It seems that 0.63a killed all the stuttering problems...if not,plz report here! [Archive]

TuRiSOft

7th October 2004, 14:58

Same here as far as I can see (But I admit i never had this problem on over than 100 titles backed up w/RB - I only stopped using it on ver .56-.61 ).

Is it the time to think about multiangles and/or menus re-encoding and re-authoring?

I think you just did a great job JDobbs and i wish to donate you my 2 cents as soon as this feature will be implemented ( Ghostbuster 1 - Pal is waiting for it on my 4th HD ).

But for now , seriously speaking , I think you got to take a rest and , like someone else said , sit down on your sofa and watch the just delivered Star Wars collection.

You got to be real tired and all you did for us is really incredible .
I wish only to thank you for all.
Bye!!!

erdoke

7th October 2004, 15:25

Originally posted by Fr4nz
You know man...Jdobbs it's a geniussssssssss :D :D

Anyway J. could you explain us what you have changed in this 0.63a last version, in order to eliminate all the stutter-related problems??

He always explains changes when posts a new version. ;)
V. 0.63 & 0.63a: http://forum.doom9.org/showthread.php?postid=548359#post552391

Fr4nz

7th October 2004, 18:49

Originally posted by erdoke
He always explains changes when posts a new version. ;)
V. 0.63 & 0.63a: http://forum.doom9.org/showthread.php?postid=548359#post552391

Yes but the 0.63a changelog doesn't explain WHAT he did in order to solve stutter problems, because he says that "for compatibility problems the "runt" VOBUs code was disabled".

FilipeAmadeuO

7th October 2004, 20:59

Problems gone for me too .
Now is time for seamless brenching support :)
Please !!!!

eriksen76

7th October 2004, 22:30

I encoded Van Helsing (PAL) and two other movies today. No problems so far :)

I only encountered one problem on an older movie which i've tryed encoding with all versions of DVDRB. It has a small drop out on some standalones. And that is still there with this version. The drop out is at exactly the same scene every time.

But I guess there's no problems with the new movies

/Julius

Faust2

8th October 2004, 08:23

from the changelog:

Version 0.63a

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

- Corrected rebuilding code to prevent "runt" VOBUs (less than .4 seconds). This will bring the newly authored stream more tightly into compliance with DVD standards. This is KNOWN to have caused audio/video stutter on some DVDs.

- Added support for Canopus ProCoder through Robot1's EclPro utility. Nice work Robot1, you continue to provide outstanding support to the DVD community! Please note that this feature needs testing.

- Modified the maximum bitrate on output so it lowers to accomodate large and multiple audio tracks. This should help to keep the bitrate bursts under DVD maximums. Maximum bitrate is now set to MAXBITRATE - AudBitrate -- so using the default it would be 9,000Kbs - AudBitrate. Note that DVD standard is no greater than 9,800Kbs for video and no more than 10,008Kbs total audio/video/subs.

- Corrected a bug introduced in v0.61 related to audio/video sync. This could have caused stutter and audio dropout on some DVDs. An option has been added to the "Options" menu called "Special Error Avoidance" for special cases in which certain #0003 or #0004 errors are encountered -- this is temporary and will be removed in future versions. It is recommended that this option only be used as a last resort after you've encountered one of these errors -- and not before, as it can cause other unpleasant effects. This option ONLY affects the rebuild phase. It is meant to counter what appears to be bad source timing parameters.

[v0.63a]

- Disabled runt VOBU code for more testing. It appears to have caused more problems then it solved.

Just out of curiosity: what settings did you stutterers use to make this happen? special error avoidance or default? Its sounds a bit miraculous, because jdobbs removed the "runt VOBU code"...

but anyhow; gone is gone :)

Congrats to jdobbs; AAA (awesome as always). Now I can maybe finally have aaaaaannnglllllless?

