|reply to koitsu |
Re: USB 2.0 Hi vs. 2.0 Full vs. 1.1
I think what is tripping me up is that it's called a "bus" ... so to me that implies that total bandwidth--or at least available time on the bus--is shared between devices. More devices = less time/bandwidth, therefore it becomes impossible to achieve maximum speeds...
The application is for multi-channel audio and I'm trying to determine if, using USB 1.1 Full Speed, it is possible to feed 8 channels (7.1) of 48kHz/16-bit audio to a device.
48000 samples per second per channel. 16 bits per sample. 8 channels ... which would be 48000*16*8=6144000 bits, or 5.85Mbps. Even with protocol overhead I think it should be do-able with Full Speed.
If we run at USB 2.0 High Speed then given the same slice of time on the bus we can move a huge amount more data and there is much less concern over getting audio starvation...
This is where the question of whether "sharing" occurs or not, and guaranteed bandwidth... i.e. if a user plugs an older USB device in, do we end up getting stuck with less bandwidth.
I believe that during USB enumeration there is a way to allocate, or at least inform the controller that you require, a certain amount of bandwidth, so that's what I'm looking further into now.
Mountain View, CA
This has gone from a general USB question to something much more complex (it sounds like you're designing an actual hardware device that uses USB). I assume you have a USB-IF membership? If so, have you asked the folks there?
Making life hard for others since 1977.
I speak for myself and not my employer/affiliates of my employer.