Hi,
I am converting AVI to NTSC DVD and experiencing sync problem.
I demuxed the avi in VitualDubMod and got Audio.ac3 file.
I encoded AVI video with Procoder to 25 fps (PAL) and then from 25 > 29.97 NTSC with DGPull down to avoid audio sync problems.
But when I multeplex the audio and video (after pulldown) with either Rejig or Muxman, the audio in the resulting VOBs is desynchronized.
I played the avi with DVDShrink without buring onto DVD disk. At the begining of themovie the audio is 3-5 secs behind the video but at the end of the movie it gets worse - it is about 35 to 45 secs behind.
Why is it happening? The original if fine. Is it possible to fix the problem?
+ Reply to Thread
Results 1 to 30 of 32
-
-
The same problem with two other AVIs. In the resulting VOBs (from both AVIs), the audio is heavily desychronized.
May be somebody experienced the same problem? Please share your experience.
I am not sure whether it is important or not, all three AVIs have one common feature: they were created from HD TV recording.
Thank you! -
It's happening because you demuxed the audio and left it in it's original form. Why did you convert the AVI to 25fps then convert again? What was the original framerate?
He's a liar and a murderer, and I say that with all due respect. -
Thank you.
"Why did you convert the AVI to 25fps then convert again?" - I am a newbie and could some things wrong, no doubt.
I read on Domm9 forum about the conversion first to 25 then to 29.97 with pulldown to avoid audio sync problems and interlacing after encoding.
This way to convert was also discussed here. Sorry, I do not have a link handy.
Do I undestand you right you suggest I should convert directly to 29.97? Will the movie length in the reulsting file be the same as in the source? -
What was the original framerate?
He's a liar and a murderer, and I say that with all due respect. -
Hi-
In the other thread you asked me to have a look at this one, but it looks to me like you already have some good help. And you haven't yet supplied near enough information. When trying to diagnose progressive audio asynch, you have to know if the audio changed length before and after, and if the video changed length before and after.
I assume you know the length of the AVI (GSpot or VDubMod). That will give you the initial length of both the audio and video. Then you demuxed it. I don't think you checked the length of the audio ready for muxing (I use HeadAC3he, although I guess GSpot can do it). The audio probably isn't the problem (especially since it's AC3), but you have to cover all the bases.
Most likely the video changed length. That's why they're trying to get the initial framerate out of you. If, for example, it's 23.976fps, but you ran pulldown afterwards for 25->29.97fps, you've speeded it up from 23.976fps to 25fps, and that could easily account for the audio lagging the video when authored.
It can be tricky to get the final length of an M2V or Vob. You can open it in VDubMod or VDubMPEG2, but the surest way is to make a fresh D2V and stick that into an AviSynth script and open it in VDubMod. Whatever you do, don't count on a player to give you the correct length. If had to guess on the basis of very incomplete information, I'd say you've speeded up the video, and it's now shorter than when you started.
I played the avi with DVDShrink without buring onto DVD disk.
Don't you mean to say you played the DVD in Shrink, and not the AVI? Have you played a vob using maybe MPC, or played the DVD using PowerDVD or some such? I don't know if it'll be any different than what Shrink shows you, but it's worth a shot. -
manono,
Thank you.
You could be right about the good help but I did supply what I was asked to supply: the initial frame rate. I am ready to provide any information but I do not know what I should provide.
If I missed something, I regret it. It is not because I do not want to do this or ignore the advice. It is just for the simple reason that I do not understand many things.
"I assume you know the length of the AVI" - Honestly I even did not know that I am supposed to check it. I never did.
" I don't think you checked the length of the audio ready for muxing" - You are right: I neve did it.
"If, for example, it's 23.976fps, but you ran pulldown afterwards for 25->29.97fps, you've speeded it up from 23.976fps to 25fps, and that could easily account for the audio lagging the video when authored." - Most likely, that's what's happened!
"Most likely the video changed length. That's why they're trying to get the initial framerate out of you. If, for example, it's 23.976fps, but you ran pulldown afterwards for 25->29.97fps, you've speeded it up from 23.976fps to 25fps, and that could easily account for the audio lagging the video when authored." - Thank you! But could you please explain - what should I do instead of what I did.