(not being serious)

dannyv

8th October 2004, 16:03

I've done star trek original series and star trek voyager season 4 as well as several other single movies and all came out perfect with 63a. Great job jdobbs.

jdobbs

8th October 2004, 17:01

It's getting there. I think the stuttering was not caused by a single problem, but instead by different small issues that may have manifested themselves differently on certain players. The combination of fixes over the last few versions has corrected those individually.

There was also a portion of the fix I put in for runts (that was problem free) that remained in 0.63a.

FilipeAmadeuO

8th October 2004, 17:58

Are you planning to add seamless breamching/multi-angle support in the next version ???
PLEASE !!!!!!!!

jdobbs

8th October 2004, 18:18

I will come out with a version that will let you choose an angle to keep first. I've gotten sidetracked several times on what I consider important bugs (like stutter/audio dropouts)...

Morien

9th October 2004, 06:14

Stuttering remained for me using 0.63a. (Played in PowerDVD 4 and 5)
Funny thing is, when I tried reverting back to an old version which I never had trouble with, stuttering continued :S

Fr4nz

9th October 2004, 08:24

Probably you are using dvds of poor quality. ;)

FilipeAmadeuO

9th October 2004, 13:21

Jdobbs - One question :
Is it possible to encode with dvd Rebuilder a seamless breanching film only by encoding the VOB ID´s that aren´t exacty semaless (ILVU).
You know that in this kind of movies a big part of the movie isn´t in seamless brenching.
Can it be done ??

jdobbs

9th October 2004, 14:14

@FilipeAmadeuO

Pretty easily. In fact that may be a good alternative for consideration on most discs. One of the difficulties of seamless branching is that since two or more streams are interleaved -- and the spin-speed of the disc is constant, the maximum video bitrate gets cut at least in half making a reduction to DVD-5 less likely to be acceptable in terms of quality.

FilipeAmadeuO

9th October 2004, 14:29

You maybe right. But usualy the ILVU vob ID´s are very short so you don´t need to compress the others vob ids very much.
I´ve already tried manualy with Alien SE and it could be done . The problem is that i need a lot of time expending
Is it very dificult to implement in DVD Rebuilder ??
Just to see who it results !!

Morien

9th October 2004, 15:37

Originally posted by Fr4nz
Probably you are using dvds of poor quality. ;)

:devil: :p
Nah I don't think I am (maybe, but no prob with DVD Shrink aside from the terrible outputted quality, and I've tried numerous other titles from the same authoring studio), as before I upgraded there was no stutter (tried with old (0.56 ver) and no stuttering, but another prob, so upgraded to 0.63a and stuttering came, then downgraded again to 0.56 and stuttering remained).

HOWEVER, I've uninstalled everything DVD-RB related (instead of just copying the exe over the existing files) and reinstalling and am testing again. SO, forget my initial post and I'll see if the problem persists. :) (Fingers crossed tho)

eriksen76

9th October 2004, 16:38

I had some minor dropouts in sound today, on a movie on which I had kept both DTS and 5.1 sound. The drop outs where only in the beginning where (according to my dvdplayer) the bitrate was just about 10000 sometimes. Which I believe is critical and causes stutter. The orginal dvd had a max of 9800 or so. It never reached 10000.

Pretty strange as Dvdrb automaticly set the max bitrate to apx 8000, then how can it still be higher than 8000????

But now I will try to lower the default (9000) to 8500 and see if this helps. Then it should automaticly go below 8000.

Any ideas??

/Julius

jdobbs

9th October 2004, 17:47

QuEnc or CCE?

eriksen76

9th October 2004, 22:16

CCE 2.67.00.27

I have never seen this kind of problem before. We are talking minor overall compression (apx 90%) and an average bitrate of 6000 or so (According to dvdrb preperation).

