Distorted colors during capturing?
 

I am capturing PAL Hi8 tapes with a Sony CCD-TR820E ---S-Video---> green AVT-8710 ---S-Video---> Pinnacle 710-USB ---USB---> Windows XP with VirtualDub (more details here). However, I sometimes get extreme, blue-ish/purple discoloration, as if a color channel drops out (though I checked whether all connection where properly plugged in):

Attachment 12913

Attachment 12914

• It only switches between normal and weird colors during transitions, so from scene to scene, never during a take.
• It is not deterministic during which transitions this happens, so a scene can be normal in one capture and weird in a following attempt.
• I tried another camcorder (a Canon UC-X65Hi(E), without line TBC) which doesn't have this problem, so it can't be the tape or a cable.
• It also doesn't happen when I playback with the broken camcorder on a TV, so it can't be the camcorder.

What the hell is going on? :question:

I'm suspecting that the AVT-8710 is at fault but it can be easily verified by removing it from the workflow, Also some Pinnacle devices are built in MPEG-2 encoder chip so you can only get MPEG-2 out of them and maybe vdub is not coping well with MPEG-2 and decoding it on the fly, If it's the case get the one without the chip, it's usually a white shell, If you have the black shell chances are it is built in MPEG-2 chip.

A serious chroma error like that is usually either caps in a unit, or bad connection (internally, externally). I've see this exact problem before, and it's almost always for those reasons.

That would be unusual.
JVC PAL > green AVT-8710 > 710-USB is one of my usual PAL workflows.

Yep.

Not this card, not the 710.

It's almost 100% the AVT-8710; bypassing it doesn't yield the issue.

For some tapes it happens almost all the time, for some it only happens after a while. The problem exists for both S-Video and composite connections to the AVT-8710. When both are connected, either unplugging or plugging the S-Video connection restores the colors to normal (I assume the AVT-8710 automatically detects the composite input when the S-Video disappears), and pausing and playing on the camcorder also restores the colors. The (composite and S-Video) connections from the AVT-8710 to the 710 never changed anything, so I don't think the problem is there.

Anyone got any ideas? I'm quite concerned, since it is by far the most expensive component.

Has it ever overheated?
What is the longest that it's ever been plugged in?

Have you considered modding it with heatsinks on the chips? The plastic Cypress units overheat, and the heat dissipation is lousy.

Be very careful plugging/unplugging power without a rest/cooldown period.

Did you try the broken camera > TBC > TV/computer?
It may be a workflow conflict issue, two pieces (or more?)) of hardware hate one another. That happens.

Don't be too quick to place blame, that's not science. Careful testing/experimenting is needed to confirm observations, don't just run with the first hastily guessed cause from a single observation. It may still NOT be the AVT-8710 at this current time.

If you don't have audio sync problems and the video is stable capture without the AVT at least for the tapes that trigger this phenomenon.

Not that I know of, but pretty sure it hasn't. How would I notice?

I think 5-6 hours maximum. Is that too much?

That sounds like a good idea! Thanks :)

I'll try that this evening.

Good point!

Yeah, I considered that, but then I won't get the reduced image jitter, though I haven't necessarily found that to be an issue.

I have the black 710-USB and it outputs AVI to Virtual Dub. Wouldn't the MPEG format for capture be enabled only by the Pinnacle software?

I just tried this, and the problem still occurs, though it looks different:

Attachment 12918

Again, pausing and playing or plugging and unplugging resets the colors back to normal.

Could it be an overheating problem? What can I try next?

What happens if you turn off the line TBC on the Sony TR820E?

Just a general troubleshooting thing but you might try another suitable power supply for the AVT-8710. It uses 12V 500mA, which is standard enough. The problem is that it's center negative polarity, which is quite non-standard. See this thread (from one of your compatriots in NL) for more specs/links:

http://www.digitalfaq.com/forum/vide...lug-green.html

No, no control, These were designed back in the Pentium era where capturing lossless was impossible due to limited hard drive space, so the only way to do it is to capture to MPEG-2 on the fly, For a fast system it was possible but most low cost PC's didn't have the power to encode to MPEG-2 on the fly so Pinnacle put that task inside the capture device instead of the computer CPU, Such cards have no use today.

I lost track of which one does what, it's been years now.

Holy shit, that's it! Thank you so much for pointing that out! :laugh: I just recorded an entire tape without problems!

Does anyone have advice on what to do next? The color difference is marginal (though may this vary with other, older tapes?):

Attachment 12919

And I would of course like to have distortions fixed in tapes, should those pop up.

