Audio Dropouts.

[gizzin]

DVD Rebuilder v.57

- Made a timing correction that may affect possible audio dropouts. It also may
reduce (but not eliminate) the liklihood of encountering an error #0003 or #0004
on preprocessed (3rd party edited) streams.

I Did the movie Summer of Sam 7 Pass CCE (v2.67.00.23 SP). Dynamically Assign Cell Bitrates checked, ConvertToYUY2() checked. Not only is the audio dropouts not fixed, it seems like the movie is littered with em. Not at any specific point. I used nero to burn the output and it was burned a 2x philips potex rewriteable. Please jdobb take a closer look at this problem if you aren't already have. Test the output on a pioneer dv343.

[gizzin]

Is there anything specific i could do to help you eliminate this? I forgot to mention this someone told me that those missing packets of audio sometimes appears after using rebuilder. The missing packets are scattered throughtout the audio file in the remux phase and cause the dropouts in only certain players. Im not exactly sure how true this is as i know nothing about the remux phase it was just what i was told awhile ago.

[Oldeman]

I still have audio dropouts on several Toshiba players 4800, 4900 and 3950. No dropouts on Apex 1500, Philips DVDR75 or software players.

Using DVD-RB and Quenc .53 . Rejig does not seem to have this problem at least in my testing. Audio drop outs cause the audio to get way ahead of the video. Its unwatchable. The video however is great.

Someday we will find a simple answer for this.. til then I'll keep Shrinking or Instacopying...

[brashquido]

Hi gizzin,

Have you tested the end product from DVD-RB in a software player such as PowerDVD or WindDVD? I also used to use Nero to burn my DVD's but found that every now and then it would stuff it up (although my problems where in regards to video). If you can test the output of DVD-RB directly from the hard disk in a software player and reproduce the same drop out issue, then I'd say it will be a DVD-RB problem. With Nero in the equation there it just raises doubt in my mind.

I now use DVD ISO Creator and DVDdecrypter as my method of getting my DVD-RB projects to DVD. It is a two stage process, but it has not failed me once in 50 odd conversions, and what's more all the software is free (after you make your donation of course).

[erdoke]

Originally posted by brashquido

I now use DVD ISO Creator and DVDdecrypter as my method of getting my DVD-RB projects to DVD. It is a two stage process, but it has not failed me once in 50 odd conversions, and what's more all the software is free (after you make your donation of course).

Sorry for being off-topic but I can find nowhere this DVD ISO Creator, and I would like to try it. Can you help me where to look for? THX (Is it sg like MagicISO Maker?)

[wmansir]

I'm not sure about DVD ISO creator, but I use ImgTool Classic (http://www.coujo.de/), which I assume does the same thing (take a Video_TS folder and make a DVD ISO).

[erdoke]

Originally posted by wmansir
I'm not sure about DVD ISO creator, but I use ImgTool Classic (http://www.coujo.de/), which I assume does the same thing (take a Video_TS folder and make a DVD ISO).

Thank you wmansir, I already have it, just like MagicISO Maker. But I'm still curious about DVD ISO Creator. If it exists at all, for Google finds nothing similar.

[gizzin]

yes i've tested the output on powerdvd a software player. It works flawlessly on the computer no dropouts at all. Its just on my specific standalone. We already nailed this one down along time ago brashguido, its only with certain dvd players this happens.

[erdoke]

Originally posted by gizzin
yes i've tested the output on powerdvd a software player. It works flawlessly on the computer no dropouts at all. Its just on my specific standalone. We already nailed this one down along time ago brashguido, its only with certain dvd players this happens.

Then it should be considered that it is a quality difference between standalone players or something with the handling of DVD structure.

[brashquido]

Originally posted by wmansir
I'm not sure about DVD ISO creator, but I use ImgTool Classic (http://www.coujo.de/), which I assume does the same thing (take a Video_TS folder and make a DVD ISO).

Yep, my brain is officially cactus! ImgTool Classic is the tool I use, I had just renamed the shortcut on my desktop to DVD ISO Creator :rolleyes:

[brashquido]

Originally posted by gizzin
yes i've tested the output on powerdvd a software player. It works flawlessly on the computer no dropouts at all. Its just on my specific standalone. We already nailed this one down along time ago brashguido, its only with certain dvd players this happens.

