VirtualDub large file size?
 

Hello there,
I'm digitizing all my family VHS recordings, and I need a little help with Virtualdub. I'm capturing to Lagarith, about 2 hours of capture is ~70GB. This is fine. But then I want to apply some filters. So I open the recording in VDub, select filters, select compression (Lagarith again), save as .avi and the result is ~200GB file. I'm applying deinterlacing filter, which gives me double the framerate (50FPS, source is PAL), but 200GB is still huge. What am I doing wrong?

After de-interlacing that's the size you would expect, Get a large hard drive or encode your captures one at a time and delete the original files as you go.

I do realize deinterlacing will increase file size, however I would expect something closer to 140GB (twice the framerate = twice the size) instead of 200. I just thought my settings are wrong. Size itself is not a problem, I have several spare HDDs I can use. Thanks.

It would be good to compare the MediaInfo output from the "before" and "after" videos.

VirtualDub is doing a low-quality deinterlace. Use QTGMC, in Avisynth or Hybrid.

Simple math, as you know. So now have 2x frames, 2x data rate minimum. Not exact, minimum.
But other variables may also exist, bloating size.

I'm not always fond of Lagarith.

Your output has been converted to RGB prior to Lagarith compression.

One way to get around this: in Video -> Color Depth... you can force input and output to YUY2.

200 / 70 = 2.85x

I opened a 142MB LAGS capture, deinterlaced at Double Rate rate, and output:

LAGS YUY2 = 200MB (1.4x)
LAGS RGB = 352MB (2.48x)

I have output set to YUY2. I'm now trying QTGMC with Hybrid, and with UT Video I got the file size down to 35GB after deinterlacing. That's half the size of original Lagarith-encoded capture. I know UT Video is lossless, but I'm a bit concerned.

Please disregard that last question, I just realized I processed wrong file. That's why it's significantly smaller. Guess i have to give them better names. But now I have another problem - audio desync. After processing with QTGMC audio is now about ~400ms ahead of video. I am able to compensate that, but it takes time, is there a solution to avoid audio desync in Hybrid/QTGMC?

I would only expect this to happen if your original capture file had inserted/dropped frames. What's your capture workflow and is something for frame TBC included?

I do not have external frame TBC. Where I live (Central/Eastern Europe) these things are almost nonexistent. Virtualdub does not report any dropped nor inserted frames during capture. Unaltered capture file has no problems with sync. After processing desync is constant throughout entire 2h video, it's about 400ms at the beginning and 400ms at the end. It can be easily corrected with any video editing software, by just moving audio track backwards in relation to video, but it's just annoying, and takes time.

Uh oh, tried another go with the same file, same settings and now sync issues are gone. No idea what is happening. One-time bug maybe? Guess the problem solved itself.