What are the chances of buying another camcorder and having its line TBC be compatible with my AVT-8710? Is the same problem likely to occur? Is that worth it, or should I just forget about line TBC altogether?

Great! Glad you got it sorted (although maybe double check with a couple other tapes that were problematic before to make sure it's fixed).

It was just a wild guess... a line TBC in the VCR/camera and a frame TBC in the AVT-8710 shouldn't conflict, but man... with analog video you just never know. Sometimes you just have to try different combinations of things to get a good transfer.

You might want to check prices on a used Panasonic ES10 or ES15 DVD recorder to add to your toolkit. They're dirt cheap compared to an AVT-8710 or TBC 1000. You don't use the DVD burning function of them--you pass your signal through them just iike your AVT-8710. They have some amount of line TBC (lordsmurf calls them TBCish) as well as frame sync. So for cases like these problematic tapes, you would do:

CAM (TBC Off) --> ES10/ES15 --> AVT-8710 --> Capture card

No guarantees that would help, but an ES10 or ES15 is not a bad investment to have around if you will be doing analog capture.

Edit: You could of course also try a different Hi8/D8 camera with line TBC. But most of the good ones are Sonys, and I fear that they probably use the same TBC process as your TR820E. Maybe someone else will have other suggestions.

I own the Sony TRV66 and never had to put nothing in the workflow but the Pinnacle 500-USB capture device in Windows 7 and TBC/DNR ON in the camcorder, I captured V8 and Hi8 tapes from different households with different camcorder brands used to record those tapes.

Does the Pinnacle let you know when it drops frames? Unless the Pinnacle does some frame sync, you might have just been lucky with your tapes not having much/any damage. I have the TRV65 (one generation older than yours I think) and the line TBC in it did not prevent dropped frames on some tapes I have with damage. But with a Panasonic ES15 in the chain, it captured with 0 dropped frames. Of course the damaged spots are still damaged in the capture, but at least the transfer was stable, meaning that audio stayed fully in sync with the video. Each dropped frame makes the audio start diverging more and more from the video.

Ideally you want both line/field and frame TBC, but there are cases when only one or the other gives you the best capture. And perhaps cases where neither is the best choice. This hobby seems to require a lot of tinkering to get quality results!

Sort of.

Software MPEG capturing sucks, the end, no exceptions. The best it ever got was the buggy beta-quality MainConcept 1.3/4 offering.

Pinnacle Studio captures AVI, and compressed in software realtime to MPEG. It's terrible. I forget if Studio forced-deinterlaces (or if it only did that in some versions), but the MPEG was blurry and blocky.

It's not like the AIW (hybrid hardware-assist MPEG, sort of like modern GPU encode/decode), or actual hardware encoders (some of which still look terrible, so being hardware isn't a guarantee for quality).

I saw this earlier, and knew it had to be a camera<>TBC issue. It wasn't impossibly to be a capture card issue, but far less likely.

Why didn't I think of that? Duh. :smack:

Ah, yes, good call, it could have also been that. :congrats:

That really doesn't matter.
- Po-TAY-to, po-TAH-to.
- Six one way, half a dozen the other.

Being different, and in the minority, doesn't really make it "non-standard". Both are standards.

Wait ... what? I was capturing lossless for restoration in 2002. I had a P4 system, 60gb OS drive, and added a new 100gb drive (largest available to at the time) to the system. That gave me at least 2 hours of footage recording space (2-hour, SP tape). A lot of file/space juggling was involved back then, to arrive back at the restored authored DVD.

You're referring to those old never-good/never-quality Pinnacle/Dazzle PCI cards, aren't you?

Ah! Excellent. But also interesting.

ES10/15, or better camera. (Both? Never hurts to have ES10/15 laying around. Cheap sub-$150 for what it can do in various situations.)

I've used many setups, as have others, and never come across this exact issue.

No! No, no, no. :no2:

Both hardware and tape.

Some users, novices especially, are too eager to go for "brands" and "models" of hardware, without any regard to condition of the 15-25 year-old items. So while a certain brand/model of camera or VCR is suggested, that exact unit may be a lemon. But you won't learn that until testing and usage. Or in the case of many novices, no testing, learn in usage. Or worse, learn only AFTER usage, when finally watching/viewing their unattended captures.

In this situation, with what we know now, after the process of elimination (good job, everybody!), it's probably a bad capacitor in the camera. That happens.

BTW, before the culprit was found, I was thinking an ES10/15 being injected may have narrowed down the issue. But (again) duh, just turn off the line TBC in the camera, for the troubleshoot. In my defense, I've been really tired this week. :o

:congrats:

In another of our OP's threads, I had my concerns about this camera. Mostly the age. I was proven correct here. My advice there was to try it, it's probably fine. But after trying it, it's obviously not well.

