The side effects of hardware filters (screenshots)

[Brad]

I'm curious about the degree to which a TBC or another passthrough device in a capture chain irreversibly "colors" the video signal. Here's what I've found so far.

Please contribute your own reference & passthrough images if you have another device. I have attached the ISO file that you can play from DVDR.

Settings

Panasonic DMR-ES15:
Set to Lighter input + Darker output. It slightly brightens the image, which I offset using the capture device's Brightness Proc Amp control.

AVT-8710 (green + black):
Chroma is always shifted down 2px. At first I tried correcting the levels change using the built-in hardware proc amp: Brightness -1, Contrast +2 is slightly too low while changing that to Contrast +3 is blown out. So instead, I reset it to neutral settings and used Avisynth to try correcting the levels, hue, and chroma shift. I doubt the hue fix will hold up for every scene.
ColorYUV(off_y=-7,gain_y=20,cont_u=27,cont_v=26,off_v=1).ChromaShi ft(L=-1)

Toshiba D-KR4:
The following settings, followed by the AviSynth correction noted.
Brightness = 3
Contrast = 7
Color = 2
Tint = 0
Edge enhance = Off
Gamma Level = off
ColorYUV(off_y=-5,gain_y=21)

Screenshots

Reference (no hardware filter between source and capture device)


Panasonic DMR-ES15


AVT-8710


Toshiba D-KR4


Thoughts

All introduce a certain amount of Y/C crosstalk to the resolution test patterns. Mismatched S-Video cables, maybe...? But I can't say that I can see the effect in the image itself.

I couldn't find a hue fix for the AVT-8710 that didn't discolor the greyscale.

The D-KR4 gunks up high-frequency details and adds notable chroma noise.

[SixFiftyThree]

Thank you for your research, it is appreciated. There is much speculation out there in regards to hardware so I think these are good threads you're making.

I intend on making some of my own contributions here, but in the meanwhile interested to see more. Keep up the good work.

[lordsmurf]

Just remember something like the AVT-8710 has many variations.
So hardware may be perfect on one, and imperfect on another.
It's true of most VCR and TBC hardware.

Not so with DVD recorders -- those are usually the same across models or even whole generations.

[jagabo]

Another issue is that the changes introduced by devices aren't consistent over different sources. Changes to the signal from one VCR will differ from changes on the signal from another VCR, cable TV box, etc. So you can't adjust your capture card's video proc amp using, say, a DVD player, and assume that those same setting will be perfect for your VHS deck (even if the signal from the VHS deck looks identical to the output of the DVD player).

[SixFiftyThree]

Just remember something like the AVT-8710 has many variations.
So hardware may be perfect on one, and imperfect on another.
It's true of most VCR and TBC hardware.

Not so with DVD recorders -- those are usually the same across models or even whole generations.

Yep, agree with both points. My own experiences back that up.

It is still good to know about those variables of course, as an indicator of the potential quality - ideal or not so ideal. As we know, most (all?) AVT-8710 units, for instance, fall in the latter category these days.

[Brad]

I was interested in seeing how close to transparent they could get with the "best case scenario". Presumably it's downhill from here with tape sources.

I'm not seeing the posterization effect that has often been claimed of the Panasonic recorders. I would expect that to show up on the luma ramp, but all I see is added noise.

[jagabo]

I'm not seeing the posterization effect that has often been claimed of the Panasonic recorders.

I haven't seen it either.

https://forum.videohelp.com/threads/319420-Who-uses-a-DVD-recorder-as-a-line-TBC-and-wh...=1#post1982320

Posterization reports may have been from people who leave the noise reduction enable.