skippy
New member
I posted this originally down in the "which audio card" thread, which probably wasn't appropriate, and got no responses. So I'll try again as a new topic. Sorry if you've seen this before. Please help out a computer-recording neophyte...
I'm an old analog tape machine guy, so my room is set up in that manner: a standalone HDR (Fostex D1624) and a standalone mastering deck (Alesis Masterlink) with an analog board stuffed in between them, and not a single computer in sight. Now that I've used the D1624 for a couple of projects, I've come to realize that it is a very good machine with incredibly braindead data backup "capabilities". So I'm looking for a multichannel ADAT interface. Argh! A computer is about to enter the studio. Blasphemy.
What I naively propose doing is having a 16-channel ADAT digital I/O in a PC of some sort so that I can simply snapshoot the HDR's digital bitstreams during operation, *on the fly*. This will allow me to back up the data on the HDR pretty much transparently during normal operation, even while overdubbing. Basic concept: the HDR and the computer acting as its backup should have essentially redundant copies of the data at all times, providing a sort of disaster-recovery mechanism beyond the HDR's "undo" capability- which doesn't do *jack* once the HDR's internal disk drive dies, or gets optimized, or gets formatted, or whatever.
I could also then do editing and screwing around with the data with some of these newfangled audio packages on the backup machine, if I feel like learning how to do it later. But that's really a secondary concern, believe it or not. I didn't want to do computer recording (which is why I bought the D1624), and I still don't: PCs piss me off too much, and I stare at computers all day for a living. However, the HDR and a PC will have different failure modes, so one of the two should generally survive any blowup.
My experience has always been that "if backup takes an extra operation, it won't happen reliably enough". I walk away from the box without burning backup CD-ROMs far too often, and the process of manually dragging that data onto a backup machine is too painful to bother with. All the rest of the machines here (a bunch of Sun workstations and a couple hundred gig of SCSI disk, which host my real-world EE consulting business) are already networked, and I have a centralized backup server and Exabyte tape jukebox that handles dumps of my business data for me every night under cron.
Always having a 24-hour-old tape backup of everything important has saved my professional life a number of times. I can't stand the idea of not being able to automate the backup of the damned audio data as well: the one day I don't back it up is the day I'll screw it up. If a current copy of the data resides on a machine on the network, the nightly backup can be completely automated, and my ample butt should be amply covered.
After several weeks of archive searching, I've pretty much settled on the Sonorus StudI/O (followed by the Hammerfall, and more distantly by the Frontier Dakota). I haven't settled on a host machine or software package yet, but I haven't found any ADAT driver software for Solaris, and I don't have time to write it. My experiences with trying to get Macs to talk to Unix boxes have been rather unpleasant, so I'll probably go with Cakewalk on some sort of Wintel box. I'm a Unix guy, but I do currently have a couple of those talking to Samba on one of the servers.
Has anybody else played with this sort of setup? I want to keep my options open for future growth (like learning computer-based editing/munging/MIDI stuff: It could happen. Dinosaur or not, I ain't that old yet), so Cakewalk/Wintel seems like a defensible framework to put in up front. According to the hype, it understands transport control, which means that (theoretically, anyway) I could use it to notice that I've said "go" to the D1624, and to snapshoot the data it sees on the ADAT inputs to the card...
This propsed rig has much more power than I probably need for this initial disaster-prevention use, but offers a lot of room for growth. Am I completely off base with the choice of methodology, cards, software, and hardware? Let the tomatoes fly.
[Edited by skippy on 10-17-2000 at 13:54]
I'm an old analog tape machine guy, so my room is set up in that manner: a standalone HDR (Fostex D1624) and a standalone mastering deck (Alesis Masterlink) with an analog board stuffed in between them, and not a single computer in sight. Now that I've used the D1624 for a couple of projects, I've come to realize that it is a very good machine with incredibly braindead data backup "capabilities". So I'm looking for a multichannel ADAT interface. Argh! A computer is about to enter the studio. Blasphemy.
What I naively propose doing is having a 16-channel ADAT digital I/O in a PC of some sort so that I can simply snapshoot the HDR's digital bitstreams during operation, *on the fly*. This will allow me to back up the data on the HDR pretty much transparently during normal operation, even while overdubbing. Basic concept: the HDR and the computer acting as its backup should have essentially redundant copies of the data at all times, providing a sort of disaster-recovery mechanism beyond the HDR's "undo" capability- which doesn't do *jack* once the HDR's internal disk drive dies, or gets optimized, or gets formatted, or whatever.
I could also then do editing and screwing around with the data with some of these newfangled audio packages on the backup machine, if I feel like learning how to do it later. But that's really a secondary concern, believe it or not. I didn't want to do computer recording (which is why I bought the D1624), and I still don't: PCs piss me off too much, and I stare at computers all day for a living. However, the HDR and a PC will have different failure modes, so one of the two should generally survive any blowup.
My experience has always been that "if backup takes an extra operation, it won't happen reliably enough". I walk away from the box without burning backup CD-ROMs far too often, and the process of manually dragging that data onto a backup machine is too painful to bother with. All the rest of the machines here (a bunch of Sun workstations and a couple hundred gig of SCSI disk, which host my real-world EE consulting business) are already networked, and I have a centralized backup server and Exabyte tape jukebox that handles dumps of my business data for me every night under cron.
Always having a 24-hour-old tape backup of everything important has saved my professional life a number of times. I can't stand the idea of not being able to automate the backup of the damned audio data as well: the one day I don't back it up is the day I'll screw it up. If a current copy of the data resides on a machine on the network, the nightly backup can be completely automated, and my ample butt should be amply covered.
After several weeks of archive searching, I've pretty much settled on the Sonorus StudI/O (followed by the Hammerfall, and more distantly by the Frontier Dakota). I haven't settled on a host machine or software package yet, but I haven't found any ADAT driver software for Solaris, and I don't have time to write it. My experiences with trying to get Macs to talk to Unix boxes have been rather unpleasant, so I'll probably go with Cakewalk on some sort of Wintel box. I'm a Unix guy, but I do currently have a couple of those talking to Samba on one of the servers.
Has anybody else played with this sort of setup? I want to keep my options open for future growth (like learning computer-based editing/munging/MIDI stuff: It could happen. Dinosaur or not, I ain't that old yet), so Cakewalk/Wintel seems like a defensible framework to put in up front. According to the hype, it understands transport control, which means that (theoretically, anyway) I could use it to notice that I've said "go" to the D1624, and to snapshoot the data it sees on the ADAT inputs to the card...
This propsed rig has much more power than I probably need for this initial disaster-prevention use, but offers a lot of room for growth. Am I completely off base with the choice of methodology, cards, software, and hardware? Let the tomatoes fly.
[Edited by skippy on 10-17-2000 at 13:54]