Having trouble with C distortion poly reset

[Synthpopalooza]

OK, so I decided to try an experiment:  16-bit C distortion using mod 5 frequencies.  These will either produce a 33% duty cycle wave, or silence.  The 16-bit notes are playing on first two channels, with first channel silent, second one playing Distortion C @1.79 mhz clocked from first.

 

I tried a poly reset code (RESETPOLY routine in the .s file) to reset the polycounter before each note play, but I think it's not working, as the bass notes will still occasionally drop out.  Any ideas how I can improve the code to get a solid note each time?

 

 

pacland-homeward2.asm pacland-homeward2.s pacland-homeward2.xex

[Rybags]

Is your delay sufficient when doing the SKCTL = 00 then 03 ?   And when setting that low AUDF value, the delay should be considered to start at that point shouldn't it ?

 

And how are you getting the desired/stable type C notes - are you doing the trick to ensure the cycle alignment is right or doing the Pokey Init each time ?

 

Can't say I've really played with this stuff much at all and not for some time either.

[Synthpopalooza]

My routine to reset  the polycounter takes 18 cycles.  It may need more because of 16 bit mode. 

 

I am using mod 5 tones here instead of the 3's so I can more easily hear the instability.   In this case you get either 33% duty cycle, or silence. 

[Synthpopalooza]

My idea was to be able to select three tones at will in 16-bit C distortion. 

 

Using standard 8-bit, there are three tones available in mod 3.  Using channel 3 gets you the 20% duty cycle every time.  But in 16-bit mode, you only get 2nd or 4th channels.  So my idea was mod 3's to get the 01100 and 01011 patterns, and mod 5 to get the 010 pattern which is fairly close in timbre to 01000.  If I can get this to be reliable every time, the rest should fall into place.

Edited August 2, 2021 by Synthpopalooza

[Synthpopalooza]

So I made a few changes.  It's 95% there, but a couple of notes still get dropped.

 

I added in a couple of calls to the poly reset routine RESETTIMER, and played with the frequency settings in the routine.

 

RESETTIMER

	lda #0
    sta SKCTL
    lda #5  
    sta AUDF0 
    nop
	nop
    sta STIMER
    lda #3
    sta SKCTL
	rts

Changing the AUDF0 value seems to shift which sections have dropouts.  Anybody have any other ideas on this?

pacland-homeward2.asm pacland-homeward2.s pacland-homeward2.xex