Maby that gives you a hint what is wrong

EDIT: Just ran the first vob file through bitrate viewer and it gave me a max of apx 8500. My 2 dvd players give me peaks @ 10000 - This should not be possible - especially with your version 0.63a

I have vbr_brate_max=7298 in rebuilder file

/Julius

jdobbs

10th October 2004, 03:31

You may have to be careful when referring to what the DVD player says. For example.. if you are running an movie encoded as FILM and it is playing back using telecining, it can be only 8,000Kbs peak could appear to have a peak of 8,000 + (8,000 * .25) - or 10Mbs. That's because there are pseudo frames being created by the telecining that don't actually exist on the disc.

The 7298 doesn't include audio or subpictures.

Morien

10th October 2004, 06:21

I've tried what I can think of and lag still remains. It's in the exact same spot each time I encode though and the exact same moment in every episode (same part of the intro credits) and even at that point in the creditless opening (extra feature). Is it an authoring problem with the original you think?

eriksen76

10th October 2004, 13:56

Originally posted by jdobbs
You may have to be careful when referring to what the DVD player says. For example.. if you are running an movie encoded as FILM and it is playing back using telecining, it can be only 8,000Kbs peak could appear to have a peak of 8,000 + (8,000 * .25) - or 10Mbs. That's because there are pseudo frames being created by the telecining that don't actually exist on the disc.

The 7298 doesn't include audio or subpictures.

No matter what I do I still get the "glitches" in sound in the first minute of the movie. But I tryed extracting the audio/video and made the vob set from those (Dvdmaestro). Then I in-corparated these vobs into the titleset and updated the ifos. No glithces now.

So there might still be a small bug in the rebuilding process. Now I'm watching "Monster" which has both DTS and 5.1 audio and I have not had any probs with playback so far :). If I remember right the compression was apx 75%)

The movie with the problems has apx 90% compression and I'm keeping both DTS and 5.1 which gave me a avarage bitrate of 5900. Maby the high avarage bitrate together with keeping two audio tracks causes the errors??

What do you think?

/Julius

Msc_Alex

10th October 2004, 21:16

Thats what I did on SW4 (http://forum.doom9.org/showthread.php?s=&postid=554638#post554638) worked for me too. But i left out 6 subs, try to make a 1:1 compare may tell more. But what to look for ??

wfn1

10th October 2004, 23:12

Originally posted by eriksen76
Maby the high avarage bitrate together with keeping two audio tracks causes the errors??

What do you think?

/Julius

i'm not so sure about the high average having to do with it because i had titles where my avg was below 3000kbps and i still got those glitches.

eriksen76

11th October 2004, 12:38

Originally posted by wfn1
i'm not so sure about the high average having to do with it because i had titles where my avg was below 3000kbps and i still got those glitches.

Well I did a backup once again, but kept more of the extra material and therefore achieved a lower avg bitrate.

No stutter now!

DVDRB might have probs rebuilding material with high avg bitrate

/Julius

Mojo300

11th October 2004, 14:19

I am also getting stutter, tried doing JuWanna Man and got stutter with 63a.

eriksen76

11th October 2004, 15:40

Still only on some standalones? What can it be in the rebuild phase that causes this?? I guess thats what we all keep asking.

Everything is working like a charm

/Julius

wfn1

11th October 2004, 17:22

its worse with some standalones than it is with others. i have 8 discs that i did with .63a, futurama seasons 3 and 4. because the originals have a lot of extras and are around 8gb to begin with my average bitrate is not that high. usually it hovers right around 3000kbps. i tested my resulting discs in several players and here's my breakdown:

pioneer dv-c503 (old changer) - some stutter, very little
pioneer dv-525 (old as dirt my 1st player) - no stutter
panasonic rp-82s (2 yrs old, higher end player) - noticeable stutter
philips dvp642 (newer player, bought 3-4 months ago) - lots of stutter on every episode usually 5-6 minutes in

tremens

12th October 2004, 05:35

I'm getting a little stutter with a few movies still. I do not see any common denominator in any of them. Definately not just a high bit rate issue though. I've tried in multiple standalones, they all do it in same place with varying degrees of severity. Kaffiene(Xine) play the DVDs on the HD with no stutter.

jdobbs

12th October 2004, 13:47

I just ask everyone to make sure the stutter is not related to bad discs... I had the problem once last week -- and it went away when I burned the same files to a different disc.

eriksen76

12th October 2004, 14:58

Originally posted by jdobbs
I just ask everyone to make sure the stutter is not related to bad discs... I had the problem once last week -- and it went away when I burned the same files to a different disc.

