43 replies to this topic

[LinkoVitch]

Are you meaning to control overall music and over SFX channels volumes? or per channel volume control?

Both are implemented just the docs don't cover the overall master volumes If you want to play with them then you can adjust the volume settings with the variables:

U235SE_music_vol
U235SE_sfx_vol

These are both long words in DSP RAM, so you will need to access them as longs from the 68K. The ranges are 0-63.

I haven't documented them yet as they are still on my list of stuff to check and sort, those labels may change also, or vanish.. (same with much of it at the moment I suppose )..

HTH

[rush6432]

Nice! Ill have to give it a go. Is there any need to call a subroutine to update the values or does the player auto check the values as it goes?

Edited by rush6432, Mon Dec 26, 2011 2:29 PM.

[LinkoVitch]

They are the actual values used at the time of mixing, so once they are set they are live.

This new rework is coming along nicely too

[LinkoVitch]

Version 0.17 released (changelog), Major changes to the API adding simpler calls and greater flexibility as well as improved module accuracy.

[DrTypo]

Hello!
I tried version 0.17. It doesn't produce any sound on Virtual Jaguar. v0.16 was still producing sound. You may have hit something that break the emulation. Maybe you could check this with Shamus?

Also a small thing: in the delivery, it would be nice to rename DSP.o to DSP.a. Otherwise a "make clean" deletes the DSP code!

Anyway, thanks for your work on this sound engine.
I wrote a module player in C#, which is probably a slightly bit easier to code than DSP RISC, and I found module playback rather tedious to really get right.
Good luck with effect 0x0 (arpeggio), I "cheated" to implement this one and used float point math. A LUT should be used but I was too lazy to bother generating one at that point...

Edited by DrTypo, Tue Dec 27, 2011 2:44 PM.

[LinkoVitch]

Thanks for the feedback, if he doesn't read this i'll poke Shamus when I next see him, I suspect I know what may be upsetting VJ (it now uses divides).

Yeah some of the effects are 'fun' but I have already prototyped most of them in 68k asm before porting to RISC, so hopefully nothing should be a sticking point.



(null)

[LinkoVitch]

DrTypo, on Tue Dec 27, 2011 2:44 PM, said:

I tried version 0.17. It doesn't produce any sound on Virtual Jaguar. v0.16 was still producing sound. You may have hit something that break the emulation. Maybe you could check this with Shamus?

Hi, just got home from the pub (where I was when I 1st replied ) and tested against VJ myself, and it works fine for me..

What version of VJ are you using? I have 407 here and it is happily playing the alf.mod. I have only tested with the example code I have supplied with the Sound Engine

[LinkoVitch]

OK, looks like I may have stuffed up the release of 0.17 somehow :/ I will confirm and rectify as needed when I get home. Sorry about this.. I'll post again when I have an update..

[LinkoVitch]

OK, more on this.. I haven't stuffed the release of 0.17! YAY! It seems I have possibly tickled a bug in SLN!

For my test I have assembled main.o with madmac.

I have then linked with ALN, and also linked a version with SLN using the same binary blobs (main.o and dsp.o) which were produced by MAC.

The output from ALN works as intended, the output from SLN doesn't. I think SLN is introducing an alignment issue within the DSP code.

[JagChris]

Could you post a detailed as possible description of whats happening with SLN bug?

[LinkoVitch]

I don't think it is honoring the .DPHRASE alignment directives correctly, it may also not be honoring the .LONG directives either which will present all kinds of wacky fail when the 68K tries to write to a long that is not long aligned, and as you have to pretty much always write long words to the DSP from the 68K this is going to cause all manner of problems.

This is my assessment based on the behaviour that I have seen when linked with SLN, which matches behaviour I had when there were alignment issues whilst I was working on the code.

[JagChris]

LinkoVitch, on Wed Dec 28, 2011 12:57 PM, said:

I don't think it is honoring the .DPHRASE alignment directives correctly, it may also not be honoring the .LONG directives either which will present all kinds of wacky fail when the 68K tries to write to a long that is not long aligned, and as you have to pretty much always write long words to the DSP from the 68K this is going to cause all manner of problems.

This is my assessment based on the behaviour that I have seen when linked with SLN, which matches behaviour I had when there were alignment issues whilst I was working on the code.

ok my thoughts were to just make the bugs public. Since it seems SubQMod has abandoned the project I dont believe just sending him a private message of the details will resolve anything.

[LinkoVitch]

TBH I am not 100% convinced with the bug yet, I am still picking away at binary files etc trying to figure this out. I have just looked and for one of the data areas I was thinking would be wrong the SLN output matches ALN.. but the files are not the same. It just *feels* like alignement issues

[JagChris]

LinkoVitch, on Wed Dec 28, 2011 1:33 PM, said:

TBH I am not 100% convinced with the bug yet, I am still picking away at binary files etc trying to figure this out. I have just looked and for one of the data areas I was thinking would be wrong the SLN output matches ALN.. but the files are not the same. It just *feels* like alignement issues

I use to have the tools he made to compare ALN output to SLN when I was helping him a while ago but I have lost them.

[LinkoVitch]

Version 0.17 released (changelog), Major rewrite of the sond rendering code. Smaller, faster and now with greater tollerance of bus latency.

[DrTypo]

Great news! I tested it a little bit and didn't found anything obvious.
However there is a bug in your post, it's version 0.18, not 0.17.

I spotted a few typos in the manual: page 4, the object file is dsp.obj and not dsp.o. By the way thanks for changing the name, it's more convenient this way.
In page 5, the references to dsp.o must be changed to dsp.obj.
Page 7, there is another reference to dsp.o. There is also a reference to the PERIODxx.s file, which isn't used anymore.
That's all I got for now.

[LinkoVitch]

DOH! I changed the URL at least If a kindly mod passes by this way could they correct my post for me please

Thanks DrTypo for the typo's.. heh living upto you name I'll go fix them ready for the next release.. Glad to hear its working for you.

[rush6432]

Hey, i got around to playing with the newest version of your sound engine. I did find an issue with mod file i was using just to see how it would play on your sound engine and it seems that i plays ok for the most part except when the melody starts to come in the notes are totally flat/off sounding. Suggestions?

Edited by rush6432, Thu Jan 26, 2012 10:40 PM.

[LinkoVitch]

rush6432, on Thu Jan 26, 2012 10:40 PM, said:

Hey, i got around to playing with the newest version of your sound engine. I did find an issue with mod file i was using just to see how it would play on your sound engine and it seems that i plays ok for the most part except when the melody starts to come in the notes are totally flat/off sounding. Suggestions?

Probably the mod you are using is using an effect that I haven't yet implemented (there are a load missing :/) If you send me the mod I can have a look and probably prioritise adding them to the player if you need them.