Protracker player with support for external sound fx

[Antiriad_UK]

Originally Posted by phx

Why do you include your custom.inc into ptplayer.asm?

Told you, crazy demo coders
I must admit I've gone the other way now and have separate objects and link them.

[Antiriad_UK]

Originally Posted by DanScott

Just popping Antiriads version into my demo system.

I'm getting a "repeatedly defined symbol" error, as the EQU's for the hardware registers are conflicting with my "custom.inc"

Might be worth popping an "CUSTOM_REGS equ 0/1" or similar, around those (and the CIA defines)

So could do something like this:

; Custom chip registers - try and catch predefined EQUs - if it fails sort this
; out yourself :)
	ifnd CUSTOM
CUSTOM		equ	$dff000
	endc	;CUSTOM

	ifnd INTREQR
INTREQR		equ	$01e
INTENAR		equ	$01c
DMACON		equ	$096
INTENA		equ	$09a
INTREQ		equ	$09c
	endc	;INTREQR

	ifnd AUD0LC
AUD0LC		equ	$0a0
AUD0LEN		equ	$0a4
AUD0VOL		equ	$0a8
AUD1LC		equ	$0b0
AUD1LEN		equ	$0b4
AUD1VOL		equ	$0b8
AUD2LC		equ	$0c0
AUD2LEN		equ	$0c4
AUD2VOL		equ	$0c8
AUD3LC		equ	$0d0
AUD3LEN		equ	$0d4
AUD3VOL		equ	$0d8
	endc	;AUD0LC

; Audio channel registers
	ifnd	AUDLC
AUDLC		equ	0
AUDLEN		equ	4
AUDPER		equ	6
AUDVOL		equ	8
	endc	;AUDLC

; CIA registers
	ifnd	CIAA
CIAA		equ	$bfe001
CIAB		equ	$bfd000
	endc	;CIAA

	ifnd	CIATALO
CIAPRA		equ	$000
CIATALO		equ	$400
CIATAHI		equ	$500
CIATBLO		equ	$600
CIATBHI		equ	$700
CIAICR		equ	$d00
CIACRA		equ	$e00
CIACRB		equ	$f00
	endc	;CIATALO

Which would catch everything in my old ".inc" file that I was using until I switched to CBM includes. But it still will fail for some asmone/include everything in one file type people.

Separate objects/linking files is better aside ( ) Maybe it would be easier just to redefine all those values as something like

__CUSTOM equ	$dff000
...
__AUD0LC		equ	$0a0
...
__CIATALO		equ	$400

Edit: I've updated the source with the first option.

[DanScott]

My solution was to just delete them from the player source

Anyhow.. the MINIMAL setting lopped off about 8k from my test.exe Every little helps...

[Antiriad_UK]

Originally Posted by DanScott

My solution was to just delete them from the player source

I'd be disappointed with anything less

[h0ffman]

Found a bug. The 9 command isn't being honoured when using sample 0. The trick allows for playing a sample offset at different volumes. It works by running a setup line first with no note and the command with the instrument number. Playing a note with sample 0 then plays the sample at the previously set offset. For example..

--- 5 910
D-3 0 C20

There's a few variants to this aswell. I'll roll up a demo mod with them in tomorrow.

[phx]

Originally Posted by h0ffman

It works by running a setup line first with no note and the command with the instrument number. Playing a note with sample 0 then plays the sample at the previously set offset.

From looking at the source I would say it cannot work, because command 9 is only executed when a note was given with it. And unless I made a stupid mistake I should have copied that behaviour from the official V2.3B player source.

Which players or trackers are known to work like you said?
Although, probably I will implement it neverthess, as long as it doesn't break other modules.

There's a few variants to this aswell. I'll roll up a demo mod with them in tomorrow.

Thanks.

[h0ffman]

Yeah it does look that way in the source but actually when it's used with a note the command gets executed twice. It plays the note at the offset and subsequently moves the offset on again.

Pretty sure all the versions of protracker have this "feature" including the original replay source. Bubsys clone also supports this along with the player 6.1

[daxb]

Originally Posted by h0ffman

--- 5 910
D-3 0 C20

Usually, I quite often used the opposite command order like this (PT3.62):

--- 5 C20
D-3 0 910

However, I remember that PT2.x vs. PT 3.x handled (non-) existence of sample number differently in some cases.

[h0ffman]

Originally Posted by daxb

However, I remember that PT2.x vs. PT 3.x handled (non-) existence of sample number differently in some cases.

I don't trust any of the v3's. I'm still using 2.3d

[phx]

I think I made a mistake and optimised that feature away.
According to the original PT2.3B source it should work without a note.

I got nothing to reproduce the problem, but maybe you want to test my fix from the latest beta-archive:
http://sun.hasenbraten.de/~frank/TEST/ptplayer61beta.lha

[girv]

- New function mt_loopfx() for playing looped sound effects.

Just what I needed! Many thanks Using it in Turbo Tomato with modifications to set up the interrupts via the OS.

[h0ffman]

Originally Posted by phx

I think I made a mistake and optimised that feature away.
According to the original PT2.3B source it should work without a note.