As far as I'm concerned its not related to bad discs. I burned the same movie (Scream 3) twice and got the same stutter in the beginning.

Again, I extracted the video/audio and re-compiled with Dvdmaestro. Put the vob files into old archive and updated with ifoupdate.

Burned the sucker, and the stutter was gone

/Julius

jdobbs

12th October 2004, 15:46

You're doing something different than others... among the DVD-RB user base this is a rare problem, but apparently not for you. We have to identify what that something is.

eriksen76

12th October 2004, 16:37

Originally posted by jdobbs
You're doing something different than others... among the DVD-RB user base this is a rare problem, but apparently not for you. We have to identify what that something is.

Is there any way to compare the output by maestro and the one with DVDRB, cause like you wrote, something different must be in the output? Cause now I've used the "Maestro Method" lots of times and it eleminates all errors for me every time.

/Julius

wfn1

12th October 2004, 17:41

it's definitely not the media as i'm using ritek g04 that ive burned literally 1000s of with pretty much 100% success rate. plus my problems are spread over 8 dvds and on top of that some players have the problem while others do not.

it would be interesting to see what maestro changes in the files over dvdrb. maybe if the change is small enough it could be easy to analyse and single out. say if eriksen and jdobbs had the same movie backed up via the same version of rb with absolutely the same settings it might be possible to make an image of it and an image of the maestro compiled one and run a file compare on it or better off make a ppf patch that would have the differences introduced by maestro. the jdobbs might be able to narrow it down. but again that would only work if the changes are small enough.

Sir Didymus

12th October 2004, 18:25

@wfn1
Not so simple, believe me.
If it would be simple jdobbs would have fixed it months ago.

There are navigation, video, audio and subpicture packs involved in the remux.

Each packet has its proper timings, and their interleaving should obey a lot of constraints to be performed properly.

If you use a different remux engine, packet placement and timestamps (SCR, PTS, DTS) change a lot even for short segments.

Then what you discover once understanding that if the remux is performed by Scenarist or DVD Maestro and IfoUpdate things are working ?

This is not a solution, it is just a workaround: the remux should be done internally by DVD-RB.

It would be just very interesting to understand how many users suffer from this type of trouble. My perception is that they are quite rare...

For those I will really suggest (at least for the moment) to stay relaxed and to apply the workaround, initially suggested by Msc_Alex...

All the best,
SD

J-Wo

12th October 2004, 18:45

I too am having problems with audio stutter. I don't know if it may be because I was using DVD Stripper to remove some VTST's, so I tried VobBlanker the other day and still the same problem. Today I'm going to try PgcEdit, and I hope this will solve it. The stutter occurs at the same points, frequently, and on both a newer Panasonic DVP-S25 and an older Sony DVP-NS315. I wonder if the cause may also because I am using RB-Opt to change CCE settings and to run OPV prediciton. I prefer to encode with the KVCD notch matrix, changing DC precision to 8, Gop length to 15, Quantizer Characteristic to 25, and min/max bitrate to 64-5500. If I am still having problems I will next try to encode again but this time using CCE standard matrix.

