Chromatic "burn-in" with AG-1980 VCR
 

I've owned a late 2001-built AG-1980 for a few years at this point-- in excellent shape with all its' innards in good standing- but there's an issue that I have never noticed until several weeks ago came up as I was reviewing footage frame-by-frame while running the QTGMC deinterlacer.

When a drastic color change occurs between frames or scenes, the last frame of the old scene's colors tend to "bleed" into the next few- it almost looks like a bad telecine or the smearing on a CRT/tube camera, but this problem is prevalent in every tape I test with it, filmed, videotaped 60i footage or otherwise- this is not an issue with the de-interlacer as running footage from other VCRs turns out fine, and these other VCRs I have tested (e.g. a HM-DH30000U, HR,6700U, PV-1530 and an HR-S5902U) all do not present this quirk.

Here are some comparisons- the first frame of a new "scene" vs. a few frames after.



( There were some images embedded here, but that wasn't good enough and the server whined about it, so here's a link. https://imgur.com/a/6wr8iTf )



I know they aren't the best examples but it's from the files I had on hand at the time. You can make out details from the scene that came prior in the first example (the entrance to a building) and in the second, the black frames prior to the iris effect bleed into the color image that comes after causing the first few frames to look quite de-saturated.

I have replaced all of the SMD capacitors with new ones on the Y/C board, and while it made some "rainbow" noise issues with badly duplicated or poor SLP tapes go away, this did not fix the "burn-in" effect present on all tapes in all modes. I'm curious if the tantalum/ceramic capacitors could be the source or if it might be the power board itself- but regardless most importantly I want to know if this is a common AG-1980 issue as I see it in videos on Youtube often but nobody really speaks of this problem.

What are you capturing with? If it outputs in some kind of compressed format you may see chroma subsampling.

It's a KONA AJA LHe+ capturing uncompressed 10-bit RGB (overkill, yes, but the software doesn't allow for 10-bit uncompressed YUV anymore and the card has done me well).

The issue does not subside with YUV or other formats present and/or with different capture cards as it's innate to the AG-1980 itself. There are five boards that I think could contribute to the issue; the power, Y/C, main, head amp and possibly the RF board (there's a Sony daughterboard on it which contains 10-15 or so surface-mounted capacitors which may or may not be interfering). I've heard there's another SMD hidden away on a "display" board somewhere which could potentially cause dim VFD output on earlier models, but upon opening up the front panels and checking both boards I could not find it- possibly a revision made between the earliest '95 models and my '01.

To further explain this issue I can present several videos that show the same chroma "delay" problem I'm having. (By using the . and , keys while paused, frames can be advanced to the exact points where cuts occur.)

https://www.youtube.com/watch?v=CMz_wM5UhUU This is a comparison between an unserviced AG-1980 and a JVC HR-S9900U using animated footage- the JVC "snaps" better between cuts though in some cases (i.e. the close-up on Belle's face to the shot of bird-feeding) both have that drastic "frame bleed"- the JVC resolving its' chromatic noise a bit faster.

https://www.youtube.com/watch?v=faGajzZrMq8 A capture of a 1986 HBO commercial break. Look at the movie promos and see how the frames after each cut get quite noisy and colorful.

https://www.youtube.com/watch?v=nuDAsSsYhc8 Some 60i footage to show that this issue is not prevalent on just 3:2 pulldown filmed footage. While most cuts here have fade transitions, the ones that do not (e.g. the black screen of a DOS computer to the white computer itself) show that same "bleed" issue for a few frames.

This chroma "ghosting" is a byproduct of the noise reduction system in the VCR. The system uses temporal noise reduction techniques that may cause ghosting.

Hm, two places I can think of where such a problem would occur would be in crosstalk elimination (video heads pick up a bit of color signal from adjacent video tracks, and VCRs have some circuitry to cancel it out.), or in the noise reduction bit.

Does it happen no matter what the picture quality switches are set to?

Yup. Expected behavior, not a fault with your particular unit.

http://www.digitalfaq.com/forum/vide...lor-smear.html

Moving the slider to EDIT should disable the DNR.

It's curious that you don't notice this artifact with your JVCs featuring DNR, though.

The JVC DigiPure units definitely have it, as do the late model Sony Hi-8 and Digital 8 camcorders with DNR.

It used to be my main VCR for some time, I switched off every single "feature" in the main menu- all the digital gubbins, the 3DNR, and from "NORMAL" to "EDIT" as always with the '00s JVC VCRs- and surprisingly it didn't really have that ghosting problem (it had other noise issues and terrible normal audio, but you can get a sense of the frame ghosting with this clip).

I'll try ingesting with the EDIT switch on but from what I've seen it adds quite a bit of noise/patterns in some cases, otherwise it's worked fine- I haven't tested that mode since I recapped the board but I ought to try it once more. Thanks again.