Inserted frames with Dazzle DVC-100 in VirtualDub?

[simbapenn]

Wondering if anyone out there has successfully used their Dazzle DVC-107 or lower with VirtualDub and not had significant inserted frames reported when capturing? FWIW, I also have an IO-Data GV-USB2 and do not get inserted frames from either device. I could continue using the GV-USB2, but I would like to run each on different computers while I capture old VHS and Video8 tapes from my family.

Equipment: JVC 7800U, Sony Handycam TVR65, Windows 10 and 11 PCs.

Things I have tried:
Different versions of VirtualDub 1.9, 1.10, VirtualDub2
Different computers running Windows 10 and 11
Different cables (S-Video, Component, different sets of both)
Different USB ports (2.0 and 3.0, and different ones on each computer)
Different driver versions for the Dazzle.
Different codecs (Lagarith, huffy)
Both the VCR (a JVC-7800U both with and without TBC on) and the Sony Handycam I have

The weird thing is, even with no tape playing and choosing the Dazzle, if I test capture, it still inserts frames. About 1 every 10 seconds or so, so it's a lot of frames. Is it possible VirtualDub is misreporting the inserted frames and everything is actually fine? Could be the 1.07 is no good, or just my particular one is having issues. To my eye, the actual capture seems fine when I try it.

TLDR: Wondering if anyone has captured in VirtualDub without inserted frames from a DVC-107 or lower version of the Dazzle. Also if VirtualDub could be misreporting inserted frames, or how to check for inserted frames in a capture.

[lordsmurf]

Welcome.

Dazzles are known junk cards, with multiple issues, including bad video quality.

GV-USB2 is "better", but it's almost a meme'd card in recent years, and gets far more praise than it deserves. "Not as bad as others" is akin to "I wouldn't kick her out of bed" (ie, not a compliment). Note that most praise is also for PAL use, not NTSC. It's a Japanese card that tends to fail in any capture software, aside from the imperfect Japanese capture software AmaRecTV.

VirtualDub can be setup badly/wrong, and many Youtube guides settings those bad/wrong settings. For example, post-XP, you need to disable DirectX in the main settings. And then the capture timing settings are extremely important.

And you lack frame TBC. That's also a primary reason frames are being dropped.
Line TBC is not frame TBC.
line = cleans the image = essentially operates on per-frame basis
frame = cleans the signal = essentially operates temporally, frame-to-frame basis

Right now, you're driving a 1985 Ford, with a wheel missing. A bad experience is expected.

The solution is
(a) better card, both of those are lousy
(b) some form of frame correction, preferably an actual TBC (for your own sanity)

[simbapenn]

I appreciate the quick reply. They're good enough for my needs. I was just asking for people's experience so I could either stop troubleshooting the reported inserted frames or move on from trying to stop them and probably the device.

Do you know if there a way to see if the frame insertion stat is accurate on a capture?

[lordsmurf]

A screenshot of your VirtualDub timing settings would be a start.
And confirm DirectX is disabled.

The Dazzle really is trash. If you consider the GV-USB2 "good enough", then that's your path forward. You'll probably still face issues, both quality and usability, but it's the least-worst card here.

[Gary34]

Quote:

Do you know if there a way to see if the frame insertion stat is accurate on a capture?

You can watch your captures back on VLC or Virtualdub and look for the inserted frames.

[Aya_Rei]

Quote:

Originally Posted by Gary34

You can watch your captures back on VLC or Virtualdub and look for the inserted frames.

More specifically, if VirtualDub reported frame inserts then you can use ctrl + ] and ctrl + [ to jump to the inserted frames. ] jumps forwards, [ jumps backwards.

[aramkolt]

I have seen in general where in certain setups (wasn't using the dazzle when I saw this), there can be a single frame inserted at a very specific interval, like every 3000 frames there will be 1 frame "inserted" like clockwork. Also of note, this happens despite zero dropped frames. I presume this is due to very fine differences in clock timings, but sounds like many more inserted frames which I'd have to also guess is due to lack of Frame TBC. I will say that the 7x00U TBCs can be kind of different than the rest of the JVC line in my limited experience. They are really the only ones that on some tapes I've seen the picture get less stable with the TBC enabled, though your mileage may vary.

[lordsmurf]

Quote:

Originally Posted by aramkolt

I will say that the 7x00U TBCs can be kind of different than the rest of the JVC line in my limited experience.

Good observation, you're not wrong.

The 7000s has only 2mb of SD RAM, whereas the 9000s had 4mb. It made a difference. The difference is seen with stability "breaks" from buffer overrun (error overruns RAM). The latter post-7x00 "Pro" series was really a continuation of the 7x00, but with DDR RAM (effectively doubling RAM).

However, none of that trivia negates the fact that, here, in this situation
(a) frame TBC is missing here, and
(b) Dazzle is a well-known/infamous POS