On Sun, Dec 03, 2000 at 12:19:17PM -0500, James Davidheiser thus spake: > > > > > Core: MZ104 from tri-m. The power consumption kicks ass, its got barely > > > enuff power, I think, to squeeze out some MP3 playing capability. > > > > I dunno about mp3 but I think it will be fast enough to drive > > my LCD menus with Java and monitor my home sensors from SF ... > > (it may be too slow for mp3) > > > > I've gotten mp3's working on a pentium 60 with 24 megs of ram already, it > was a bear, I hadda downsample everything, and if I clicked on a menu > (this was under win95), the audio would skip and jump and other fun It's an efficency problem more than CPU power. I have a P100 and playing MP3s using XMMS uses 20-30% without any skipping or other performance losses while running X, Netscape, ssh, etc.... mpg123 and it's lib form used in XMMS is the most efficient player out there that I've ever seen. IIRC, a 486DX4/100 is just enough to play 128kb/s MP3s at full quality. Also, in Linux, you have the advantage of a good scheduler, unlike Windows. Usually, just doing something like the following ensures good playing on smaller systems withh little/no jerking: nice - -20 mpg123 -b 2000 foo.mp3 which gives mpg123 the highest priority in the scheduler. So you could run Emacs at the same time, but it woudl be slower and if you start to get I/O bound by swapping to disk, then all bets are off. If you want to be extreme about process priority, you could easily alter mpg123 to use a few of those Linux scheduler calls that would effectively give it first pickings at *every* time slice. (immuteable priority) Even more so, given that x86 is the platform of choice, one could write code under Real-Time Linux, and give some code priority over Linux itself (kernel and user space). Actually, for hardware interfaceing, I'd reccomend RT-Linux. It's bare metal control with all the niceties of UNIX. :) Eric LaForest -- Subscription/unsubscription/info requests: send e-mail with subject of "subscribe", "unsubscribe", or "info" toWear-Hard Mailing List Archive (searchable): http://wearables.blu.org please, Please, *PLEASE* don't subscribe through a forward/false domain
From Wear-Hard Mailing list Archive (WH)
Maintained by R. Paul McCarty
Archive created with babymail