Sir Didymus

12th October 2004, 18:52

@J-Wo
Applying all that preprocessing, tuning and tweakings all together, I am really surprised you are suffering from just some stutters... :eek:

J-Wo

12th October 2004, 19:24

@sir didymus:

Well isn't removing unwanted vts such as trailers and extras standard practice if one is making a DVD9 to DVD5? Also, the kvcd notch matrix is a perfectly acceptable matrix with almost 20% compression over standard matrices, with absolutely no visible difference. You should read more about it at kvcd.net. Besides isn't it the whole point of DVD-RB to be able to use CCE to encode your material? I simply don't like the default CCE settings. BTW I have never had any problems with "so much pre-processing, tuning, and tweaking" before using manual methods, except with DVD-RB.

J-Wo

12th October 2004, 22:37

I read in the Star Wars thread that people have had better success with v0.60b. So I downloaded that and used it to rebuild the D2VAVS which was created by v0.63a after the Encode process. The result had the same audio dropouts in the same locations. They appear on both my standalone players (one Sony, other Panasonic) but not with PowerDVD. I'm going to try preparing and encoding using 0.60b. Has anyone else found more success with an earlier version of RB?

Faust2

13th October 2004, 01:00

Originally posted by J-Wo
@sir didymus:

Well isn't removing unwanted vts such as trailers and extras standard practice if one is making a DVD9 to DVD5? Also, the kvcd notch matrix is a perfectly acceptable matrix with almost 20% compression over standard matrices, with absolutely no visible difference. You should read more about it at kvcd.net. Besides isn't it the whole point of DVD-RB to be able to use CCE to encode your material? I simply don't like the default CCE settings. BTW I have never had any problems with "so much pre-processing, tuning, and tweaking" before using manual methods, except with DVD-RB.

That comparison isn't valid. DVD-Rebuilder has its name for a reason. It rebuilds the original DVD structure, which is approved by the fact that the dvd is sold like this. If you strip content, this structure is likely to be screwed up. If you are able to use this data whith an authoring package to produce a functional dvd, this has nothing to do with rb's approach; rb isn't an authoring tool which has to cope with nearly every material.

Nonetheless it could be a bug in rb, so just make an encode without any preprocessing and you know more.

wfn1

13th October 2004, 01:04

hmm, i was using .60b and got a stutter so i guess i spoke too soon.. tracing back i did change one thing whereas i had 0 problems before now i get the stuttering problem. i went from cce 2.50.01.00 to 2.66.01.07 .. can anyone using 2.50.01.00 confirm that they've had stuttering issues with it? i'm going to try it anyway tonight.

dragongodz

13th October 2004, 01:58

Well isn't removing unwanted vts such as trailers and extras standard practice if one is making a DVD9 to DVD5?
there is no such thing as a "standard practice". i think you will find the majority will do movie only, shrink extras to extra small size or split a dvd. removing unwanted material in a relativly easy fashion has only become easy with releases of programs such as vob blanker etc. even then they are not 100% bug free (is any program ? :D ) or guarenteed.

the kvcd notch matrix is a perfectly acceptable matrix with almost 20% compression over standard matrices, with absolutely no visible difference. You should read more about it at kvcd.net.
no visible difference to you but other could easily tell you they do see a difference. the actual film used can have an impact on this where using 1 matrix may look better while on a different film a different matrix again can look better. so please dont talk about absolutes when it is not a fact.
as for kvcd.net, hmm oh yes i am sure they are unbiased. thats sarcasm for anyone that didnt get it. the people at kvcd are already well known for thinking their matrices are the best thing since sliced bread. :D

jdobbs

13th October 2004, 01:59

Originally posted by eriksen76
Is there any way to compare the output by maestro and the one with DVDRB, cause like you wrote, something different must be in the output? Cause now I've used the "Maestro Method" lots of times and it eleminates all errors for me every time.

/Julius They couldn't possibly look alike... does the Maestro source have the original subpictures (no), the original soundtracks with exact timing parameters (no)... you have to realize that what you are asking is virtually impossible. It also wouldn't look alike if you compared DVDMaestro to Scenarist...