As for DVDShrink, you are right: I played DVD, not AVI. But I can hardly imagine there would be any difference. I played AVI in MCP and it was fine. -
OK, let's try it again.
1. What's the framerate (fps) and length (in minec) of the AVI?
2. What's the length of the audio after being demuxed (minec)?
3. What's the length of the M2V after coming out of DGPulldown 25->29.97fps (minec)? And how have you determined the length?
These are all things you need to know how to find out when trying to diagnose problems of this type. -
manono,
Thank you!
What you are telling me is close to step by step procedure - exactly what I need.
Length of AVI: 2:03:32 (GSpot)
Frame rate: 704 X 384 (GSpot)
"What's the length of the audio after being demuxed (minec)?" - GSpot does not show this. Sorry, I do not know how to determine this.
"What's the length of the M2V after coming out of DGPulldown 25->29.97fps (minec)?" - It probably sounds weird but GSpot shows 3 seconds. I do not know why. May be GSpot is a wrong tool to determine this?
I would appreciate if you could please explain how I can use the specs to handle the dysincronization problem.
Thank you! -
If you had read my first post in this thread, you would have found how to determine the answers to the questions I asked.
Sometimes teaching you things isn't very rewarding for us as we have to keep telling you the same things over and over. It seems to go in one ear and out the other. steveryan asked the framerate and you gave the resolution. guns1inger explained framerate and even gave you the 3 most likely answers from which to choose. I asked for the framerate. Again you gave the resolution.
OK, the framerate of the original AVI is 23.976fps. What made you think that applying DGPulldown 25->29.97fps was the correct thing to do? What made you think this was somehow a PAL2NTSC conversion? Take that M2V (the same one you used DGPulldown on already) and run it through DGPulldown again, this time for the default 23.976->29.97. Then mux/author again. There's a chance the audio will be slightly off (let's hope not), but it can be easily fixed by adjusting the delay. You shouldn't have progressive asynch, where it gets worse as the video plays along. You created the progressive asynch by applying DGPulldown incorrectly, for 25->29.97fps. The effect was to speed up the video from its original framerate, to make it play at 25fps, and not 23.976fps, to play faster and end sooner. Had you followed my instructions and figured out the length of the M2V, you could easily have come to the same conclusion. Let's hope it works correctly this time and winds up in synch. -
I bet you are muxing video 25FPS with audio 23.976FPS, you must convert the audio to match video framerate
If you don't know how to read GSpot video information, use VirtualDub instead. Just open your video file and then File Information (the framerate is the first line )
use Besweet to convert audio to correct framerate
like this :
besweet -core( -input MYAUDIO.AC3 -output AUDIO_FIXED.AC3 ) -ota( -fs 48000 -r 960 1001 ) -ac3enc( -b 256 ) -
If we go all the way back to very beginning of this saga (some 3 or 4 very longs posts ago), this began as an exercise to convert PAL to NTSC, then to sharpen the image on the way.
Now it transpires that it was NTSC all along . . . . . . .
Stop now.
Go back to the beginning and re-encode your video as 23.976 720 x 480, with 2:3 pulldown. Use LSF or aSharp or whatever works for you. You will have no audio problems.Read my blog here.
-
You think it's the very same AVI we've been discussing for a week or so now? I hope not.
-
Guys,
Thank you for your patience and advice.
I apologize for missing some things that have been already discussed. The volume of information and the fact that is mostly new for me makes it a little overwhelming.
I do many things without understanding how it works and this leads to problems I have.
All of a sudden in the middle of the night I relaized that I could try 23.976 720 x 480, then use pulldown to 29.97. In the morning, I read guns1inger's advice about it...
My understanding (wrong) was that Procoder will change avi to 25 then to 29.97.
Thank Delta2 for advising another way to fix the problem. -
Hello everybody,
I enquired about the audio synchronization problem some time ago. That time, I was making a gross mistake during encoding AVI file to DVD NTSC but this time, I believe, I am doing it right but still have the problem.