Ok, cool. From what you've said here though it still does not conclusively point the finger at DVD-RB as being the cause of your grief. What methods have you tested with getting your finished DVD-RB project to DVD?

[erdoke]

Originally posted by brashquido
Yep, my brain is officially cactus! ImgTool Classic is the tool I use, I had just renamed the shortcut on my desktop to DVD ISO Creator :rolleyes:

Then I've already dowloaded DVD ISO Creator from coujo.de... ;)

[gizzin]

i have used clonedvd to burn it, imgtool then dvd decrypter, used dvdshrink to make iso then also used decrypter to burn, um intervideo dvdcopy to burn it... all seem to give dropouts, tried burning at 1x, and used various different types of media. Its likely that it is a dvd rebuilder problem... but its hard to find because like you said the dvd structure or whatever it is cant be handled in my player (pioneer) and toshiba from what i read in lots of forum posts.

[Toranaga]

I just tried another Deep Space 9 DVD using min. and max. bitrate values, and it did not help eliminate the stutter.

[brashquido]

Originally posted by gizzin
i have used clonedvd to burn it, imgtool then dvd decrypter, used dvdshrink to make iso then also used decrypter to burn, um intervideo dvdcopy to burn it... all seem to give dropouts, tried burning at 1x, and used various different types of media. Its likely that it is a dvd rebuilder problem... but its hard to find because like you said the dvd structure or whatever it is cant be handled in my player (pioneer) and toshiba from what i read in lots of forum posts.

Well that cetainly seems to narrow the field a bit. I have a Pioneer 355 without any such issues. How long do the drop outs last? Are you able to test your projects on any other set top DVD player other than your 343?

[gizzin]

I've tested it on a 5disc sony player(also a ps2), apex and panasonic. I don't remember the certain model numbers. I wish i did it'd be more helpful. The apex and the ps2 don't recieve the dropouts. The movie that was tested was rules of engagement. I really don't get it because I have made good movies with rebuilder without dropouts but i can't remember with what version. I've also downloaded a movie done with rebuilder and it had no dropouts at all so it really makes me think what the hell it could be. Everything passsed .40 always has dropouts and in this last case its way worse (Summer Of Sam).

[brashquido]

Have you used a program like WinMerge to compare the IFO's of the orignal to the IFO's produced by DVD-RB? Or do you think it is something else?

[lamster]

I was just posting about the audio dropout problem, and then checked on another machine which is doing a rebuild, and happened to notice that on the "Audio streams to keep" section, it's stripping out French.

Just wondering - does anyone who's experienced the dropout problem have everything unchecked in the "Audio to Remove" section of the DVD-RB Setup panel? If not, maybe I'll retry but keep all languages, and see if this makes a difference.

[gizzin]

I doubt it is the ifo files.

[jdobbs]

Originally posted by gizzin
...Not only is the audio dropouts not fixed, it seems like the movie is littered with em. I agree. I tested the changes on a few DVDs and it appeared to be ok. But yesterday I tested it on my copy of "Once Upon A Time In the West" -- and for the first time I also got audio dropouts and video jitter. I'm going to post an update (v0.57a) tonight.

Sigh... I thought I'd nailed one.

[erdoke]

Originally posted by jdobbs
I agree. I tested the changes on a few DVDs and it appeared to be ok. But yesterday I tested it on my copy of "Once Upon A Time In the West" -- and for the first time I also got audio dropouts and video jitter. I'm going to post an update (v0.57a) tonight.

Sigh... I thought I'd nailed one.

Does it mean that you exactly know what causes these problems?
If so, would you please tell us? I'm really curious about this, though I've never heard these droppouts with my player.:)

[b1ackmai1er]

Hello,

I just wanted to advise that I too am having problems with audio glitches using version 0.57. The movie uses 5.1. The glitches seem to be scattered through the whole move and are common at scene changes.

Mr jdobbs your program is fantastic. Looking forward to 0.57a and also to version 1.00.

Love to see a movie only option.

Regards b1m1

[Noah]

Just curious, but do you anticipate 0.57a being a new fix or the previous code which was better in this regard but still problematic?

