spin speed of the disc and buffer space used in reading from the disc. It only applies to the video as it is stored, not as it is interpreted. Since the 2:3 pulldown is applied after the disc is read, the apparent "illegal" bitrate probably doesn't really exist or apply. This is the very reason you can get better quality on a FILM based source than a standard NTSC source.

This kind of question should be aimed at one of the true DVD geniuses like MPUCODER. He has a lot better understanding of the standards than most.

I'd say the best test would be to run this maximum bitrate test against some DVD masters that have pulldown applied. My guess is they might fail also.

[DDogg]

Original video stream - So much for my theory as this will play fine on the Apex. Now I am even more confused.

http://img10.imageshack.us/img10/1589/Original.jpg