Alright, I ran a handful of tests, a handful of times against each test case. Here's what I found with my gear...
Upstream Gear 1 - field
MK4 / KC5 / CMC6 >
Lunatec V3 (24-bit / 44.1kHz) >
Coax S/PDIF
Upstream Gear 2 - PC
Foobar2000 (24-bit / 44.1kHz) >
EgoSys Waveterminal 2496 >
Coax S/PDIF
Record Settings
MEDIA : Kingston 45x CF Card
PHANTOM POWER : Off
FILE FORMAT : 24-bit
INPUT SOURCE : S/PDIF
INPUT MONITORING: OFF
MONITOR VOLUME : OFF
DISPLAY LIGHT : OFF
L/M/H GAIN : L (n/a)
In all cases, I left the coax S/PDIF cable connected to the MT, with the upstream gear constantly outputting a signal during all operations (power up, record, stop, format, etc.). This was to simulate my field use - I will always have my V3 outputting a signal prior to performing any operation with the MT. For this first round of testing, I only generated very short files: 10s - 5min, depending on whether I was preoccupied with work and conference calls.
[01] Power up MT, record
Whether the CF card is empty or contains data, the MT properly reports the Record Time Remaining when initiating the first recording after power up.
[02] Power up MT, record, stop, record again
Whether the CF card is empty or contains data, the MT improperly reports the Time Remaining when initiating a second (or third, or fourth...) recording. In all cases, the MT reported the entire remaining capacity of the media as the Record Time Remaining, rather than the actual Record Time Remaining given the 2GB file size limit.
[03] Power up MT, record, stop, format media, record again
Whether the CF card is empty or contains data upon power up, the MT improperly reports the Record Time Remaining when initiating a second (or third, or fourth...) recording - even after formatting the CF card. In all cases, the MT reported the entire remaining capacity of the media as the Record Time Remaining, rather than the actual Record Time Remaining given the 2GB file size limit.
[04] Power up MT, record, stop, delete file, record again
Whether the CF card is empty or contains data upon power up, the MT improperly reports the Time Remaining when initiating a second (or third, or fourth...) recording - even after deleting the file(s) on the CF card. In all cases, the MT reported as Time Remaining the entire capacity of the media.
[05] Power up MT, record w/o signal
MT does not freeze upon recording a file without a signal, or stopping the recording.
[06] Power up MT, record w/o signal, stop, record again w/ signal
MT does not freeze upon recording a file without a signal feed, or stopping the recording - as long as the S/PDIF cable is connected.
With my field gear upstream, the MT successfully started a follow-up recording every time - sometimes displaying the Record Time Remaining correctly, and sometimes displaying the entire remaining capacity of the media as the Record Time Remaining.
With my PC-based gear upstream, the MT sometimes locked up when starting a subsequent recording (with or without a signal) unless first powering off/on. When it successfully started the second recording, the MT displayed the entire remaining capacity of the media as the Record Time Remaining.
Note: recording a file without a signal feed and without the S/PDIF cable connected, and it locks up *every* time upon "stopping" the recording - can't even initiate a second recording.
Not gobs of test cases, or gobs of runs against each case, but enough for my tastes for the moment. Nothing definitive here, really, just thought I'd pass along my results. Basically - it seems aside from the very first recording after power-up, my MT displays the entire remaining capacity of the media as the Record Time Remaining. Not always, but most of the time. From now on, I'll power off/on the MT between sets, as a safeguard in case the MT decides to report a too-short Record Time Remaining instead of the full media capacity as in my tests above.
Still not sure what the MT will do if it reads too much Record Time Remaining time (e.g. 4:23:10) and hits the 2GB marker. Haven't gotten that far in my testing, yet.
And just in case something screwy happens over time, and since it reflects the scenario I generally find myself in when recording, I'll repeat test case [02] with longer initial recordings and then stop and restart recording only after the level meters go haywire.
PLEASE don't take these results as indicative of what *your* MT will do. It seems others have entirely different experiences than mine, with the Recording Time Remaining generally reading shorter than it should (while mine generally reads longer than it should).