Luminance clipping in capture Pt1: PAL VCR comparision
 

I hope everyone understands what I am writing here because I used an online translator.

I have spoken about it several times in some forum posts that some video recorders clip the luminance signal during playback.
I have video tapes recorded from 1986-1995 with a Siemens FM mono recorder or Sanyo recorder where the brightness levels are too high (blown out whites).

Possible reasons could be:
-too high Video signal from satellite receiver, antenna signal
-the recording videorecorder (sharpness settings?...)
-cheap, bad video tapes
-Copy of a videotape
-video amplifier etc...

To capture VHS video we need at least 2 (3) devices. VCR & capture card (and for better quality TBC or passthrough DVD-recorder).

Any device can cut the brightness. Here in the forum it is often reported that some DVD recorders & capture cards cut off the brightness or other users ask why.

Does the clipping happen with every cassette? NO!!! :)

Only in those where the brightness was too high when recording.

Part 1: The brightness clipping on some PAL video recorders.

Test setup:

-same s-video or composite cable for dubbing
-same video cassette
-capture card Canopus NX (this has a real proc-amp to adjust the input levels, this card can also record the full area including super black and super white)
-all captures were adjusted to made in the 16-235 range

-13 different SVHS video recorders (sorted by release date)

-JVC BR-S811E (1989)
-Telefunken A1200 (1989 JVC HR-S5000 clone)
-Nordmende V6400S (1992 JVC HR-S4700 clone)
-Blaupunkt RTV925 (1992 Panasonic FS-88 clone)
-Blaupunkt RTV950 (with TBC) (1992 Panasonic FS200 clone)
-Blaupunkt RTV966 (with TBC) (1994 Panasonic HS1000 clone)
-Hitachi S390E (1994)
-JVC HR-S7600 (with TBC) (1999)
-Philips VR1500 (with TBC) (1999 similar to JVC HR-S8600)
-Sharp VC-S2000 (with TBC) (2000)
-Panasonic NV-HS860 (with TBC) (2000)
-JVC HR-S5950 (2002)
-JVC HR-S8960 (with TBC) (2003)

2 different VHS video recorders:
-Panasonic NV-HD650 (1995)
-Thomson VPH6990 (1999)

Attachment 11869

Here is a picture with a good capture without clipping on the left and a bad capture with clipping on the right.

(Note the font WTA-Tour Essen, can you read it?):D

Attachment 11870

including the display in the waveform monitor. On the right side (bad recording) you can see at the white line (at +100) how the recording is clipped when playing in the video recorder.

It doesn't matter whether you switch the TBC on or off for video recorders with TBC:

Here is a lossless sample from the Nordmende V6400S:
Attachment 11872

Here is a lossless sample from the JVC HR-S8960:
Attachment 11871

To save storage space, all recordings (some double with and without TBC) saved as a split screen and as an mpg file.
Attachment 11873

Should anyone want to see another lossless sample. Let me know and I will upload it.

Maybe now some people see their captures from another side.:question:
This is one of the reasons why you should have a second video recorder when you have a large amount of videotapes to capture. Not every video recorder plays every cassette well.

There is no way to restore the clipped area in post-processing if the video recorder already cuts off the brightness. :depressed:

Part II would be the capture cards if there is interest.

On the JVC's you may have gotten some settings wrong, They do have some picture settings, I think it should be on "Edit" for capturing.

No wrong settings.
All the JVC recordings are made with edit mode.
I although have made captures with B.E.S.T on-off, with Picture mode auto, edit, sharp, soft with tbc on-off with video stabilizer on-off. The luminance is cut off. The philips which is a jvc perform the same way. :depressed:

Would like to see the capture cards used.

Unless you have a well calibrated studio monitor that's not how you judge VCR's, Capture cards react differently to each VCR. Also VCR's electronics need adjustments every once a while, As components age values drift.

So quickly looking at the capture it's mainly the JVCs (incl philips), Sharp and HS860 (last one only when TBC is on) that are clipping?

Looks like the Sharp uses the same JCP8017 video chip that's used in the HR-Sx600 JVCs (and also hi-fi models from the same lineup). It's not a JVC clone though as it's otherwise quire different.

On the more modern VCRs (in this case then 1999+) models, there is very little that can be adjusted. On the older ones, e.g FS200, HS1000 you can adjust the video level in/out of TBC, and a bunch of other things with trimpots, but they dont' seem to be clipping anyhow. On the modern jvcs you can adjust a few things with a service remote but I don't know if there is anything that could help avoid clipping.

If there is an A/D conversion as part of the video chain in the VCR that is one point where clipping can happen, though it does seem to happen even with the TBC off. I don't know whether the video go through the digital processing regardless or not. It could also happen in the demodulation stage, but I think that would cause other artifacts too.

Have you tested if the same happens with JVC hi-fi models from the same era? (Like JVC HR-J668, 680 and similar) Most of them have the same main video/audio chips as the SVHS models, but most lack any of the extra digital processing of their bigger brothers.

Is there a good tape to test for luma clipping? "VHS Video Essentials"?

do you know how to use capture software and histograms?
Believe you me, it really helps to consult capturing guides and find out what you're doing. The capture wheel was invented decades ago and has been in use ever since.

..................

Calibration tapes are essential for repairing and adjusting VCR's to factory spec especially if a recap job has been done or the VCR hasn't seen an adjustment for decades, Histograms don't tel you $hit about the tape itself if your VCR is wrong to begin with.

This seems like an interesting thread (one that I've not read yet).

However, behave, please do not derail it. :2cents:

While latreche34 is correct, where special test tapes are required for factory spec comparisons, sometimes the obvious unavailability of said tapes will require a test bed of personal tapes. However, it must always be remembered that the recording values of said tapes can also cause variations on playback. Same for VCR condition, capture card AGC/etc response, etc. So what a test like this can definitely state is "for my tapes, and with my hardware workflow", this is the results. However, it may not always be the case where you can definitely state that XYZ is always true. That takes process of elimination. Science, constants and variables.

I would like to see this test with another capture card, namely ATI AIW, given that it has proven itself to have some of the best levels.

Interesting thread, Bogilein. I appreciate it. :)

as requested by lord smurf a comparison of AIW 9600 AGP vs Canopus nx as a splitscreen mpg file.

Players used Blaupunkt RTV950, Hitachi S390, JVCHR-S8960, Nordmende V6400S, Philips VR1500, Thomson VPH6990.

A short note about the Thomson, since it is a VHS player it's connected via composite cable. I tried to adjust the level during the recording, no chance with the AIW.

If someone needs the lossless capture or want to see what the files look like under the waveform monitor, let me know and I will upload them.

The very first capture, RTV 950 (NV-FS200 clone) on AIW, in my favorite. Best capture of the lot. Values are stable, colors true, image stable, sharp without excessive haloing. Chase with Avisynth chrome offset correction, and it's good. Maybe NR if you want, but be easy/light.

Some of those are just plain ugly.
- the first was the 8960 and the NX card,
- then VR1500 was NX was even worse,
- with the VPH6990 and AIW being the ugliest yet

I just do not know if any altruisms can be made from these samples. Very interesting samples, but it mostly shows how important testing on multiple decks and capture cards can be.

Now I'm curious how interjection of the ES10/15 may affect each capture.

AGC inside VCRs and capture cards is the issue, and the nth gen source signal is causing the variations. This is all normal.