BTW, I'm also having problems with 0.57 which are even visible/audible in PowerDVD unlike the typical dropout problem.

[jdobbs]

It's posted.

@Noah -- returned to previous state... almost no one got the problem then and it was only on certain players, now I think it might happen on any player. (Note: I am reporting this on faith that it has been accurately reported and was not caused by helper software -- as I've never been able to repeat the problem in previous versions).

@erdoke
I don't know for sure the "fix" was even related to the reported dropouts -- I only knew this area of code "could" cause a problem (as was proven in the fix) -- not even that it did.

[erdoke]

Originally posted by jdobbs
It's posted.

@erdoke
I don't know for sure the "fix" was even related to the reported dropouts -- I only knew this area of code "could" cause a problem (as was proven in the fix) -- not even that it did.

Well, I have no clue about programming, just find it interesting. Can we know that "this area of code" corresponds to the VOB remuxing part of RB or rather the issue relates to the rebuilt DVD structure?
(I hope this time I expressed myself correctly in English. I really don't want to seem offensive, just curious.)

[jdobbs]

It's the part that remuxes -- and builds the SCR/PTS/DTS information for the output. I'm not sure I know the difference between that and the rebuilt structure.

[Noah]

Originally posted by jdobbs
@Noah -- returned to previous state... almost no one got the problem then and it was only on certain players, now I think it might happen on any player. (Note: I am reporting this on faith that it has been accurately reported and was not caused by helper software -- as I've never been able to repeat the problem in previous versions).Well, I know if PowerDVD chokes on it, my hardware player isn't likely to do any better.

I have done loads of discs where the only previous processing done was ripping to ISO in DVD Decrypter. I've even done a few with RB-Loader which should be even less processing than that and in both cases I'll end up with dropouts on some discs. For me, that seems to eliminate that factor.

Anyway, I'm glad you're making some efforts on this problem even if 0.57 didn't quite pan out. ;)

[erdoke]

Originally posted by jdobbs
It's the part that remuxes -- and builds the SCR/PTS/DTS information for the output. I'm not sure I know the difference between that and the rebuilt structure.

It can come from my insufficient knowledge on DVD, but when I speak of remuxing I only think of building the VOB files, and the "rebuilt structure" means to me the VOBs and the correct IFOs together. :)
I thought that maybe this dropout issue can come from invalid references in IFOs. I still don't really understand why only with some players this happens and why mainly at chapter transitions.

[Noah]

Dumb question jdobbs, but before I run another encode, are you sure 0.57a is posted? Neither the archive nor the program are designated "a" and the file size of both are identical to 0.57.

[jdobbs]

You may have a caching problem. The archive is called DVD-RBV057a.zip and the program has "v0.57a" displayed on its header. I just downloaded it myself to be sure.

[jdobbs]

I still don't really understand why only with some players this happens and why mainly at chapter transitions.A lot of things can happen on the original DVD at a chapter point. Sometimes the clocks restart at zero, sometimes they reset to other times. Also, each chapter is encoded individually so its clock (within the .M2V files) begins at 00:00:00:00 -- so DVD-RB uses the clock value from the original source to replace the values in the .M2V. You also have to space the packets (video, audio, subpicture, and NAVPACKS) inserted in the stream in a manner consistent with the 1X read rate of a DVD player while considering the model buffer size. That throws a lot of variables into the equation -- especially since the new video stream is smaller in size than the original.

In 0.57 I'd decided to rely more heavily on the original stream and give greater weight to its values -- bad call, my calculations turn out to be much more accurate... so I went back to them.

[Noah]

Originally posted by jdobbs
You may have a caching problem. The archive is called DVD-RBV057a.zip and the program has "v0.57a" displayed on its header. I just downloaded it myself to be sure. Ahhh, you're right. Funny, my browser (Firefox) cached the attached file but not the page.

[erdoke]

Originally posted by jdobbs

...
In 0.57 I'd decided to rely more heavily on the original stream and give greater weight to its values -- bad call, my calculations turn out to be much more accurate... so I went back to them.

Oh, I see. Thanks for the info.

