performance of PS3M module replayer for amiga
category: code [glöplog]
did anybody ever use ps3m
http://aminet.net/package/mus/play/ps3m312
http://aminet.net/package/mus/play/ps3msrc
and if yes, how is the performance and support for pattern effects?
http://aminet.net/package/mus/play/ps3m312
http://aminet.net/package/mus/play/ps3msrc
and if yes, how is the performance and support for pattern effects?
@neoman
Performance in what sense?
In general I would expect S3M and XM to be far superior to MOD (more samples, bits and channels) and in fact they sound very cool.
Have you had a look at the program documentation and assembly source code?
Performance in what sense?
In general I would expect S3M and XM to be far superior to MOD (more samples, bits and channels) and in fact they sound very cool.
Have you had a look at the program documentation and assembly source code?
performance in sense of cpu-time for the replayer obviously.
never used it, sorry.
but what exactly hinders you to simply implement the replayer-code into a simple basecode and poke a color into $dff180 after you called it in your 50fps-loop ? This will show you exactly how much "rastertime" it eats up.
then you can run some tests with pattern effects aswell.
should not take more than an hour all in all. ;)
never used it, sorry.
but what exactly hinders you to simply implement the replayer-code into a simple basecode and poke a color into $dff180 after you called it in your 50fps-loop ? This will show you exactly how much "rastertime" it eats up.
then you can run some tests with pattern effects aswell.
should not take more than an hour all in all. ;)
@ɧคɾɗվ.
Yes cpu-time obviously for you.
I interpreted "performance" also as accuracy of the effects thus my question.
Yes cpu-time obviously for you.
I interpreted "performance" also as accuracy of the effects thus my question.
PS3M had audible noise and a reduced volume due to its software-mixing. It got worse the more channels your module had and it was quite CPU hungry. IIRC Hippoplayer had it included, so you could quickly test a few modules on your target hardware and see if they play fast enough and sound good. Which CPU are you targetting by the way, 68000 or 68060 (or anything inbetween)?
For me, PS3M would be a no-go if you were planning to have any demo-effects running while the music is playing, as I think it was too slow. If you were just thinking about a music-disk, you could be fine. Also, your mileage might differ.
Maybe also check for Digibooster, OctaMed, Art of Noise, MusicLine, TFMX, or any other multi-channel replayer.
For me, PS3M would be a no-go if you were planning to have any demo-effects running while the music is playing, as I think it was too slow. If you were just thinking about a music-disk, you could be fine. Also, your mileage might differ.
Maybe also check for Digibooster, OctaMed, Art of Noise, MusicLine, TFMX, or any other multi-channel replayer.
hardy: I know I can measure it on my own but I was more asking for someone who already has some experience with PS3M.
noname: Thanks, that's exactly what I wanted to know. I already took a look at DigiBooster 1.7. Its replayer source looks pretty cool and faster than the one of PS3M. Also for me quality is OK. It's for an intro for 68040/60s :)
noname: Thanks, that's exactly what I wanted to know. I already took a look at DigiBooster 1.7. Its replayer source looks pretty cool and faster than the one of PS3M. Also for me quality is OK. It's for an intro for 68040/60s :)
Yeah, I tested it back in 1997 when the code was out. The playroutine was able to play pretty much everything I tested (xm and s3m mods) back then. But there is likely still some bugs or unimplemented features in the player. I remember some 16-32 channel XM mods took around 50% of CPU time on my 50 MHz 68030. But with less amount of channels, like 8, the CPU requirements are decent, still too much for an unexpanded A1200, if one wants something else to run simultaneously.
As someone already wrote, the audio will become more noisy and volume reduced with more channels, because of software mixing, and there is no "volumeboost" implemented, if I recall correctly.
As someone already wrote, the audio will become more noisy and volume reduced with more channels, because of software mixing, and there is no "volumeboost" implemented, if I recall correctly.
pgm: thanks, good to know. PS3M doesn't feel too complete. I don't like it to be limited to just 4 channel but still I love module music for intros which is why I'm researching my possibilties. I'm pretty fine with DigiBooster and learn how to use it now. Also for 8-ch music it's just using a quarter frame for playback on an unexpanded 1200.
I used it (with XM's) in a couple of demos in '00 - '01. The results were really quite horrible but I was too young and stupid to realize.
Korvkiosken: You didn't upload these demos here? ;)
As far as I can remember they were:
- Hypnopolis
- Smurphine (except for the banjo piece in the middle)
- the *first* version of Scrimm. The final version just converted the multichannel XM to a streamed format instead. Better sound quality and lower CPU usage at the cost of just a tiny bit more disk space.
- Hypnopolis
- Smurphine (except for the banjo piece in the middle)
- the *first* version of Scrimm. The final version just converted the multichannel XM to a streamed format instead. Better sound quality and lower CPU usage at the cost of just a tiny bit more disk space.
