720X480 Was Better??!!?!?!?

[sammie]

Ok guys, I've been a firm believer in CVD for VHS captures. I understand the arguments perfectly. You can't create information that isn't there. You are getting a higher bitrate per pixel. I did a quick experiment for a friend who was getting into the ummm hobby this weekend to prove this point. Guess what capturing then encoding at the same bitrate, the file captured and encoded at 720x480 was hands down better. I was capturing several of the kids Christmas cartoons. I used VirtualDub to capture and de-interlace. Then tMPEGenc highest quality 2 pass to encode VBR with an average of around 5000 (same for both). What's up???? Am I missing something? Am I doing something wrong with the CVD capture.

[House de Kris]

Just to make sure I understand your question, you are surprised that capturing and encoding at more than twice the resolution looks better? Also, why are you deinterlacing? Is your final viewing on PC or TV?

[wway]

Sammie,

The more information you supply the the encoder, the better the final video will look. So, you are not doing anything wrong.

Someone posted a very detail reason why capturing at a higher resolution than your final resolution is better. I dont remember the thread.

For a more dramatic example, capture a few minutes of video at 352x240 and at 720x240. Encode both to vcd 352x240 at 1152 bit rate. Then check the average Q value of both mpeg files with BitRate Viewer. The file created from the 720x240 capture should have a lower average Q value. The lower the average Q value, the better the video.

wway

[sammie]

Well, the reason I'm surprised is that I've always read that VHS is only 352x240 interlaced so the max original information you could capture is 352x480. Anything beyond that is made up, doubled or interpolated as I understand it. So the argument is that you can't get any additional original information beyond 352x480. You spread the bitrate around among a smaller amount of points and you should be getting a much better picture.

Primary target is a TV. I was de-interlacing because there were some clear interlace artifacts in the 720x480 capture visible on the tv. I presumed due to the higher capture size because I've never noticed them before. I de-interlaced both to be fair in the comparison. De-interlacing didn't seem to make much difference either way for the 352x480, except in the qulity of the paused picture of course.

I'll have to check your Q value wway. See what is going on there.