24.6.198.63
In Reply to: RE: Made Perfect Sense posted by kgiessler on May 27, 2007 at 20:47:12
My understanding of the history is that the USB designers were not convinced that USB 1.0 could handle "CD quality audio" in the "request packet, send packet" model and keep up with it, especially if there was any other traffic going on on the bus. Thus the isochronous mode which reserves the bandwidth and doesn't require the host processor taking part in handling each packet request. With USB 2.0 there is plenty of bandwidth and the USB chips with integrated processors can easily handle the overhead, but that was not all in place when the specs were being written.
An attempt has been made to come up with new specs based on USB 2.0 that would probably do what you are talking about, but the major players cannot agree on anything, so we are left with the original specs. What that means is that anyone doing such a project has to both program the processor in the USB chip to implement their private protocol, but they also have to write, support and maintain windows and Mac drivers to implement that protocol. Thats a LOT of work for a small company. And the big mass market companies don't care, the adaptive mode is fine for them.
The only companies making a stab at this are ones trying to do many channels (either studio companies or home theater companies) which cannot get the bandwidth over USB 1.1, so they HAVE to use 2.0 and come up with their own protocols to take advantage of that. And even if they DO implement somethig in 2.0 there is no guarantee they are going to do it in such a way that a local low jitter clock is used to clock the data. The others I know of that are doing this are using frequency synthesizers to generate the clocks (thats the easy way to do it) which is not particularly better than the adaptive mode. The 0404 is the only one I know of that actually has separate crystals for the different frequencies.
John S.
Follow Ups: