Understanding VirtualDub (editing video and audio)

[27am_dot_com]

Just looking to wrap up this DVD project but I've run into a wall.

In short, I have about 2 dozen AVI files that I will want to join (at the same time, chop first few seconds or last few seconds etc for editing purposes). Then I'd like to extrac the WAV to work with it all from there.

I was told to use VirtualDub and am surprised how well it seemd to work. Very easy to Append AVI and easy to edit (chop) where needed.

However (and this is my challenge), I must be missing some logic on how to work with the finished file. Let me give you an example. Right now I have 13 of the files joined and chopped and ready to be saved as one new full AVI.

The 13 files run approx 24 minutes and 665 mb. Yet when I save in VirtualDub, the resulting file is 45 GB!!! What the??

But I don't want to compress in here (or do I?) as I was told that TMPGenc delivers a better video compression and I want to maintain video quality without fuzziness etc.

Am I supposed to compress? How do I get these 13 files to remain around a workable file size (not 45 GB.. I ended up aborting half way through the process by the way)

[27am_dot_com]

Also, adding to my Virtual Dub questions... I ran a test with just 2 AVI files being combined and then resaved. What's interesting (and confusing) is that when I double clicked the combined AVI file, it automatically opened with Windows Media Player (my default player) and it didn't work... it was distorted. I thought it was a corrupt file. But then I opened it with 3 other players and it runs just fine.

Should this concern me?

After I compress it into MPEG-2 it will run fine on a DVD either way right?

Any specific reason why Windows Media Player can't play my combined video??

[yoda313]

Hi,

There are other avi codecs besides divx/xvid that will be more or less uncompressed but not that massive. Try something like huffy or someother avi codec. It still has to know what to compress it to. That large of a size is the totally uncopressed nature.

Kevin

[27am_dot_com]

Last VDub question... when wanting to append the 14th video file to my first 13, I received a message saying it can't be done due to different sampling rates. I saw a similar question on this forum regarding audio but I didn't understand the reply to well and am not sure if the answer is the same for working with videos.

Here's what's confusing: all of the video segments were recorded with the exact same setting (I didn't tinker with anything) so why would they have different "sampling rates" of 15.0000 vs 8.00000??

Thank you.

[junkmalle]

If you're not using any filters in VirtualDub use Video -> Direct Stream Copy mode. That way it won't decompress and recompress your video. (Actually you probably saved as uncompressed RGB, hence the very large file.)

[Poplar]

VirtualDub is mainly used as a linear video editor and a frame server. Under certain circumstances, it can be used as an AVI joiner, i.e., all the files have the same compression, sampling rates, FPS, etc. I am not a fan of joining video segments. What can you gain? Make a play list and they will play like one.

[27am_dot_com]

Kevin... so how do I convert them to a different codec? I'm as new at this as they come. They are Camtasia recordings of one of our presentations.

Junkmalle... I'm sure I'm not using any filters as I don't really even understand the software. But I selected Direct Stream as you suggested. What now? Do I still have to save it or what's the next step?

Poplar... do you recommend a different software for video joining? Essentially, as I mentioned, they are Camtasia recordings of a presentations. Some have startup microphone noise. Some need a cut off at the end. ANd all, eventually, need to be joined so that I can extract the full audio (speech). Why do I need the full audio? Because I need to interleave background music behind it.

I'm going with the understanding that TMPEGenc is the best video compressor (or one of the best) but I know it's poor for audio compression. So I need separate videos and separate audios to work with. If I should be looking at an alternative route, please let me know.

[junkmalle]

Junkmalle... I'm sure I'm not using any filters as I don't really even understand the software. But I selected Direct Stream as you suggested. What now? Do I still have to save it or what's the next step?

Make all your edits and Save your output. With Direct Stream Copy set VirtualDub will only cut and paste. There will be no decompression or recompression so you'll lose no video quality. And the file size won't change except of course for the bits you cut out.

[manono]

Hi-

As mentioned before, you saved them as uncompressed. If you were to have set Video to Direct Stream Copy, and then save, the time to save would only have been a minute or two, and the final size would have been about the sum total of all the original sizes.

