--- In logic-users@yahoogroups.com, "ivandeaton" <ivandeaton@y...> wrote: >> > Well, > You are absolutely WRONG on every point except #3(...yes,96k does impeed performance). > I've had years of experience with this, > and what I've found is this... 45 tracks on an interanl ATA will playback fine. 90 tracks on > the internal ATA will stop the system... ie...core audio error message. I can take half the 90 > tracs and put them on the OTHER internal ATA and all 90 playback as easy as the single > drive playing 45 tracs. Ivan with all due respect, you're making my point for me because what you've failed to mention is that a single ATA drive can playback around 90 tracks. i've tested them a bunch and that was always the case - they can handle between 89 and 92 tracks. so that you can playback 90 on 2 ATA drives makes my point exactly - you can use multiple drives, but they don't improve performance (OK, if you are claiming that by using 2 drives you can gain a performance increase of 1 track - i'll concede). now, if you could playback 170-180 tracks with 2 ATA drives, then you'd have some kind of proof about multiple drives. look, i've tested these a lot (along with using them). for more testing i did a long time ago on my dual 1.25 with ATA drives (among others) go here (you will see that ATA drives can handle 89-92 tracks): http://community.sonikmatter.com/emagic/ultimatebb.php?ubb=get_topic;f%; t0262#000000 >>>>> The same 90 on a SCSI array playback even easier with no probs. Well, of course they do - because the SCSI drive is faster. but if you've ever really taken the time to test a RAID vs a single scsi drive then you'd know that a RAID makes no real world difference for logic. you get almost exactly the same performance from a single SCSI drive. again - i'm not the only one who has experienced this. go here: http://discussions.info.apple.com/webx?14@46.SoRPaBRjAsO.0@.689d9169 and check the posts of a guy named Gerald Stringer - he thought RAIDs improved things too, until he actually tested it. seriously - just because you've spent a lot of time on your system doesn't mean you know how it performs. i've spent a lot of time on my system and i've spent a lot of time testing my system. i've done a LOT of testing on all of my systems and i've spent a LOT of time documenting that testing and posting it for the benefit of other logic users (primarily over at sonikmatter and at the VSL forum). and frequently these tests have been confirmed by other logic users. if you'd care to actually run controlled tests, document them, and post the conditions and results - i'd love to read them. i'm sure others would too. but until you do, i think suspicion should be cast on what you're claiming - because i have a great deal of direct experience which contradicts what you say. and let me stress - i DON'T WANT IT TO. i really WANT to be able to use all of these hard drives with logic - they cost a lot to be of no benefit. but as of today, they don't improve performance. so - to recap: 1. RAIDs DON'T improve performance with logic. well, to be clear, i've encountered one other user who was using a hardware RAID card (which a UL4d isn't) who experienced improved performance in OS 9. there is NO situation in which a UL4d will improve track counts with logic. i've tried single channel 8 drive RAID. dual channel 8 drive RAID. single and dual channel 4 drive RAID. single and dual channel 2 drive RAID. tried the RAIDs with UL4d. tried em with UL3d. tried all combinations in the 133 MHz PCI slot of a G5. tried all in a 100 MHz slot in a G5. and, with the exception of the UL4d, i've tried all the combos on a dual 1.25. and NONE of them showed any real world performance gains. 2. multiple hard drives DON'T improve performance with logic (beyond, at best, a couple of tracks). again - i've tried multiple (up to 8) 15K scsi. i've tried a 15K scsi on the UL4d and a 10K scsi on a UL3d (separate PCI busses). i've tried 15K scsi and a fw800. i've tried 15K scsi and an internal 10K SATA. tried internal SATA drives. again - in NO set-up could i improve track count playback performance by more than a couple of tracks. 3. EXS streaming does NOT improve when using multiple hard drives (as measured by the EXS Virtual Memory window - where is shows the 'could not get data in time' figure). 8 15K scsi drives on 2 channels of the UL4d drop data at EXACTLY the same point that 1 15K scsi drive does. again - tried this with all combinations (scsi and firewire, scsi and SATA, etc.) and not one combination ever improved performance. a LOT of people have assumed that using multiple hard drives improves track count and streaming performance simply because you can use multiple hard drives for both things. but i've tried many times to confirm this on many systems and have never been able to. hope this helps