What VCR worn heads looks like? Before/after JVC head swap!

[aramkolt]

I am refurbishing a JVC HR-S9800U for a member here and noticed right at the end of the process when I was dialing tracking in with an oscilloscope that the picture and waveforms were pretty underwhelming - in particular, the familiar THX logo on a commercially produced SP speed tape had "full line luma shadow bleed" (though I'm sure there's a more technical name for it) that went all the way to the edges of the screen and the picture also seemed to lack as much detail as it should have.

It did not have lots of dark/static streaks after bright areas on screen or lots of full line dropouts oddly enough which is more what I've always thought of with worn heads, but I haven't personally come across many machines with worn heads, so this one was interesting to me.

It wouldn't track well to an EP tape despite all the adjustments I could try and I now believe that the EP heads wear faster because they are thinner (so less surface area needed to wear away a given thickness of the protruding head). All heads wear simultaneously whether they are in use or not since they are still moving past the tape.

I later learned that the owner did not use the machine for EP at all, so they were unaware that EP was a problem and only had me refurbish it because it would power down after inserting tapes and also had black and white video output over S-Video, so it was no longer useable in that state.

Of note, this monitor is an LCD that does show sub-black (less than 7.5IRE signals) so I don't think you'd see the dark shadows on high contrast areas on a regular capture that clips or displays anything at/under 7.5IRE as black and those shadows would be invisible during playback normally.

I'll do a short video capture with both the old head and the donor (less used) head in ProRes422HQ (10 bit) of the full THX logo via S-Video to an AJA KiPro, but here are the initial preview "screen shots" of the LCD-displayed playback. This isn't exactly the same frame/angle/focus, but the differences in detail and "luma shadow bleed" are pretty apparent to me anyway.

Posting also as a sort of public service announcement to raise awareness that you might be using your machine with very worn heads and not even know there's a problem, or it could present in a way that you might not immediately think of. "THX logo test" is worth giving a try if you're curious, especially if you can boost luma to the point where you can see below 7.5IRE to see how uniform the "black" background is and how much bleed of both light and shadow there is without it being clipped to black.

Original Head THX Logo SP.jpg

You must be logged in to view this content; either login or register for the forum. The attached screen shots, before/after images, photos and graphics are created/posted for the benefit of site members. And you are invited to join our digital media community.

[lordsmurf]

Keep that tape. We may need to discuss it later.
Set it aside.

Does this luma offset/bleed issues exist on all tapes, other tapes --- or just this one?

FYI: If just this one tape, that bodes well for your testing desires. I can help guide you in some experiments. But ONLY if it's an error on just that tape, and repeatable in many/all other VCR models/makes. That tape represents an edge case I've not had time/ability to look at myself just yet.

[aramkolt]

It's the same tape both times, same machine, only head was swapped to a less worn one. But yes, I definitely will save the tape. Detail and luma seemed off with all content of the worn head that I tried. Luma also seemed to be noticeably high in any sort of bright scene and there was generally lower contrast with the whole image appearing more "washed out." I think any commercial tape with the THX logo or even using it during standard credits where it's white text on black background should give a similar look on a machine with worn heads.

I speculate that all brands of VCR will do this when the head signal is uniformly weak (from heads being very worn) is the head amp has to boost the entire signal multiple times higher than it would with heads in better condition. What could differ is if they have circuits in place to try to mask/compensate for it differently. Could be some other models or brands will prefer to show static over a "washed out" signal.

Essentially, I think this is a signal to noise problem. All that follows is speculation, I'm not an electrical or radio engineer haha.

These are made up numbers as an example, but let's say the background noise level is always 10mV, and normal heads scanning a good tape produce a total waveform amplitude/height of 100mV. That means the signal represents 90mV of the total signal fed to the head amp before amplification. This gives a signal to noise ratio of 9:1 (90mV:10mV).

Now with very worn heads, let's say the total waveform reading the same tape is reduced by 50% and is now 50mV, but the noise still represents 10mV of that (so 40mV signal, 10mV noise), your signal to noise ratio is now 4:1 which when amplified to standard levels by an op-amp, a far larger percentage of your resulting waveform used for everything else in the VCR is noise. this over-amplification gives a more "overprocessed" look with details lost.

It's kind of like when you have a very quiet audio playback and you raise the volume to hear it better, the background noise gets amplified along with it and since there's less signal to begin with, more of what you hear is noise in the amplified signal.

To further complicate things, very worn heads probably get a little wider a they wear down or at the very least could have different pickup properties - similar to how a sharpened pencil graphite gets wider as you write with it. This could cause the head to pick up even more noise from neighboring helical tracks if the heads are now wider than they should be, so your noise floor might actually go up with worn heads too. In the example above, let's say noise is now increased to 25mV instead of 10mV (from the heads being the wrong width and picking up information from adjacent tracks) and since the total waveform height is still now only 50mV, this would give a signal to noise ratio of 1:1 (25mV signal to 25mV noise).

At some point when the signal to noise ratio gets low enough, the VCR just can't tell what is signal and what is noise and you're left with mostly dropouts or static. What's interesting here is that the head amp still makes something of that weak signal without actual typical "static dropouts", though it was extremely picky about where the tape guides had to be adjusted in order to get that stable image (possibly due to wider head width or change in pickup characteristics?).

Reminder - all of that above could be wrong, but seems like a reasonable explanation to me. Seems the JVC as long as it is tracking appropriately can recreate an image with a noisy signal without dropouts, it just may not be very accurate or contain as much detail as it should.

The other takeaway is that EP heads *probably* wear out first on most machines due to being physically thinner, so if you are concerned about head wear, do these tests with an EP recording. I wasn't able to here because the worn heads would not read EP reliably at all. I'm guessing this has been discussed elsewhere, I just personally haven't stumbled across evidence for it.

The test of value here for head wear is really any black background with white text (credits usually) or THX logo on a commercial tape and then to boost luma or black level up (usually this would be done with a proc amp or you might be abel to do it with contrast settings on some monitors) so you can see how uniform the black background is, and to see how far luma bleeds past the edges of white areas. From these tests, I think some bleeding will always be there just because of how VHS works, but the less, the better. This will definitely be going into my VCR head-to-head comparison, though I'm sure other test patterns will also reveal the same defects.

[Aya_Rei]

So what would be a good sign to know when it's bad, the shadow offset covering the whole screen?

Do have some offset, is this amount normal? Intentionally increased the contrast to make it most noticeable.

2006 Sci-Fi ECW Wrestling Recording.00.png

You must be logged in to view this content; either login or register for the forum. The attached screen shots, before/after images, photos and graphics are created/posted for the benefit of site members. And you are invited to join our digital media community.

Opening to Jaws 1997 Widescreen VHS.00.png