As for the sampling rate error, evidently VDub thinks that final audio file has a slightly different sampling rate from the rest. It may or may not be true, but you still won't be able to Append it. About all I can suggest is to extract that final audio from the AVI, and then join it to the other audio when you get it into your Audio editing program, some kind of a WAV Edior, I assume. You may be able to get it to Append to the rest in AVIMuxGUI, but I wouldn't try with that huge uncompressed AVI you have now.

If you want to Direct Stream them, just Append as before, set Video for Direct Stream Copy, and then Save under a new name.

Edit: junkmalle beat me.

[junkmalle]

By the way 27am_dot_com, If you use Direct Stream Copy you can only cut on key frames. If your source files are something like Divx there may be 300 or more frames between key frames. If your source is DV, every frame is a key frame.

[Nelson37]

under environmental settings in TMPgenc, you can specify TooLame for audio encoding and SSRC for frequency conversion. Very important you obtain these two free tools and specify their use in TMPGenc. Makes the audio encoding as good as any you will find.

[27am_dot_com]

You guys have been of really great help. The video editing/chopping/joining is now child's play. However, I logged back onto the forum to ask a few more questions about audio converions.

Nelson37, are you suggesting that I don't need to do anything outside of TMPEGnc... just have the two programs you mentioned installed and then bring in my AVI files and my WAV files (or even the AVI with the WAV still interleaved) and then have TMPGEnc sort it all out?

Don't I need to compress the audio to AC3 format? I guess I understood that AC3 is the best quality and the smallest file for my DVD.

Really my only reason for extracting the WAV was to 1) the interleave background music 2) be able to compress into AC3 for smaller file but top audio quality.

[Nelson37]

For Ac3, you can use BeSweet or one of the many GUI for it. SFAIK, almost all freeware tools use the same basic ac3 encoder.

Frequency conversion is something you need to be careful to avoid, few progs do this well. This would be the 44 to 48 audio frequency.

MPG audio will work in most players but is not to spec, at least in the US.

Filesizes roughly comparable.

[27am_dot_com]

Nelson37, I think I'll take your original suggestion.

I even need to resample an audio from 8.000 to 15.000 and think TooLame could do this.

So I downloaded TooLame and then I grabbed SSRC. But SSRC was in a .tar file which I'd never seen before. So I searched for it and grabbed some GNU file which was in .gz format. Now I had to look up .gz and I guess I use that gz uncompressor to then open the gz file so that I can grab the .tar compressor and open the SSRC file.

(man am I tired of these abbreviations)

But how do I get TMPGenc to "call upon" these programs now?

[27am_dot_com]

Actually it looks like SSRC has 4 different programs to download on its page... which do you suggest? And it's SSRC where I read that it can resample audio (not TooLame) but that file seems to download as a corrupted .tar file.

Any pointers?

[Nelson37]

Options, Environmental settings, external tool.

Been a long time since I found SSRC.exe, but that's the one you want.
Many freq converters will add a hollow, metallic sound to the audio. SSRC generally regarded as the best.

TooLame is just the MPG audio encoder, but seems to work better than the one built-in to TMPGenc. Not as noticeable a change as when using SSRC.

That sampling rate difference is odd, is the last one perhaps mono rather than stereo?

I always edit my AVI, then strip the audio, then convert. Besweet will also integrate SSRC, using it for freq convert, and then encode to AC3.
Then I encode video seperately and mux. New version of TMPGenc with AC-3 plug-in will convert to ac3, but I have not used this.

Infact the besweet package may be your best bet as I think it includes all needed progs, even if you don't use it you can get the SSRC this way.

Stands for Shibatch Sampling Rate Converter.

[27am_dot_com]

The sampling might be mono... but I'm just as confused. In fact, I just finished all of the recordings and out of 18 files, 2 turned out with 8.000 rates instead of 15.0000. So maybe most are mono and 2 are stereo? I'm lost already. All I know is my settings never changed during recording.

But thanks for the suggestions. I have lots of software to toy around with. I'll get this done yet.

[Safesurfer]

You might also want to look into frameserving the video portion from VDub to TMPGEnc, that way you don't have to create an AVI first. This guide might help explain how to go about it:-

https://www.videohelp.com/forum/userguides/87270.php