Is the [VCR Input] box on the 700-USB a line TBC?
 

I just made a clip of a rather dramatic surprise on cheap hardware.

Granted it matches the quality of the tape but here's the setup:

1. Panasonic PV-4657 (4 head VHS, composite output, nothing special, no DNR, no TBC, lousy tracking)
2. Pinnacle 700-USB Porsche Blue Box, USB 2.0 powered
3. Hand drawn Graph Edit, crossbar, capture device > NTSC_M, YUY2, 720x480

So its having line length problems, flagging, but looks pretty good otherwise.. very little noise.

I normally ignore the [ ] VCR Input check box as non-functional.

Checked on.. and lo & behold.. the image straightens up and locks solid, every single line the same length, both frames in perfect sync. And no color skewing.

I wasn't going to actually do a capture.. just trying things.. so what gives?

Strange, that checkbox usually doesn't do anything on other devices (even though some A/D chips have like fast-locking modes where it would make sense).

Have you had a look inside the pinnacle box to see what hardware there is in there?

Not really.. there are tons of inside pics all over the web.. and I have disassembled one.. it has simply never grabbed my attention. I like its aluminum heat sink heavy blockiness..I've used it mostly as a paper weight.

I made a short clip of the difference.. but I am rather embarrassed to post it.

The difference is quite remarkable though.

Where did you find pics? I couldn't seem to find any pictures of the insides, was curious what chips it has. From what I can find, the Pinnacle pci cards (and some dazzles) from that era seem commonly have philips SAA7113 or SAA7114 chips, but those only have a very limited ability to correct line jitter. Curious about what chips the USB 700 use if it has a functional line-tbc.

I can't find the images off hand.. they must be buried in a Forum some place. I've certainly taken pictures of them though.

I do know Linux drivers were attempted and failed a number of years ago assuming it was based on a Connexant chip.. but they have a custom signal processor chip before the decoder which pre-conditions the signal, its also a hardware crossbar processor.

The line was called the Marvin Pro (Avid Liquid Pro BOB) Marvin (700-USB) and Marvin-Lite (500-USB)

The Marvin Pro has 5.1 sound and component inputs, AV in/out to its crossbar

The Marvin dropped the 5.1 and component, but retained AV in/out to its crossbar

The Marvin-Lite dropped AV out from its crossbar

But they all seemed to have that signal pre-processor chip, it is quite massive.

The software was generally lackluster depending on a lot of proprietary hardware embedded features rather than software.. it had uncompressed and MPEG2 code/decode in hardware.. but they sold at BestBuy and Circuit City.. so I discounted them as serious boxes.

More or less I threw out the software since after Corel acquired Pinnacle the activation was non-functional anyway.. and started playing around with the DirectX drivers in GraphEdit instead.. the hardware keeps surprising me with the features accessible only from the hardware drivers.

Oh.. it doesn't have a DV codec.. but it does have a firewire i.Link port.

If Avery Lee hadn't spoke up for and supported the Plextor ConvertX 402U he might well have liked these boxes.

(ugh).. I attached the capture sample.. (yeah.. laugh all you want.. regarding the content..).. I was looking for a bad tape.

This is (not) S-VHS and not via S-Video, it is using Composite, no proc-amp or any other filtering.. its straight off a crummy 1996 vcr that wheezes like an organ grinder.

Oh, that clearly does something. I suppose it may be some custom solution then, like some DVD-recorders do. E.g the Panasonic DMR-ES10 has a chip that does raw analog to digital, and then sends the digitized signal to a different processor for further processing.

Granted, if it's conexant-based, I know some of their chips some programmable handling of video lines, so that may also be a possibility.

Let us know if you manage to dig up the pictures of the inside.

Yeah.. since it jitters from line to line, side to side.. and the edge on the Left crawls and weaves like a snake.. I figure its horiz. line correcting and not vertical sync or 'frame' correcting.

Its also not simply deinterlacing.. the edge is shifting back and forth. I am not sure about the mice teeth however.

Since the capture is "uncompressed" there is no MPEG macroblocking artifacts, so its easy to see that checkbox is definitely affecting the clean up of the input signal.

I would almost dismiss it as as a sync level switch.. except for all the weaving.. and what you can't hear.. the wow-wow fluttering of the vcr.. the drive belt is probably worn out. Which you would expect to create line length errors.. which a line TBC would tend to correct.

(attached) Is the Marvin Pro, the board on the Left is the video board, the board on the Right is the lower deck mostly 5.1 audio board. The Marvin should be just the Left "upper deck" board. The Marvin-Lite is an entirely different small matchbox board affair in a small plastic housing about the size of a deck of cards.

I guess I should have mentioned this is all being done on the Win7 x64 operating system.. but I don't think that would effect signal treatment/

I have the Pinnacle USB 500.

It was often a recommendation on german forums since 2005.
The box was bundled with the Pinnacle Studio 10 software. But it was possible to capture without the pinnacle software with Virtual Dub in YUV2 uncompressed.
Pinnacle wasn't pleased about this and with later boxes it wasn't possible or just only with the composite input with Virtual Dub. SVideo wasn't selectable. Maybe there are some registry tweaks possible to make the Svideo input selectable.

But the USB 500 have some limitations. It works best with Windows XP. With newer driver it isn't possible to capture under Virtual Dub. The capture range is only from 16-235. The super black and super whites are clipped. The box has proc amp features but it's more a fake proc amp. If you have video content where the white level is higher as the luma range 255, you can't make em visible to capture. For example the old Brooktree tv cards or Canopus NX, ATI AIW have an real proc amp to make the whites visible.
By my own recordings made with an old Siemens VCR and an Sanyo VCR in the 80's the white level is always out of range.

Yes the box has some light jitter correction and locked audio function. I have open my box and have made a picture from the board.
Inside is the philips 7113 chip.

If you use a Sony/Pioneer DVR (680,870,555....) in passthrough mode the 500 box can be a good choice if you can grab one cheap, because the Sony/Pioneer DVR's have white control and the output range is from 16-235.