View Full Version : Audio Modeling on GPU (newb)
nbohr1more
02-Jul-2008, 04:49
Perhaps I've overlooked info about this in the forums but I'd like to know a little about GPGPU's potential as Pro-Tools/Digidesgin DSP . I fondly remember my Roland VS840 Digital 8-track that my wife made me sell :cry: , it had an audio modeling technology called COSM. I'm sure my crusty old P4 could pull it off but I'd love to know how old of a GPU could do that level of Audio modeling. I wish I had more references. Nothing on Google or Wiki about the underlying chips... :oops:
out of interest how would an xfi dps compare to one of those ?
cosm Composite Object Sound Modeling" technology that shapes audio by applying the sonic characteristics of popular or classic microphones, guitars, guitar amplifiers and studio reference speakers.
while eax shapes audio by applying enviroment characteristics like room size and the material its made of so there is some similararity
nbohr1more
02-Jul-2008, 18:34
The marketing materials for the XFI indicate that something like COSM should be do-able, I hate Creative drivers and have serious misgivings about the false claims of 24bit Audio in my Audigy Platinum Ex. (Crappy settlement requires you to purchase more hardware)
nbohr1more
02-Jul-2008, 19:22
I kinda see what you're implying. The audio modeling hardware should be in proximity to the dedicated Audio card... I agree, except for the fact that proprietary Audio hardware vendors charge more per FLOPS. Which also begs the question, "Video Cards with HDCP require Audio hardware, so why not make the GPU availble for Audio Modeling since it's nearby?" (I know, my use of BTQ is borderline...)
vBulletin® v3.8.6, Copyright ©2000-2013, Jelsoft Enterprises Ltd.