[jdobbs]

Would one of you audio drop-out guys be willing to try the test version I've loaded here. (http://forum.doom9.org/showthread.php?postid=541685#post541685) My hopes aren't high, but I made a change that could possible have an affect on audio... If I get no response I'll remove the test version.

I made the change hoping it was related to a #0003 error that was received by djan.

[Noah]

I'll give it a go. My reference drop-out source disc is out on loan, but I should be able to report back by tomorrow night.

[gizzin]

I did Gang Related With 1pass vbr w/analysis(cce v2.67.00.23). Besides the output size being only 4gigs thats besides the point. Ive done this movie over and over. I tried the test version with this one the blips or dropouts are at the same point as every other time no more no less. But there is only like 3 dropouts in this one not that it matters. So I guess no dice on this one.

[jdobbs]

Ok. No joy. Thanks.

[lamster]

Originally posted by lamster
[...] maybe I'll retry but keep all languages, and see if this makes a difference.

Didn't make a difference. Tried with 0.57a; will be checking the results of 0.57b tonight.

[jdobbs]

Download and try version 0.58...

[gizzin]

i just finished gang related, all the dropouts are still in the same spot. sorry to give you the bad news. v58

[gizzin]

do you want the bad cell sent to you are something for analysis? anything to get this fixed.

[Sir Didymus]

@gizzin

I think we should take the whole matter with lot of care, since the audio dropout problem is too relevant and I do not want to generate any expectations from such low level investigations...

Nevertheless, could you use Vobedit to do just a simple test for me ?

1. Open with Vobedit one of the VOB containing by sure some audio dropouts

2. Scroll the left panel until you get occurrencies of a sector with an audio pack, immediately followed by a sector with a navigation pack

3. Click on the audio pack and look in the topmost part of the right panel the SCR values (starting from byte 4)

4. Click on the navigation pack and look in the topmost part of the right panel the SCR values (starting from byte 4)

5. Repeating some times the steps 3 & 4, you should find couples of SCR values differing among them for less than 50.000 units

Please be very careful in using Vobedit, since it is reported it is not totally reliable in the showed results, so if you get some anomalous conditions, please double and triple check the values in order to be sure about what you see...

All the best,
SD

[gizzin]

What info you want me to post?

[Sir Didymus]

I just want to know if you are seeing couples of SCR values differing among them by less than 50.000 units...

This situation should NEVER be present for a DVD in order to be MPEG compliant...

Cheers,
SD

[gizzin]

Almost every group i looked at differed less than 50.000. Is that right? even though i didnt look at alot of groups just like 10 of them

[Sir Didymus]

Originally posted by gizzin
Almost every group i looked at differed less than 50.000. Is that right? even though i didnt look at alot of groups just like 10 of them
First of all thank you for the check. I know it could be a little disconfortable to use Vobedit, if you are not familiar with it...

Then, if you have followed precisely my indications, what you discovered are absolutely anomalous conditions: the differencies between two contiguous packs should NEVER be less than 50.000 (except for some situations, like the chapter changes, when the SCR is reset).

About the high number of occurencies of this anomalie, please follow the other thread (about the Pack Header SCR)...

I will post in a half a hour some results showing very surprising results...

If this is confirmed also by other people, it could be an useful indication to Jdobbs for ironing this timing issue.

By the way, are you a PAL user ?

Again: I suggest we'd better not to have strong expectations respect to the audio drop out troubles, let's just talk about some timing anomalies (at least for now...).

I say again: THANKS A LOT FOR YOUR CHECK.

Cheers,
SD

[gizzin]

I'm a Ntsc user

[Sir Didymus]

Originally posted by gizzin
I'm a Ntsc user

Hei, mate, thanks a lot for the support, you have been very kind...

As I told some posts before, in order to be correct, we'd better not to have specific expectations for the solution of the Audio Dropouts problem...

Nevertheless cross your finger; like many other users and testers of DVD-RB I have some positive feelings on the subject, and Jdobbs is a god in the coding, so maybe that release 0.59 will make you an happy Ntsc user ...

Cheers,
SD

[jdobbs]

Anyone tried this yet?

[lamster]

Originally posted by jdobbs
Anyone tried this yet?

If by "this" you mean 0.59, I've got it running on 2 machines now; won't know results until tomorrow.