I had no problems recording & playing back 24 simultaneous tracks with a MOTU 2408 mkII with my 700MHz Pentium III, 500Mb of RAM and a 80GB 7200 rpm Western Digital FireWire drive. And so, how fast, how much, do you require that you should be so concerned? There are very few machines these days that are so underpowered they have trouble with 24 tracks. Most problems generally boil down to "OPERATOR ERROR". So you really have nothing to worry about. It doesn't seem like FireWire 800 is taking off? But then neither did high-definition DVD and maybe not even Blue-Ray? Dolby beat out DBX and where are they now? In the realms of history lessons that's where.

You don't hear much from Ray or Dagmar Dolby anymore?
Ms. Remy Ann David

Firewire is plenty good. It is not at all a bad thing.

Who's complaining and what specific complaints do they have? This buffer thing you mention, do you know more about it?

I use firewire as do tens of thousands of other people. I'm sure there are problems but generally it's not due to the firewire connection.

If you want to step up from Firewire, there are PCI audio interfaces.

The issue is with MIDI delay which has been somewhat problematic since we moved away from serial/MIDI adapters (ah..sigh)

Unless you're scoring films with tons of MIDI info, you're probably fine. With a little tweaking most people don't have lots of problems. Its the sync to MIDI tracks that is an issue, but also one of the easiest things to fix via MIDI.

Phil

Ah yes, I see. Thanks again all!
Great forum with great answers!
-mark

One recent post on these forums misleadingly referred to FireWire causing latency because it used buffers.

Any I/O on a DAW workstation that has to process a stream of data in or out in a clock-accurate fashion has to use buffers, whether the physical interface is FireWire, USB, PCI or some other type. It's the size of the buffer that influences latency, and FireWire is no different from other I/O methods in this respect.

Most sophisticated DAW software can compensate for buffer size during tracking by time-advancing the replay (output) relative to the recorded input. However, the problem comes with software foldback monitoring, where the output cannot be time-advanced as the data to be output has not been recorded yet. The message in this case is clear: to avoid buffer latency, use I/O hardware that offers direct monitoring of live signals mixed with the replayed tracks.

Boswell wrote:

Thank you fine sir, that cleared it up for me 100%. I'm sure others will learn from this knowledge as well!

Thanks,

I use a RME FF800 running at S400 with Cubase 4 and can do as many tracks as I like at 48Khz with an AMD FX60 2.8Mhz 4G RAM XPro SATA drives with no problems. Input latency is 6.271msec and output latency of 7.333msec...I would like to upgrade to S800 cable and PCIe card but so far this works solid and stable with no hiccups. Other USB units I've used in the past (Lexicon, Alesis etc) had way higher latency numbers as well as driver issues and yes the USB bus system has too much crap on it already!! Unless someone comes up with a better bus stream for digital recording besides FW it will still be the way to go for the foreseeable future IMHO....