interestingly, i just checked the files from the wilco show i taped with michael miller's 722. i ran it with the auto-split at 2gigs. because michael's recorder has firewire issues and has to be sent back to sound devices, i had to transfer the files to my comp using the digi out to my sound card and record them in real time. i started wavelab recording about 15 secondsafter the 7 series started playing back with wavelab set to split at 1999 mb. the 7 series jumps to playbak the next file after 1 is done with a small little gap. so the first split appears in wavelab as dead space in the file. there are no lost samples and the files merged fine. but the left channel had 1 more sample than the right channel so i had to delete the dead space separately in each channel.
It doesn't seem to matter at all, but i wonder if the 7 series is off by 1 sample in each channel or this is something in my computer.