home *** CD-ROM | disk | FTP | other *** search
- Nntp-Posting-Host: kolsaas.ifi.uio.no
- Newsgroups: comp.sys.amiga.audio
- Path: sparky!uunet!mcsun!sunic!aun.uninett.no!nuug!ifi.uio.no!nilsco
- From: nilsco@ifi.uio.no (Nils Liaaen Corneliusen)
- Subject: Re: Random modplayer?
- Message-ID: <1992Nov16.142014.7212@ifi.uio.no>
- Sender: nilsco@ifi.uio.no (Nils Liaaen Corneliusen)
- Organization: Dept. of Informatics, University of Oslo, Norway
- References: <1992Nov14.024056.69318@ns1.cc.lehigh.edu>
- Date: Mon, 16 Nov 1992 14:20:14 GMT
- Lines: 25
- Originator: nilsco@kolsaas.ifi.uio.no
-
-
- In article <1992Nov14.024056.69318@ns1.cc.lehigh.edu>, elm4@ns1.cc.lehigh.edu (Ed Mackey) writes:
- > Annoying busywait bug?! WHAT annoying busywait bug???? There is no point
- > in EdPlayer that I do any busywaiting when another task could be running.
- (...)
- > If you're talking about the DMA wait, then you have no concept of the
- > time involved -- there needs to be a pause between certain writes to
- > the audio chip, but this pause is so small there would not be time
- > for a single task switch on most systems. To exit the interrupt, resume
- > the current task, suspend the current task, and re-enter the interrupt,
- > would take much more time than the DMA wait itself, so there is no
- > point doing that. By the time you got out it would be time to go back
- > in. No work would be accomplished and time would be wasted.
-
- Assuming that you use the normal protracker replay, it is wrong to use a
- busywait. If you play a module that requires REAL tight timing, it will
- play wrong. The right thing to do is to use audio interrupts. At least, thats
- what is done in the coming IntuiTracker and it WORKS.
-
- BTW, what should the busywait be set to on a 4000?
-
- ------------------------------------------------------------
- INCLUDE "disclaimer.i" Zrouoggathog
- Nils Corneliusen Cthulhu in '96!
- internet: nilsco@ifi.uio.no Hasoithshuna
-