J-Wo

13th October 2004, 02:16

Okay well, I won't even bother to go into a debate over matrices. But I did try reencoding using the standard CCE settings supplied by Rb-Opt. Audio dropouts were still there at the exact same points but were a lot shorter, i.e. less noticeable. I'd really like to know if people have had any luck with previous versions of RB, because the problem definately hasn't gone away for me with 0.63a.

lamster

13th October 2004, 06:10

Originally posted by J-Wo
Besides isn't it the whole point of DVD-RB to be able to use CCE to encode your material?
Gee - I wonder what those "Rejig mode", "QuEnc mode", and "Procoder mode" options are for? :rolleyes:

Sir Didymus

13th October 2004, 09:25

Originally posted by J-Wo
Okay well, I won't even bother to go into a debate over matrices. But I did try reencoding using the standard CCE settings supplied by Rb-Opt. Audio dropouts were still there at the exact same points but were a lot shorter, i.e. less noticeable. I'd really like to know if people have had any luck with previous versions of RB, because the problem definately hasn't gone away for me with 0.63a.

Ok. Your post gives now some positive contribution, so let me first say I was a little bit acid, but it was not my intention to start a discussion with you. If you feel offended, please accept my apologies.

1. You'd better use the last recommended release of DVD-RB. No point in using previous releases since they may have bugs more relevant than the one you hope to magically disappear moving to a previous version.

2. About the "standard practice" of preprocessing, everybody know that RB is in beta testing, so it could be reasonably sensitive to screwed input. It has been told and repeated many times that, at least for the moment, it's better to post and report troubles ONLY ON ORIGINAL, UN-PREPROCESSED MATERIAL.

3. If you want to use RB in its actual stage of development, you may obtain exactely the same results applying cuts and manipulations of the title contents AFTER THE REBUILD AND NOT BEFORE. Just need to tune TargetSectors in order to rebuild the title with a size exceding the DVD-5 limit by the amount you will cut later.

4. Thank you for your link to the matrices at kvcd.net. I suppose to know the potential benefits of applying non standard settings to CCE. My experience is that these tuning are rarely necessary in the DVD arena; they are in some way necessary for SVCD, but for DVD they could be just useful in some specific and rare situations where the title bitrate drops in the region of 3000 Kb/s or less. Also be careful when applying non standard settings, since (especially for the GOP structure set at 15 frames) the size reduction benefits could be very relevant, but it is debatable that the produced DVD material would still be DVD compliant.

SD

J-Wo

13th October 2004, 18:10

Originally posted by Sir Didymus
3. If you want to use RB in its actual stage of development, you may obtain exactely the same results applying cuts and manipulations of the title contents AFTER THE REBUILD AND NOT BEFORE. Just need to tune TargetSectors in order to rebuild the title with a size exceding the DVD-5 limit by the amount you will cut later.

I actually came to that same conclusion just before reading your message. So last night I ran 0.63a on my Alias season 2 disc from a DVD Decrypter ISO image without any pre-processing. I use very simple avisynth filters:

Crop(28,20,-28,-20)
FluxSmoothT()
AddBorders(20,20,20,20)

This crops the framesize and add borders to make the end resolution 704x480 (also applying light filtering). I suppose it's possible RB does not like this resolution, but from my understanding it is perfectly DVD compliant. I suppose I could add NO avisynth filters and see what happens.

I then openned my rebuilder.inf into Rb-Opt and changed the CCE settings to conform with the notch matrix and manually set the Q value to 10 for OPV encoding. Again, it's possible that RB doesn't like the usage of non-standard CCE matrices, but I've posted this question to the kvcd forum and so far nobody has reported these problems.

Then I rebuilt with 0.63a. I used PgcEdit to blank the titlesets containing extras/trailers in order for it to fit onto a +RW, then burnt it. But audio dropouts were still there in the same spot. So I then used 0.60b (reported in the star wars thread to be better with these dropouts) to rebuild the already encoded D2AVS folder. The result had the same audio dropouts BUT they were much shorter.

