PTIII audio glitching

[David Wright]

I'm using PTIII on an 8100 with G3 accelerator PT4.1.1 and OS 8.1 (please don't bother telling me to upgrade!) and I'm getting occasional glitching when reading or writing to the PT drive on the Digi bus. I've tried many approaches to this, including swapping the Disk I/O card and drive, but no luck so far.

The glitching occurs in exact sync with the disk activity light, and consists of short bursts (about 400ms) of audio from elsewhere in the same file. Occasionally there is a single isolated glitch, but most often it continues indefinitely once started.

If it happens during data write, then the file is corrupt and any attempt to play through that point in the file will produce regular glitching. However, you can play OK from just after that point. It's normally impossible to copy or archive the file.

If it happens during playback, you can stop and try again and it will be fine. A typical situation is when bouncing down a long session (e.g. complete album). If it happens towards the end of the session, I let it continue to the end, then bounce again from just before the glitch, and paste it over the damaged portion of the bounce file. Always works fine.

Seems to me it must be possible to reason out exactly what's going on here, but I just don't have enough knowledge of the SCSI read-write processes. This may be irrelevant, but I've only had this problem since starting to use larger capacity drives e.g. 18gb.

Any thoughts on this one would be really appreciated.

David

[Russ M.]

Back about 1996 I had those problems. It was the drive. The older scsi drives were fussy. A wide drive with a narrow adapter on it caused all sorts of problems. The problems went away when I replace it with a narrow drive, same drive, same firmware just narrow.

[Russ M.]

One other thing, I seem to recall the older systems having a 2 gig limit on size so you may need to partition your drive.

[David Wright]

Many thanks Russ. The drives I'm currently using are recent Seagate narrow SCSI drives (ST318418N) - they started producing them again last year. I too had reservations about using 68pin drives with adaptors and stopped using them.
Likewise I had the same thought about OS and partition sizes. I upgraded from OS7.6.1 to 8.1, and tried small partitions, but it didn't help, sadly. I am still using HFS (not HFS+) as I need to retain compatibility with some older apps.
BTW, I use a similar Seagate drive on the Mac bus for SDII and MLCD and it gives no problems.

[Russ M.]

Hmmm! Have you tried formating with variuos disk utilities.. I had great results with FWB. Also tried setting the various parameters in protools.. buffers etc. Just thinking out loud.. but the drive may be too fast for PTIII. It may be feeding data faster than it can be used.. There is a way to slow that down but my brain is a bit foggy this morning.. I seem to remember running into that at one point.

[Russ M.]

Also check that the drive is terminated correctly.

[xray]

Dear David,

The number "1996" rang a bell with me, because I also purchased around then, and had glitching problems: very delicate but noticable glitches.

I don't know if you're using an adaptec scsi accelerator, but if you ARE, you need to be sure to use the Version 4.1 firmware. (Downloadable) If you've got the old (1996-7) v 2.5 - that's the problem and V 4.1 will solve it.

I hope this helps you. I know how vexing this can be. Particularly the lack of support or understanding of this issue.

Dick

[David Wright]

Thanks again Russ. I've used various formatting utilities including FWB. I actually use Apple DriveSetup now as I had other problems with FWB. I've also tried a variety of terminators, active and passive.
Your comment about the drive being too fast sounds very interesting though, since the problem seems to have surfaced with more recent and possibly higher performance drives. If you can recall what the solution was I'd be really really grateful!
Thanks also xray. No SCSI accelerator involved as the drive is connected to the SCSI bus on the Digi Disk I/O card that was used on NuBus systems.
I'm not clear to what extent the PT software is therefore responsible for all disk I/O operations, or whether it uses the system routines. I suppose it's even possible it's a PT bug . . . I've tried to get an upgrade to v4.2 (the last version to officially support NuBus) but so far with no luck.

[Russ M.]

RE: too fast drive. I'm just trying to recall but I seem to remember that the drives cache was the problem, feeding data too fast to PT. There are settings pages that can be changed on the drive but you need to used something like FWB to get at them. I remember doing it but don't remember exactly what they were. Try looking at the Configure Driver window in FWB. There are abunch of setting like Allow Disconnect", "Blind Transfers" etc. It may be one of those. I remember playing with those settings.

[David Wright]

Thanks Russ . . . this ties in closely with something I was told by a colleague but he too was vague as to exactly what to do about it! I'm reluctant to just dabble . . . hopefully someone can come up with something more specific?