Jason, keep pressing them, when I was testing the nomad firmware early on, it had the same identical misplaced sample issues, which were solvable in firmware. the fact that the iriver seems to have the same issues, makes me think its trivial
a few things I've learned from 100's of hours (1000?
) of testing, that can save you some time:
-first and foremost, establish that the unit doesnt resample inputs at all sample rates. You've tested 44.1, confirm it with 48K, and you'll never have to do it again (for this firmware)
-once a soundcard is established to not resample, youre only looking for glitches, this is easier than full bit-accurate testing. the 'glitch-free' testing is the testing you should be doing whenever you install a known-good soundcard in a new machine. There are several methods for this, but I'll give you mine. This is the long, detailed version, it assumes you have soundforge, but you should be able to do it with any wav editor:
I generate a 440hz sine wave using wavelab. its a 5 sec wave. I then take the 5 sec wav into soundforge and copy and paste it out to a 5 minute file, then save that file. whenever I need a wave for testing, I take the 5 min wav, can quickly copy it out to a 7-8 hour wave. (keeping it stored as a 5 minute wav keeps me from eating up several GB's on my drive, and it only takes about a minute to copy it out to a long file). Anyway, once you generate the long file, go to the beginning and 'insert silence' (10 sec.). OK, now I have a reference wave file, with silence at the beginning. You will be playing back this test wav and recording it on the device you are testing. Start the device recording while rolling the silent intro. This makes it super-easy to cut it to a known point to compare it with the original wave. OK, to do a test, play it back on the computer, digi out of soundcard, then into the gmini. The only variable here is whether the digi-out on your soundcard is good, and wont drop samples (dont do anything else on the computer while playing it back). after you record a long file on the gmini, transfer it back to the computer. at this point, you can do the quick trim and compare it with the original for bit accuracy, or just do the search for glitches (much quicker). to search for glitches, move the cursor to the beginning of the file, and do 'find'>'glitches' (the following settings will find all audible glitches: -25db slope, 80 sens.). It takes all of 10 seconds to execute this command, then come back in a few minutes. it will either stop at a glitch, or go all the way to the end of the file and say 'no events of the specified type were located'
this is really good once you get the bugs resolved and it is working well, it is trivial to do 10, 50, 100 hours of testing on the unit, which should give you complete confidence in the field.
to test the effect of different spdif sources on recording stability, just go analog out of the computer, into dat, a/d converter, etc. the sinewave is still a sinewave (tho not digitally the same), so you can do glitch testing, but not bit-accurate (like I said above, you really only need to do bit-accuracy once for each sample rate in a given firmware on a given device).
I am currently doing testing as follows:
sine wave>mini to rca cable>r500>xlr out>v3 or ad2k>2 outputs feeding nomad and vx pocket (testing both). While the procedure above seems complicated, once you get it going, its a breeze. If youve got a lot of testing to do, its the only way to go. I've done 30 hrs of testing on each of the devices in the last 2 days, spent prob 1-2 hours total doing it.
