So what will be the big deal for USB 3? more inputs probably?
Most likely.
Will audio be better on USB 3?
A few semi-informed thoughts, aimed at being improved upon by those more informed:
As already mentioned, USB (and Firewire, and any similar connection) is just a means of transmitting digital data. So long as the connection works as it's supposed to, the data received will match the data sent, and it will have absolutely zero effect on the audio that's being described by that data.
The various connections
do, of course, differ to the extent that:
(a) they
don't work the way they're supposed to, so that the data received
doesn't match the data sent; and
(b) the compromises that have to be made in order to ensure that (a) doesn't occur.
These are obvioiusly two sides of the same coin. To put it in practical music-recording terms: the first ordinarily appears as pops and dropouts. The second takes the form of the increased latency and limitations on track-count, sample rate, bit rate, etc. that you put up with to avoid pops and dropouts.
At the risk of stating the obvious: Music recording differs from more typical data transfers (like, say, moving data among disks and ordinary office applications, like word processors and spreadsheets) in that it has to happen in real time, and in a "real time" that's pretty demanding to current ordinary consumer technology. In typical data transfers, the software can recognize a bad bit of data with standard error-checking measures, and go back and try it again until the data is retransmitted without an error. When you're using a word processor and it's loading a file from an external hard drive, you don't even notice when - say - a particular characteX is mis-transmitted: even if the computer has to go back and tryi it 100 times until it comes out right, the delay is too slight to notice. If it has to go back and transmit every fifth character 100 times, the only effect will be that the file will load more slowly than it otherwise would have ... but it'll load quickly enough that you probably won't care much. When you're transmitting music - even just an ordinary CD - you
must send the equivalent of 176,400 word-processing characters every second.* If you fall behind even once the audio signal won't be transmitted in time to play perfectly. Software has enough error-tolerance to deal with dropping a single sample, but once you miss more than a few, dropouts and pops necessarily result. So speed and accuracy (two sides of the same coin, since error-correction trades speed to get accuracy) matter for music much, much more than for typical computer tasks.
Another element to throw in is the extent to which a particular data-transfer method uses other computer resources that your music software also wants to use: CPU processing, for example. The key criticism of USB 2 for audio (at least relative to Firewire) is the extent to which it burdens the CPU. USB 3 is still somewhat mysterious, but it
looks like the engineers' primary purpose was to make things like hard-drive backups faster, so that they didn't really address CPU use. This tends to suggest that USB 3 - though it has a considerably higher maximum data-transfer rate than Firewire - may still be inferior to Firewire for audio applications.
As for Firewire, my impression is that its primary champion was Apple, and that Apple seems to have lost interest in it. Which suggests the likelihood of an improved version of Firewire is low.
______________
*A more demanding example: say you're trying to transmit just 4 mono (or two stereo) tracks simultaneously in a 24/96 format. You need to send the equivalent of 1,152,000 word-processing characters per second. That's about 3 novels
every second.
