ATI 600 USB 2.0 Calibration & Test Patterns
 

Out of curiosity, I thought I would simply do the following:

SMPTE Bars (NTSC) Test Pattern DVD [Playing In DVD Player] -> [DVD Input] DVD Playerr [S-Video Out] -> [S-Video Input] ATI USB 600 USB 2.0 [USB Out] -> [USB Input] Into computer

The question to myself was: do the colors come out correct when the test patterns are fed back into the computer from the DVD drive? And if not, why? And if not, should they? If so, how to correct it? :hmm:

The answer is apparently no! When the test patterns were played back from the DVD player into my capture card, the colors are completely off.

I'll attach images shortly.

Attached is calibrated colorbars generated by AviSynth. This is simply played directly through AviSynth and not played back by the DVD player:

Here is the SMPTE Bars (NTSC) played back through the DVD Player into the ATI USB 600 2.0 capture card:

Here are the default ATI 600 USB 2.0 proc amp settings:

Using the ATI 600 proc amp settings, I was not able to alter the colors to get them to resemble the SMPTE Bars (NTSC).

This leaves me asking a few questions:

1) Are the colors off due to my DVD player?

2) Am I passing the test patterns back into the computer properly through the capture card?

3) Why is the following greyed out in the ATI 600 USB 2.0 proc amp dialog: gamma, white balance, back-light comp, gain? Why are the 'auto' check boxes also greyed out?

4) What would be an appropriate way to calibrate the ATI 600 USB 2.0 properly to a test pattern to help ensure my colors are reading in at their proper colors?

Thoughts? :hmm:

I could help you calibrate the colors as best as possible for that source, but if the point is to calibrate for capturing VHS, you have to calibrate to the particular VHS deck and tape. Setting up an 'ideal' calibration is just a baseline and has no real meaning. The only real way to get a calibration, is if the tape itself had a recording of colorbars. If it's a home recording, you could record colorbars yourself on the same VHS deck, then calibrate off that, and assume that all your old recordings are to the same standard.

Other than that, it roughly looks like a difference between REC601 and REC709 (actually called ITU now...) color conversions, this could be due to how you captured the screenshots and so on. Could you post a clip of the capture from an mpg file instead? This gives me the raw data with no conversions.

Yes, this is my understanding too. It is 'ideal' and has no real meaning except to in this case see how I would approach adjusting whatever is needed to get the colors in line. We can assume then for this test, that it was captured from a VHS (although we know it was captured from a DVD).

But isn't there any benefits to calibrating to the test patterns? For instance, wouldn't that help me determine if my ATI proc amp settings are off? For example, what if I accidentally left my 'hue' setting to MAX in the proc amp settings when captured the video? Wouldn't I then have looked at the captured test pattern and tried to figure out why my hue was completely off?

Then, I would have adjusted my ATI proc amp 'hue' level back to the current default setting of 64. I would then believe my ATI proc amp is set to the proper levels and that I was closer to being ready to capture for real.

Or is this a bad approach?

Also, doesn't this also help me determine if the 'quality' of the video output is OK? For instance, wouldn't I be able to tell if my chroma levels are off? Or if the video is so bad, maybe I have a bad S-Video cable? And, see if there is a lot of noise in the recorded test patterns?

Ultimately, it will be yes home recordings on my VHS deck. And, also VHS movies from a camcorder.

What would be the right way to attempt to calibrate a VHS deck (for home recordings) and then a VHS camcorder? Let's also assume I never shot with color bars originally in either the home recordings or camcorder recordings? Then what?

This perhaps will also help me to try and determine how to properly calibrate my screen too. Something i'll need to think more about.

I've attached the NTSC color bars that were recorded in VirtualDub from playing the DVD through the DVD player into the capture card as explained in my first post. To keep the file size down, it's only a few frames.

*bump* :read:

Probably, yes. These are all over the place, in terms of calibration.

Again ... you could be passing the wrong colorspace here.

In what? VirtualDub? ATI CMC?

I match against Avia on output.

Were you able to un-grey these settings?

Now that I'm starting to feel functional again -- note: NOT better, yet -- I'll see about finishing that ATI 600 guide this weekend. :)