You're just lucky -- with the tapes you've seen so far.

I've used dozens of combinations of cameras, capture cards, and (with/without) TBCs. The simply fact of the matter is that Hi8/Video loves to drop frames. For example, when the home-shot tape has a scene change, with unrecorded tape snow between, even with a line TBC enabled, the video may drop 15 frames. That's a half second. You now have a noticeable audio desync. Hi8/Video8 drops more than VHS, in my year of experience.

Some people never see that, but it's mostly due to how the tapes were shot (camera work), the quality of the shooting camera, and the ability of the playback camera to match the shooting conditions. That's a lot of if/but's.

I'm neutral on that exact Pinnacle card, there are worse, there are better. It's unremarkable, nothing special. Just a card that will chug along, for better or worse. (I had one, sold it long ago.) I do wish I'd have opened it before selling, to snap a shot of the board chips. If you could do that sometime, and PM me, I would appreciate it. But also noting that Pinnacle/Dazzle had many revisions, some not publicized on the casing.

I recall that it did, yes, but that was 9 years ago.

It does not.

Yep.

I have/had the 66, same here, dropped frames. But it wasn't even damage, just scene changes and bad shooting. Also a bad shooting camera, where it'd blip at record start/stop. It's not damage, but it dropped frames. LOTS of frames.

:congrats:

Yep, almost always. :congrats:

Yep, but rare -- far less often than new/inexperienced users tend to think. (Everybody wants to think they're special, but few actually do come across these situations. There are weird scenarios that I've only seen once in 20+ years, but have been able to give advice to a handful of others that also came across the weirdness.)

:congrats:

Looking back at the thread again, color differences in TBC on/off is often a caps issue. Both something like AG-1980P, and camcorders.

I guess you were one of the very few luckiest that fills the entire hard drive of a PC with two hours of SD video when most of us were not even thinking of backing up a CD audio without the crappy mp3 compression, My computer had a 10GB hard drive in 2001.

Thanks for all the help, everyone!

Will do that!

So the line TBC in a Panasonic ES10/15 is not as good as a camcorder's? Wouldn't it be better to try a different camcorder, first, because if that succeeds I have a better line TBC? In that case, would the Panasonic still add anything (e.g. do the line TBCs complement each other, or does it have other valuable functionality)?

I will if I don't forget about it :)

What are caps?

I have noticed the image going weird on a handful of tapes on the TRV66e, though that seemed to be reproducible rather than random and I'm not sure if it was exactly the same effect as here, (while the older Sony Hi8 VCRs didn't mess up the image in the same spot.)

Yeah the TBCs in these camcorders are not full frame ones that ensure a stable output framerate. They do seem to do a bit more than the ones in most VHS VCRs though, and are extremely good at stabilizing horizontal jitter. There may bee some improvements in the video8 format that helps compared to vhs as well but I'm not sure. The bandwidth is a tad higher at least.

I usually just use my pioneer or sony DVR as passthrough when capturing 8mm tapes (as my datavideo TBC has some noise issue and a bit of image offset), but if the vertical sync / top of image area is messed up you do need something pretty solid to capture without frame drops yeah, so I use an ES10 or EH57 as passthrough for that. Those issues are not all that common though. I've found that having a hitachi D8 camcorder as a backup has been useful in some of those cases since it doesn't blue-screen like the Sonys.

The pinnacle 500 does have a video chip from NXP related to those used in the datavideo tbcs, so while it may not have the frame TBC/buffering to correct for sync loss it is at least a step up compared to many other capture cards when it comes to handling dodgy video signals.

Yeah everything I have read from lordsmurf and others is that the line TBC in the actual playback device (VCR/VTR or camcorder) is stronger than the one in the ES10/ES15. So your idea of picking up a different Hi8 or D8 camcorder would make some sense, especially if all or most of your tapes are Video8/Hi8.

But an ES10/ES15 would likely be cheaper to acquire, and would be useful if you had other kinds of analog tape to transfer (VHS/VHS-C, Beta, etc.). And unless the tapes have severe time base errors, then having stronger line TBC correction may not even matter.

If you opt to get another camcorder, you might want to look for a different brand just in case the problem you're having is with Sony's TBC implementation rather than your specific cam having a defect in its TBC circuit. hodgey seems to have had good luck with Hitachi, so maybe that's an avenue to explore.

If you decide to get a D8 camera just make sure that you get one that can play back Video8/Hi8. I think I read that some D8 camcorders aren't backwards compatible (though I myself have no experience with D8).

capacitors