OK thanks for that explanation about buffer. Two followup questions, if you don't mind.
First question is, when the write speed falls under the minimum speed during your testing, what is the resulting effect on the resulting recording? Digi-noise? Skipping? I just want to confirm with 100% certainty that what you're seeing when the speed drops too low is the same as what I'm experiencing in my unit. Thanks.
As far as I'm aware, cards that don't meet this threshold are dropped from further testing, so I don't know what the end-world result is of continuing to record in this case.
We are looking at improving this, though I'm not privy to the potential details.
The second question is just the engineer in me that's curious...I think some people have commented in this thread that a certain card works fine in their DR680 but not in the DR70D. Is it a logical conclusion then that there's more buffering memory in the 680 and that's why they're not seeing the same issues on the 680?
Unfortunately I don't know, I've not seem the design docs for either products, and no other cases of increased recording failure has been reported so far, so there's no failure report doc yet. Given that the DR-680 does more channels to start with, it makes sense that it would have a bigger buffer though.
Circumstantial evidence though still points at the card having slowed down since being used previously. If someone has a card from that exact test, they should do an erase format and try it again, and let us know what card it was (if it was on the DR-680 tested media list).