>> From memory this aspect was not mentioned in your big shootout >> recently? It would be great when new interfaces were reviewed that >> there was some kind of defacto test run to measure this aspect of >> performance - like "round trip" latency at specific buffer settings >> while measuring CPU load. Unless my thinking is somehow flawed this >> would be a pretty decent measure of a device's performance? > This is a great idea. I've had the opposite experiences as Nick, some FW interfaces I own/tried work flawless at a significantly higher CPU load than others. MOTU & MH are great in this respect, RME much less so. All this with the same buffer size. My G5's with MOTU interfaces/drivers are quite usable @ 128(I can run a Logic song with an EVB3 organ with moderate polyphony, 2 GAPs, a small EXS piano, 1 instance of Altiverb and some small fry), my FF800 will glitch & crackle with the exact same setup. Another factor; Intel based Macs perform better than Motorola(Gx) ones. It would be interesting to run the same test on G5's and Intel Mac's. Best, Zip Boterbloem Media Mechanics Zwaluwstraat 54 2025 VR Haarlem The Netherlands +31627014758 zip@knoware.nl