2004-02-11 - edit - now up to ~100 hours of field and field-simulated testing.
I'm starting to perform bit-accuracy tests on the Hosa ODL-312 AES/EBU <-> optical digital format converter. After ~10 hrs of testing (admittedly far more to go), so far so good. I wanted to start doing these tests as I would use the unit in the field, so I'm running everything powered off of DC:
MK4/KC5/CMC6 > V3 (16-bit/44kHz) split to
[1] > coax > EgoSys Waveterminal 2496 > CEP2.0
[2] > AES/EBU > Hosa ODL-312 > optical > NJB3
- V3 is powered by 6v SLA.
- NJB3 is powered by its internal battery.
- Hosa ODL-312 takes 9v DC, draws 50mA (according to the specs, not verified) and is powered off a 9.6v RC, 1600 mAH battery pack.
- Files transferred NJB3 > PC via Adaptec Fireconnect 4300
I've trimmed the files from [1] and [2] to the same starting/ending samples and run the EAC WAV compare utility. So far, no differences whatsoever.*
*Outside of the 10 bit-perfect hours, though, I've twice now encountered a problem on the *PC* side. Somehow the L/R channels are swapped and one of the channels is offset by 1 sample relative to the other. WTF. Not a clue trying to sort this one out. So at the beginning of each run I test to make sure L is recording to L and R is recording to R on both the NJB3 and PC. When this criteria is met, I proceed with the tests and so far the results have been only positive.
Lots more testing to go, but I thought everyone would like to know how it's going so far.
FWIW, I picked up my Hosa ODL-312 at Full Compass (I think) for ~$50, shipped. Not quote as cheap as some of the others, but so far so good on the bit-accuracy front and this thing is built like a
brick.
I'll post a few pics in a second here and will report back after further testing.
Edit: Confirmed bit-accuracy through ~70 hrs of testing now. I'm done - 70 hrs is good enough for me.