Yeah, the fact that there is 57 odd minutes with no sign of a glitch prior to that 3 seconds is really a PITA as well.... I wonder why it takes the NMT another 50 seconds to decide to crash.... strange.
If you do find one again (and I assume you will find another if you can't find any existing one) I've updated my writeTime utility to do a little more analysis on the contents that I'd like you to run n the PC. I'm not making a lot of headway on that sample and I must have crashed my dev. NMT hard 50 times trying different things.
Still haven't had any time to look yet...... getting a new one could take a few days as they don't happen all the time (thank the ....) & then there are ones like that sample that play through just fine without a glitch for 50 odd minutes.... but at least you now know how bloody frustrating it is (x50) when the NMT hard locks while playing back a recording