I just purchased a potentially interesting ?400 budget laptop (with TI chipsets all over, Cardbus and Expresscard connectors, etc.). Will post some details soon. A MADIFace is on its way here, and I'll test it for multitrack recording performance next week or so.

Meanwhile, this machine is running XP SP3, and everything seems to work ok, except for periodic (yet irregular) single DPC latency spikes, some below 2000, some up to about 3000 or so. I tried deactivating this and that, killing services left and right, nothing. Then more or less on a whim, I deactivated the "Microsoft ACPI-Compliant Control Method Battery", and the spikes were gone. Well, almost, there was one of 2800 just now. rant Will see whether this actually causes dropouts. Anyhow, this will now occur once in several minutes, nothing regular.

After finding out the hard way (two days of searching), I googled and found that this solution has worked well in context with other laptops and audio interfaces. So if you have inexplicable DPC spikes, try this. The downside is that there will be no display of remaining battery power in the task bar. But I personally never record on battery anyhow.

Interstingly, this battery thing causes no problems whatsoever on my old 933 Mhz PIII Toshiba Tecra 9000 (my main recording laptop). On this machine, though, there is a constant DPC latency value of around 300, instead of 12 or so on the new one. The Toshiba uses an older driver version, I might try to exchange it...

Forum reports elsewhere indicate that this does not improve DPC latency readings in Vista.

Regards
Daniel Fuchs
RME