Advanced Calibration and Meaurement

[jmac698]

Just thought I'd point out my latest experiment.
http://forum.doom9.org/showthread.ph...82#post1584482

The short story, I've created a test disk to measure and calibrate a VCR. I've also made an Avisynth plugin & script to analyze the results.

Right now, it can calibrate levels, analyze non-linearity (a little bit), and frequency response.
What could be done with these results in the future, is to adjust sharpness to neutral and suggest the ideal capture settings in terms of brightness and contrast. It turns out that using greybars or colorsbar to set black and white is NOT the best way to do it!

http://screenshotcomparison.com/comparison/137533

Information you could find out in the future includes, capture window (for setting proper aspect rating and known how much cropping you are losing off the sides), fixing non-linearity (would adjust midtones levels to proper amount), measuring jitter.

Some things you could test:
edit/normal switches on VCR
sharpness/color/level errors of TBC passthrough
TBC accuracy
dropped frames

These tests will be supported by the half dozen plugins I've been writing one at a time for the past few weeks. There's a LOT missing from Avisynth, but once I add some functions, it should be much easier to do calibration work.

[juhok]

So you're calibrating the capture device, not the VCR itself, right?

[jmac698]

Hmm.. both affect the signal, so I can measure and adjust for any distortions. But I think I would take calibrating a VCR to mean adjusting something on the VCR, which might mean taking it apart and turning some pots on the circuit board, which is probably a normal procedure in the factory or repair.

I think I learned something quite interesting today. I noticed the peak on my test of VCR frequency response. Next, I looked at the datasheet for a head amplifier. Three things I've learned;

1 I thought I could connect a digitzer straight to the heads and somehow get a better quality signal. In fact, the single chip head amplifier directly outputs the VCR signal, which means it demodulates the FM signal. So nothing as simple as touching a digital oscilloscope to a couple pins. Also, changing out this one chip for a better one might give less noise or different sharpness.
2 Some issues with layout can cause the peaking I've observed. So I know my VCR might not be good quality. I also know it can be fixed by soldering a single part. Or I can do this in software.
3 The head amplifier directly combines Y and C signals, so in fact having s-video in is pointless I want to check this with a few more chips. I've always wondered this; first I read that the "color-under" method means the signals are separate, but instead they are just separated by the amplifer. So I want to do more research.

[juhok]

Yup "measure and calibrate a VCR" 99% of the time means calibrating the unit to spec using factory test tapes.

IIRC your VCR is not in the list of recommended units for transfer and this all started to work around some of the limitations your hardware workflow has. I applaud the hard work and research and at the same time I'd like to remind the occasional reader that statements like "having s-video in is pointless" really might only apply to limited range of hardware which is not recommended for transfer to begin with.

[jmac698]

I wasn't clear, the datasheet I'm looking at doesn't apply to my VCR, however the datasheet does mention a general problem with "peaking", which I've also observed in my VCR.
The datasheet is here
http://www.datasheetcatalog.org/data...nics/mXqww.pdf

Quote:

V.2.b - Mixing Y and C input signals
The input of the record amplifiers can be considered
as a virtual ground. Consequently the input
signal VIN can be either a composite signal in which
Y and C have been already mixed, or separated
and Y and C signals. In this last case the two signals
will be added one to the other by the record amplifier.

Therefore, s-video in is useless. If you can find a chip which truly keeps Y/C separate, let me know. I've looked at a half dozen so far, and none of them do so.

I've wondered about this for a while, and I've tested my (recommended) JVC S-VHS outputs, both composite and s-video, and there was a slight tint change but hardly a noticeable difference.

Does anyone have a schematic for a hi-end VCR to check?

[juhok]

I lack the VCR electrical knowhow to validate statements on chips used. So no comment on that.

I got a bit derailed by the S-Video IN (missed the IN -part). Personally never really was in the business of recording tapes using higher-end gear. S-video OUT is what matters for most of the people doing analog to digital. Also, not seeing composite artifacts on most home video recordings I've transferred argues for seperated signal to be common, no?

I've uploaded many service manuals for high-end VCRs in here : http://www.digitalfaq.com/forum/video-restore/

[jmac698]

I tracing it to an AN443CS head amp.
http://pdf1.alldatasheet.net/datashe.../AN3341SC.html

In this chip, finally we see separate Y/C in/out. Interally, two amplifiers are fed from one head input for playback, and for recording, the two signals go to a block called mix, then to a head.
It's not stated anywhere, but this could be where the "color-under" system is applied.
It seems that a simple composite signal could be recorded directly as the luma as well.
I've also collected a sample where a VCR had dot-crawl artefacts. Maybe it was from cross-talk in the circuit-board? Yes I was wondering about that too.
One theory is that the color resolution is too low to have dot crawl.
Maybe they are always separated, but the cheaper chip does a notch filter to separate internally, then does color-under.

[juhok]

If the signal is composite, poor color resolution propably doesn't fully explain lack of artifacts because I've had them using as bad or even worse (composite) sources than VHS.