Soooo... for my next test I uninstalled Avisynth 2.5.5 and CCE 2.67.00.27, ran RegSupreme Pro to do a registry cleanup, rebooted, and reinstalled them. Now I'm starting with 0.60b from scratch (prepare, encode, rebuild) to see what happens. If you guys think any of my avs filters or changes to CCE may be causing this, please let me know!

robot1

13th October 2004, 18:26

704x480 is a valid DVD resolution, but probably DVD-RB needs for the backup the same resolution of the original.
Other filters shouldn't give problems, but it would be better trying the backup with all default settings.
If you have problems with default settings, it's a DVD-RB bug or a wrong installation problem.
If you haven't problems with default settings, you could understand what filter or setting is the cause of the problem, helping all the DVD-RB community.

J-Wo

13th October 2004, 19:12

Alright, I'm just in the middle of rebuilding an encode made with 0.60b. If there are still problems, I'll go back to default settings. Do you think it's okay for me to still use Rb-Opt in order to manually set the Q setting for OPV encoding? This way I avoid having to do any predictions since I roughly know what Q setting to use for the movie to fit onto a DVD+RW.

robot1

13th October 2004, 19:36

I you know the Q factor you need, you can use RB-Opt to avoid prediction.

Toranaga

13th October 2004, 22:49

Killed the stutter?!?!?!?

I ran a new PAL movie through .63a and the stutter was worse than ever!!! It was so bad that I had to dumb the recode and use DVD shrink instead. Before it just stuttered when there was high bitrate, but now it does it almost all the time!

I am starting a Deep Space Nine episode disc now and will report back when it is done.

J-Wo

13th October 2004, 22:54

Okay, I have some progress to report. I used 0.60b on an iso image without any pre-processing. The only avs filter I used was FluxSmoothT. I ran Rb-Opt to do OPV prediction and set the Q setting for VTS3 (unwanted extras) to 70 so they would be highly compressed (and this way I didn't need to run any post-processing blanking). I didn't touch any of the CCE settings from whatever Rb already defined. The final result had much FEWER dropouts than before, but there is one particular location where a considerable dropout always comes up. I ran a check on the original DVD and there is no dropout there. There is also no dropout when played with PowerDVD, but there is with both standalones. In case I had bad media to start with, I just bought brand new Memorex 4x DVD+RW's.

So now I'm going to rerun the rebuild stage except using 0.63a. If *that* doesn't work, I'll start from scratch, running a prepare and encode with 0.63a. The only reason I started with 0.60b is there were reports in teh Star Wars thread that stated 0.60b might be better for dropouts. Beyond that I think I'm fresh out of ideas!

J-Wo

14th October 2004, 02:08

No difference between 0.60b and 0.63a, audio dropouts still present. jdobbs are you out there?

wfn1

14th October 2004, 04:41

man this forum is agressive :) out in the big3 we just bide our time before the next rel. comes out. here it's like a post a minute.

take a break people, go out to a bar, call your mom or whatever JD needs to rest sometimes too.

J-Wo

14th October 2004, 04:55

big3? I don't know where that is...

But for me these audio dropouts are a major problem. They crop up in every run I do with Rb. So as far as I'm concerned, the time I'm taking to make endless encodes is to try and help jdobbs narrow down the problem. And hopefully other people can provide some helpful suggestions.

Anyway back to the topic at hand... Did an encode using 0.56, no preprocessing, no filtering, standard CCE settings. There was only 1 dropout compared to 4 possible known spots where other versions have had them. So that's definately an improvement.