I am converting AVI to NTSC DVD and applying the audio the following way:
- I demux the audio from the AVI file with VitualDubMod and get Audio.ac3 file.
- I check the AC3 file with AC3FIX;
- I disable the audio in AVI and create a copy of the AVI without audio;
- I rencode AVI video with Procoder to 23.976p -> NTSC (3-2pulldown). The AVI original frame rate is 23.976.
- After that I use DGPpulldown to change 23.976p -> NTSC (3-2pulldown) to 29.976;
- Then I use TSUNAMI-MPEG DVD Author Pro to merge the video and the audio streams.
In three cases out of four I experience audio synchronization problem.
The audio usually matches the beginning of the video but then it lacks behind the video. If I try to adjust the audio to the video, as result I get a movie with the audio being ahead of the video at the beginning; the audio perfectly matching the video in the middle; and the audio lacking behind at the end.
It looks like I am doing something wrong but what?
Thank you! -
because you encoded as 23.976 fps with pulldown in ProCoder, the DGPulldown step was unnecessary. That aside, I can't see anything in your process that would cause things to go out of sync.
If you open the ac3 file and the m2v file in virtualdubmod, what running times do you get ?Read my blog here.
-
I have opened the m2v file in virtualdubmod but I do not know where I should look for the running time.
It looks like I only have the file that I have got after I applied DGpulldown. I have deleted the one that I have got after encoding with Procoder.
If I understood you right, I do not need DGpulldown step. But TSUNAMI-MPEG Authoring Tool and Muxman do not accept m2v file directly after encoding to 23.976 with Procoder. -
virtualdubmod crashes after I open the m2v file in it and click "File Information".
As for ac3, it looks like I do not know how to open it in virtualdubmod. When I try to open it directly in virtualdubmod, it givesw an error message.
I'll encode it again and try the resulting m2v file that I receive after use of 3-2 pull down. -
Hi,
I have encoded the AVI again with Procoder to 23.976p -> NTSC (3-2pulldown).
After that I checked the length of both m2v files (the one that I just created and the first one on which I used DGpulldown after Procoder) with GSpot: both have the same length of 1:48:18.
Edited: I've found the AC3 runtime through ImTOO MPEG Encoder (I do not know how to read it in GSpot). It is 1:48:17. The program shows the runtime of the video (demuxed, no sound) also as 1:48:17.
Before encoding, I cut off a part of the titles from the original AVI and demuxed the audio after that. Could it be because of that?
Thanks! -
It looks like something is globally wrong in what I am doing.
I have encoded another AVI. This time the source is 25 fps.
- I demuxed the audio (MP3) and converted it to AC3 with ImTOO MPEG Encoder.
- I have checked the AC3 with AC3Fix
- I have encoded AVI (with audio disabled) to 25 fps with Procoder
- Used DGPulldown for 25 to 29.97.
The runtime of both the video and the audio (AC3) is exactly the same.
When I try to mux the video with the AC3 with TSUNAMI-MPEG Encoder, I have the same synchronization problem as before: the video and audio are synchronized at the beginning but progressively desynchronized further on.
I am totally confused!!! Please help. Any ideas would be appreciated. -
Please help!
I believe it depends on the way the audio is extracted and encoded as every time I changed it, the resulting audio also changes: it is either behind or before the video but I cannot handle it.
If any information is required I am ready to submit it,
Thank you!
Similar Threads
-
questions about audio normalizing and audio encoding
By codemaster in forum DVD RippingReplies: 2Last Post: 25th Sep 2011, 21:22 -
AVStoDVD crashed on audio encoding.. have m2v video file, but no audio
By arcooke in forum Newbie / General discussionsReplies: 3Last Post: 5th Aug 2010, 14:18 -
encoding with DTS audio stream - Audio not found in AutoGK
By extrabyte in forum DVD RippingReplies: 1Last Post: 7th Jan 2009, 16:42 -
best audio encoding?
By Remyisme in forum Newbie / General discussionsReplies: 4Last Post: 20th Oct 2008, 04:44 -
ffmpeg -- encoding while muxing new audio w/out source's audio, how ?
By vhelp in forum AudioReplies: 8Last Post: 21st Nov 2007, 19:53