Video conversion hardware matters!

[BmacSWA]

Here's a quick side by side of the same video captured from an OLD (1988) videotape that was a 1st gen captured on basic hardware and real deal "Big Boy" hardware found and preached about on the forum. The left side of the video was captured on a 2002 Panasonic PV-V4524S VCR composite out through a Diamond VC500 (probably not the good one) USB into VirtualDub. The right side was Captured through a 2004? JVC SR-MV50 through a Big Voodoo TBC-10 into a proprietary PCI Express capture card. Both were deinterlaceds in VirtualDub with QTGMC and resized in Final Cut Pro on the Mac. The difference is incredible. Sharper images from the SVHS machine, less "shaky" video due to the TBCs and capture card.

I am curious about the JVC and TBC combo trying to handle some poorly dubbed tapes. Is almost looks better, which is still poor, without the TBC than what it spits out. I'll post a video demonstrating that later. Tell me what you think after taking a peek at this video.

Also I usually take the VHS to 60fps, but this was just a quick and dirty

[lordsmurf]

Quote:

Originally Posted by BmacSWA

composite out through a Diamond VC500

AGC is staticly wrong. Yikes.

Quote:

VirtualDub with QTGMC

How?

Quote:

I am curious about the JVC and TBC combo trying to handle some poorly dubbed tapes. Is almost looks better, which is still poor, without the TBC than what it spits out. I'll post a video demonstrating that later.

I don't understand this comment.

Quote:

Tell me what you think after taking a peek at this video.

Quality has improvded, as expected.

[BmacSWA]

Quote:

Originally Posted by lordsmurf

AGC is staticly wrong. Yikes.


How?



I don't understand this comment.


Quality has improvded, as expected.

how? An avisynth script dropped into Virtual Dub…

The hardware, I’m assuming the tics remove some “static/horizontal interference” however the quality of the tape is so poor only so much can be done with hardware alone.

[lordsmurf]

Quote:

Originally Posted by BmacSWA

how? An avisynth script dropped into Virtual Dub…

Shucks, I keep hoping a native VirtualDub QTGMC will someday exist.

Quote:

The hardware, I’m assuming the tics remove some “static/horizontal interference” however the quality of the tape is so poor only so much can be done with hardware alone.

It really depends on the errors. It's why I have multiple VCRs, TBCs, and capture cards. We can build the best workflow, but it still does not have 100% coverage for the various errors.

[aramkolt]

Hmmmmm. The difference isn't quite as big as I was expecting haha. The black levels are definitely off in the left sided capture, but the whites are very obviously being clipped in the right sided capture. Blacks may also be clipped as well, that's harder to say. Jitteriness yes is improved on the right, but not completely eliminated like I would hope the full frame TBC would do. I don't think the Big Voodoo is particularly touted on DFAQ, but I could be wrong. People tend to use what is available and there's nothing wrong with that.

My question is how much jitter goes away by just using the S-VHS machine with line TBC on? While the line TBC technically shouldn't improve horizontal jitter, that player alone may just do better with it even without the Big Voodoo in the chain.

Seems like too many variables were changed at once to really know what item is causing each benefit (jitter improvement/sharpness) or detriment (clipping).

When you deinterlace with QTGMC, what format and bitrate do you have it output that is more Mac friendly?

[lordsmurf]

@aramkolt, your assessment here is not correct.

The values on both is pretty awful, but the tape is primarily at fault. The VCRs each interpreted the values differently, and that's expected. (For the record, vhs-decode would do no better here.) This is the reality of craptastic home movies, and the onus is largely on the recording camera and camera operator all those decades ago. JVC VCRs have a known profile for values, which tends to err slightly on the side of legal darks and proper contrasted highlights. The alternatve is a washed-out muddy mess. Panasonic S-VHS tends to err on the side of neutral contrast, but tend to fall into illegal black/darks. You really cannot win here. Proc amps are what you must have.

Again, you're too visual. The role of the BV is not visual, aside from proc amp values. It's for signal stabilization, the non-visuals. Certain early 10 models -- not random versinos -- are also some of the best TBCs ever made.

There are a lot of variables here. But it starts back in 1980/90-whatever, when the tape was made.

VirtualDub2 can output the hacked/reverse-engineered ProRes422, and that's what I always suggest.