Captured Video Out of Sync

[sublimelax]

I've got a Winfast TV2000 XP Deluxe capture card. I'm using a 1.7Ghz, 256mb RDRAM, 60 gb 7200rpm hd with WinXP Pro.

Anytime I capture a long video (more than 1 hour) from my cable line, I get audio out of sync problems. For example, I tried to record the Chris Rock comedy special off of HBO the other night (strictly for time-shift purposes, no warez here). It was a 90 minute show. At about the one hour mark, I noticed the audio begin to come out of sync. By the time the show was over, it was way out of sync. This also happened when I tried to capture the Super Bowl a few months ago. Short (half hour) shows are fine. The problem only begins in long captures. The audio is also out of sync when I try to burn it to DVD and view it on a stand alone.

I've tried capturing in both mpeg 1 and mpeg 2 and I get the same errors. Is there some way to fix this?

[FulciLives]

Try to capture to an AVI file.

I would use iuVCR or TheFlyDS

Both programs allow you to sync by selecting AUDIO as the MASTER STREAM

This works fairly well.

I would use either the HuffyUV codec or the PICVideo MJPEG codec (on the 20 or 19 quality setting ... not lower than 19). Audio should be 16-bit stereo 48k PCM WAV format if you intend to use this for a DVD video.

This method works very well for me.

The only problem with this method is that you will end up with an "odd" framerate. Instead of 29.970fps you will end up with something like 29.966fps or 29.967fps etc.

So here is what you do.

After capture load the file into VirtualDubMod and check to make sure that the audio and video match. Under VIDEO (from the list of options at the top) click on FRAME RATE. Hopefully they match. If not then you can probably make it sync by using the option called, "change so video and audio durations match". If you need to do that then you have to select DIRECT STREAM COPY for the video and audio then save to a new AVI file.

OK at this point you have an AVI that is in sync but has an "odd" frame rate. I use AviSynth to fix this. A lot of people think AviSynth is hard but it can be simple. Just donwload it and install it which is simple enough to do.

Now open up the windows notepad and type in the following:

AviSource("D:\capture.avi")
AssumeFPS(29.970, true)

See that isn't so hard
Just make sure you select the correct DIR PATH and FILENAME of your captured file.
Now save this file but make sure when you save it that you change the file type from *.txt to ALL FORMATS and add .AVS to the end of the filename as you type it so you get something like looks like this ---> filename.avs

Now you have a "text" file that will work with AviSynth

Simply load the AVS file into your encoder of choice. You wuld load the AVS file just as you would load your AVI file but instead of loading the AVI file directly you are loading the AVS file ... which in turn tells the encoder to load the AVI file and then adjust the frame rate.

Some things I've noticed ...

I have both TMPGEnc and Cinema Craft Encoder. I like to use TMPGEnc just for the audio. It handles the AVS script very well. Just load the avs script for the audio and select PCM output. The audio will be re-encoded from the original to a new pcm wav file that is now corrected to match 29.970fps ... after this you can then convert it to MP2 or AC-3 or whatever other format you may need. Be sure under the advanced options or properties etc. of TMPGEnc that you link to SSRC under the audio/sound tab. This works wonders with the audio doing this kind of "conversion". THIS IS VERY IMPORTANT ... do not deviate from what I just said here otherwise the audio conversion may not work properly.

I then do the video in CCE because it is very fast and I like the quality etc. but you could do the whole thing (video + audio) in TMPGEnc.

One last thing I would like to add ... if you want to edit the capture you can do this in VirtualDubMod then put the edits into your AviSynth AVS file. Very easy to do.

Follow all the steps above up to and including the creation of your AVS file.

Now load the AVS file into VirtualDubMod.

Now do your editing.

When you are done go to TOOLS (acroos the top) and under that select SCRIPT EDITOR

Your AVS script will pop up. Go to the second line and then select IMPORT FRAMESET AS TRIMS. If this option is "greyed out" then you have to go to the option called SCRIPT HANDLEING and select AviSynth.

After you do this your script will change and look something like this:

AviSource("D:\capture.avi")
Trim(717,11507)
AssumeFPS(29.970, true)

That second line is what gets inputted into the script and contains your editing info. It might be as simple as my example or it might be very long if you did a lot of edits (such as cutting out TV commercials).

Make sure you save this change before you go on.

Now you can go ahead and do your encoding step(s).

I realize this may not be the answer you seek but the capture card that you have is best suited for AVI capture. If you follow this then your chances of having no audio sync problems is very nearly 100%

GOOD LUCK !!!

- John "FulciLives" Coleman

[SLK001]

Your problem is "timing". Use VirtualVCR, or equivalent, and use "dynamic resampling" for your audio.

[FulciLives]

Your problem is "timing". Use VirtualVCR, or equivalent, and use "dynamic resampling" for your audio.

I tried that and never got it to work for me. This method resulted in sync going in and out throughout the capture. Often it would start in sync then slip out then back in then out then in etc.

- John "FulciLives" Coleman

[SLK001]

You should give it another try. I've been doing older VHS tapes and dynamic resampling is an absolute must. I capture at full 720x480 resolution at 30fps (or, 60 fields per second). Each field should occupy only 16.7 mSecs of time, and each field will have 800 samples of audio associated with it.

Sometimes, due to inaccurate timing, other than 16.7 mSec is presented for capturing, but your card only grabs the 16.7 mSecs and either junks or pads the excess/deficiency. But the 800 samples are still associated with the video field.

Let's say that the video is continuously presented at 16.5mSecs per field (instead of 16.7mSecs), so in no time, the audio will fall behind on playback, because the video is too fast on record. Now, the video will be essentially returned to normal speed, because this is the only way the capture system knows how to do it (RE: play back your 16.5mSec field at the correct rate of 16.7mS per field). But your audio is played back at the same rate that it was recorded. With dynamic resampling, the audio will be "matched" to the field rate, by examining how many samples were actually captured during the field capture (say 790 samples), then resampling these samples (here, 790) to the proper 800 samples. Thus, your audio will speed up and slow down (usually not enough that a human can detect) in sequence with the video.

I know that this is a real simplistic explanation, but this is why people have A/V sync issues, and how capturing programs correct them.

Or something like that!

[FulciLives]

Well that is all great and fine BUT I tried it with VirtualVCR and VirtualDub Sync and neither worked for me in keeping rock solid sync when using the dynamic resampling method.

Obviously it is not a 100% solution that works with all hardware configurations etc.

My method (explained in my first post in this thread) works for me and I know of at least one other user on here that does more-or-less the same thing.

- John "FulciLives" Coleman

[warlock110]

can some one elaborate on how to load the AVS file into CCE? i'm using CCE SP version 2.5. I also have TMPGenc, i can get it to work with TMPGenc but I perfer CCE with DVD2SVCD cuzz it's much faster.

[SatStorm]

My objervation on this subject.
When you have a seperate PCI Soundcard, the dynamic resampling method virtualvcr offers do an excellent job.
When you use the built in sound of your motherboard (AC97 codec) with the latest drivers, then it mess up your captures.