But right now I'm really at my wits end. Can't think of too many more combinations of tests I can run. So ya, maybe I will go call mom now... ;)

VamPYR

14th October 2004, 06:08

after many RB version that i've use, 0.63a is the most problem less version for me.
did about 50 project from my source using RB 0.63a, and didnt get a single problem. all my final project play well in various standalone player. (i burn my final project using BeAll DVD-R silver printable media)
my original source are mostly NTSC,4:3(fullsceen),some source had alot of special features, photo gallery and over 7G.
i rip it directly using DVD Decrypter and encode it using RB, CCE, default setting, 4-pass. it take about 5hrs to finish.

thanks JD for all the great job he had done for us!
(sorry for my english, it not my mother language):p

J-Wo

14th October 2004, 16:42

Just wondering... For those of you who are NOT experiencing any audio dropouts, are you doing 2 pass VBR of OPV encoding? I'm strictly doing OPV but maybe I should try VBR since I'm having so many problems.

jdobbs

14th October 2004, 17:18

Originally posted by J-Wo
big3? I don't know where that is...

But for me these audio dropouts are a major problem. They crop up in every run I do with Rb. So as far as I'm concerned, the time I'm taking to make endless encodes is to try and help jdobbs narrow down the problem. And hopefully other people can provide some helpful suggestions.

Anyway back to the topic at hand... Did an encode using 0.56, no preprocessing, no filtering, standard CCE settings. There was only 1 dropout compared to 4 possible known spots where other versions have had them. So that's definately an improvement.

But right now I'm really at my wits end. Can't think of too many more combinations of tests I can run. So ya, maybe I will go call mom now... ;) I think it's amazing that it crops up everytime you use RB -- but it has never cropped up (even once) on 90%+ of the RB users... my bet is on OPV.

Toranaga

14th October 2004, 17:31

I just ran a Deep Space Nine disc through using 5 pass CCE and the audio stutter is there but only in high bitrate scenes and not all the time like the movie I did before this.

J-Wo

14th October 2004, 17:48

Originally posted by jdobbs
I think it's amazing that it crops up everytime you use RB -- but it has never cropped up (even once) on 90%+ of the RB users... my bet is on OPV.

You may be right, I'm going to give 2 pass VBR a try. But in case anything *else* is causing problems, could you please tell me other settings you use in Rb that might be causing me problems? I'm pretty sure I have everything else at default, except I have been using Rb-Opt to run OPV prediction (that's about it).

robot1

14th October 2004, 18:02

Originally posted by J-Wo
You may be right, I'm going to give 2 pass VBR a try. But in case anything *else* is causing problems, could you please tell me other settings you use in Rb that might be causing me problems? I'm pretty sure I have everything else at default, except I have been using Rb-Opt to run OPV prediction (that's about it). Have a test encode without any tweaks (example in one click mode), and be sure to use the right .dll: http://forum.doom9.org/attachment.php?s=&postid=517471, without other mpeg2 decoders in your avisynth plugins directory.
That's the final test for your setup!

J-Wo

14th October 2004, 18:16

When I did a clean install of Avisynth 2.5.5, I already cleared everything out of my plugins directory and installed this version of mpeg2dec3dg.dll. But I did have 3 other filtering dll's (fluxsmooth, removegrain, and deen) but I'll delete them right now.

I've actually never done one click either, always 3 click. So I'll give this one a whirl... this time still with 0.59, since that has given me the fewest dropouts so far. See you guys in a couple hours ;)

edit: btw, I just finished an OPV encode with bitrates set at min=300, max=5000 to see if the default bitrates were too extreme (min=0, max=9000). No change in dropouts...

J-Wo

14th October 2004, 21:42

hey guys, sorry for making this an enormously long thread but.... What do you know! 2 pass VBR seems to have solved it. No stuttering in the 2 known locations previously seen with OPV, using Rb 0.59. I'm going to do another VBR run but this time make my CCE notch matrix adjustments thru Rb-Opt and see if that causes any dropouts. If not I'll try Rb 0.63a again and make sure the latest version has more problems or not.

So right now, at least for me, it looks like OPV is the real culprit.