I did a stupid thing this weekend - I recently reformatted all my disks to FAT32, switched from W2K to WXP, and installed DX9 (risky, I know). Then I did a 9-track recording of a live show without first testing stablity of my box. I record with a MOTU 24i.
I know the error in my ways, but simply didn't have the time up front - and the loss of this concert recording is not really a big deal.
Here's what happened. It appears that on every track a few bits were dropped at fairly regular intervals so tracks go in and out of sync with others. The mismatch is just a few miliseconds, but enough that it is clearly audible. I had recently adjusted the buffers from 256 to 128 on the MOTU console, and also had not disabled write behind caching on the drives. Also, I may have had the data directory and temp directories on the same disk, although I have not seen any evidence of Vegas using temp storage while recording.
The machine is a P4 1.8 with 512MB RAM and all disks are at least ATA100 on their own IDE channel. I have recorded many times without mishap running W2K.
Any tips on what I might look for to clear up the problem? I have disables write-behind and put the buffers back to 256, but have not had the chance to test this setup. I'll do that tonight.
Also, would ASIO in V4 be less prone to encounter this type of trouble?
I know the error in my ways, but simply didn't have the time up front - and the loss of this concert recording is not really a big deal.
Here's what happened. It appears that on every track a few bits were dropped at fairly regular intervals so tracks go in and out of sync with others. The mismatch is just a few miliseconds, but enough that it is clearly audible. I had recently adjusted the buffers from 256 to 128 on the MOTU console, and also had not disabled write behind caching on the drives. Also, I may have had the data directory and temp directories on the same disk, although I have not seen any evidence of Vegas using temp storage while recording.
The machine is a P4 1.8 with 512MB RAM and all disks are at least ATA100 on their own IDE channel. I have recorded many times without mishap running W2K.
Any tips on what I might look for to clear up the problem? I have disables write-behind and put the buffers back to 256, but have not had the chance to test this setup. I'll do that tonight.
Also, would ASIO in V4 be less prone to encounter this type of trouble?