I got nothing to reproduce the problem, but maybe you want to test my fix from the latest beta-archive:
http://sun.hasenbraten.de/~frank/TEST/ptplayer61beta.lha

thanks pal, i've plugged it in and the particular scenario I had is now working perfectly. I've been a bit tied up last few days but will try and create a proper test module sometime this week.

Also double thanks for implementing the master volume changes (which I think I missed?). I've got smooth transitions between mods now.

side note, I forgot I modified the mt_init to use a unified sample pack for all modules. i built a little app which finds all the unique samples (by name) in a colletion of modules, exports a single sample pack and adds references inside each module. Guess I'm mentioning it in case there was anybody else here who might need that? It just made the process of making the music for my game a lot easier as each mod can be written separately but also share samples.

[phx]

Originally Posted by h0ffman

thanks pal, i've plugged it in and the particular scenario I had is now working perfectly.

Thanks for the confirmation!

Also double thanks for implementing the master volume changes (which I think I missed?). I've got smooth transitions between mods now.

Nice. So all the latest modifications are confirmed working. Perhaps I can move towards an official release now.

side note, I forgot I modified the mt_init to use a unified sample pack for all modules. i built a little app which finds all the unique samples (by name) in a colletion of modules, exports a single sample pack and adds references inside each module. Guess I'm mentioning it in case there was anybody else here who might need that?

Do you have portable source code to share? Indeed, maybe somebody is interested.

But I don't understand why you have to change something in mt_init for it? mt_init allows the separate specification of module- and sample-pointers.

[h0ffman]

So, its a little more complicated I guess. My tool runs through the following steps.

1. optimise all modules, removing unused patterns and samples
2. makes a list of all distinct samples and saves them in an indexed pack
3. using the index, it runs through all the modules replacing the first 4 bytes of the sample name with the index value within the sample pack
4. each module is then saved as pattern data only
5. on mt_init, I pass in the base address of the pack, as the init runs through each sample setting pointers, it grabs the index id from the sample name and uses that to find the required sample within the pack. the found sample is then loaded into the sample pointers.

so basically without the tool its not really useable. once the game is done I'd happily wrap it up into something useable it anybody needs it.

[phx]

Originally Posted by h0ffman

5. on mt_init, I pass in the base address of the pack, as the init runs through each sample setting pointers, it grabs the index id from the sample name and uses that to find the required sample within the pack. the found sample is then loaded into the sample pointers.

Ok, I understand. This is the additional logic needed in mt_init. But probably too special to make it part of the default ptplayer.

once the game is done I'd happily wrap it up into something useable it anybody needs it.

[h0ffman]

hey Phx

Found a bug in the master volume control where it was effecting the volume of sound effects during my fades. I've done the following below which seems to have resolved it but I'm not sure it's correct.

EDIT: OK, it's not quite working I don't think. I'm not sure how you detect if a channel is being used for an SFX or not.

EDIT2: I'm an idiot.. code fixed

                  move.w     #$4000,INTENA(a6)

	; adapt all channel volumes immediately
                  move.l     a0,mt_MasterVolTab(a4)

                  tst.b      mt_chan1+n_sfxpri(a4)
                  bne         .1                                                                 ; no sfx playing anyway
                  move.w     mt_chan1+n_volume(a4),d0
                  move.b     (a0,d0.w),d0
                  move.w     d0,AUD0VOL(a6)
.1
                  tst.b      mt_chan2+n_sfxpri(a4)
                  bne        .2                                                                 ; no sfx playing anyway
                  move.w     mt_chan2+n_volume(a4),d0
                  move.b     (a0,d0.w),d0
                  move.w     d0,AUD1VOL(a6)
.2
                  tst.b      mt_chan3+n_sfxpri(a4)
                  bne        .3                                                                 ; no sfx playing anyway
                  move.w     mt_chan3+n_volume(a4),d0
                  move.b     (a0,d0.w),d0
                  move.w     d0,AUD2VOL(a6)
.3
                  tst.b      mt_chan4+n_sfxpri(a4)
                  bne        .4                                                                 ; no sfx playing anyway
                  move.w     mt_chan4+n_volume(a4),d0
                  move.b     (a0,d0.w),d0
                  move.w     d0,AUD3VOL(a6)
.4
                  move.w     #$c000,INTENA(a6)

[phx]

Yes, that's a bug. And a perfect fix. Thanks!

[no9]

Originally Posted by phx

- E8 command can be used as a trigger for your main program.

Hi @phx! Do you think it is possible to have more such triggers exposed, like for certain samples played or note values?

[phx]

Originally Posted by no9

Do you think it is possible to have more such triggers exposed, like for certain samples played or note values?

_mt_playfx returns a pointer to the channel status structure, which includes everything. So you could theoretically call _mt_playfx with a dummy effect for all four channels to acquire their status pointers. Or just export the symbols mt_chan1, mt_chan2, mt_chan3 and mt_chan4.

Although, I cannot recommend that, because I don't want to guarantee that the offsets in the channel status structure remain the same in future versions.

[no9]

I will look into this anyway , thanks!