Well - I know some of you are...
I\'m considering moving to a Firewire RAID system, perhaps initially just a two IDE drive unit, striping the two drives...
I was wondering whether there is any real advantage over going to this solution over a slightly cheaper solution of getting two external IDE drive housing boxes and then just plugging my drives into them and splitting the samples logically across both drives. e.g. woodwind on one, brass on the other?
I guess the answer comes down to how much you know about your \'usage\' of your sample data. If you \'know\' that you always use a mixture of woodwind/brass (as an example), then you can probably achieve this quite well. If however, you vary your usage quite often, then I guess a raid striped solution whereby ALL samples are shared across all drives, would be a better solution.
I\'m interested to know what is really stopping more than 160 polyphony... if - for instance - I raided up 9 fast drives on firewire, surely I\'d be able to stream/access enough sounds realtime to drive a lot more than 160 voices. I don\'t like that GS places that (soft-coded) limitation on the user. Granted some people are finding it difficult to get even 160, but it seems to me that the PCI bus can handle more, so can firewire, so can the CPU... it is just people\'s drive configs that are the choking point?
Of course I\'m willing to be shot down and corrected