Alternative capture methods: SDI capture directly from TBC?
 

Wondering if anyone has tried capturing SDI video output from rackmount TBCs that have them for analog captures? If it's a viable option, it would seem to have some real theoretical benefits at the very least - completely avoiding an around of digital to analog and then back to digital conversion.

My thought is doing that essentially makes the TBC the main analog to digital conversion (in other words, IT IS the capture card) in the chain. In a typical setup, the TBC converts to digital to store the frame, but then converts the frame right back to analog and there most likely is some measurable amount of signal loss there, plus you then have to capture it with a regular capture card which requires it to go from analog to digital yet again (more potential for losses).

By recording SDI output from the TBC, the signal I would hope remain digital the entire time and you'd skip the need for converting back to analog and then once again converting back to digital with a traditional capture card.

Using an SDI capture card I think is essentially taking already digital data and storing it without any "conversion" per se.

I'm going to attempt to try this myself and I'll report back (gotta wait for some hardware and also have to hope the stuff works since everything is sold as-is untested these days).

I know doing analog to SDI then to capture SDI is frowned upon here, but I think that's because we don't trust the analog to digital converters in the analog to SDI converter. You HAVE to be able to trust the TBCs ADC since that's in your chain no matter what.

Those are called converters not TBC's, Their main function is to capture and convert analog to digital, Analog outputs are for monitoring purposes. It was never intended to be used the way you are using it.

There are few people that are using such devices like me, Jwillis and others, These devices are of high quality broadcast standard and their big advantage is that they relief the computer from the digitizing task, it's all done in the FPGA, and the digital signal is then streamed to the computer for storage via a SDI/PCI, SDI/USB or SDI/Thunderbolt adapter.

Interesting! Out of what you've tried, is there any advantage to doing it that way other than offloading processing or does it lead to cleaner captures? I've found on my XP capture machine that CPU use rarely goes over 10%, so I don't know that I'd do it for that reason personally.

The couple TBCs I had in mind are the Hotronics AP41 (with SDI output option), and possibly the AJA FS1. I also thought this was possible with say the Leitch 475, but turns out that requires a stable input signal already and that the TBC it has can't be used at the same time as some other conversions.

SDI has fail points, it's not a panacea. But it can be acceptable under certain conditions.

I actually have several, but never use them. It wasn't even until a few months ago that I bought another SDI card, having misplaced my original.

Some SDI converters do act as non-clear passthrough (non-clear = TBC works), same for certain DV converters (that also outputs bypassed DV compression). Actual "passthrough" is actually untouched in>out, but you want some touching for TBC. Not all items have that, the TBC is not engaged on passthrough. Most, actually, are true passthrough, few route through TBC.

Generally speaking, SDI/DV/recorder passthrough is weak compared to TBCs designed for analog sources, like standard DataVideo and Cypress. The ES10/15 is really the main exception to the rule, not the rule at all. Noting the Es10/15 is far from transparent output, it touches more than TBC.

Hmm. I guess the question is how to tell if the SDI out on those devices was passed through the TBC portion of those devices or not. How would you be able to tell unless you're seeing visual dropped frames or maybe audio sync issues? Theoretically if the video had gone through the TBC, the audio sync should remain fine I would think.

Spec sheets, and hope the sheets are accurate.

No, audio sync is not a test for presence of TBC. You can get seemingly in-sync, but other issues present eventually. The quick test for TBC generally requires a library of test tapes.

Snell & Willcox have nice block diagrams in their manuals. Example for CVR600 series Attachment 17179