Flanger code, starting point

[gary00k]

Hi everyone. I found nice flanger code at spinsemi site. I modified it to make flanger depth controlled manually. It should increase delay (make fx frequency lower) with pot increasing and it works like that. The problem is, even if pot1 is set to 0 at the beginning, flanger goes itself from low to high frequency and just then I can control it by myself. Is it possible to make it starting directly from pot value?

Code: Select all

;pot1 = flange rate

mem	fladel	512

equ	fladel_138  fladel+138	;mid pointer to flange delay
equ	fladel_139  fladel+139	;mid pointer to flange delay +1

equ	tri	reg0	;triangle made from ramp
equ	flaout	reg1	;mixed flanger output
equ	fladout	reg2	;flange delay output
equ	k1	reg3	;first coefficient for linear interpolation
equ	k2	reg4	;second coefficient for linear interpolation
equ	temp	reg5	;general temp register

; Inputs control
rdax	fladout, 0.6		; get fraction of flange output	
rdax	adcl, 0.5			; add inputs * 0.5 each
rdax	adcr, 0.5	
wra	fladel, 0			; write flange delay input, clear acc

; Flanger control
sof	0, -0.13
rdax	pot0, 0.13
sof	0.00015, 0		; scale to filter
rdax 	tri, 0.9999		; filter triangle waveform so that pitch bend does not get too great at high rates
wrax	tri, 0			; write triangle wave to register, clear accumulator

; Formulate delay pointer,
or	fladel_138 < 8		; fladel^ + 138	;get midpoint address pointer
rdax	tri, 0.03125		; add triangle wave modulation, scaled to fit delay range
wrax	addr_ptr, 1		; establish address for lower interpolation sample

; Develop fraction of tri:
and	0x0000FF		; mask off integer portion of address, leaving a fractional value in the lowest acc byte
sof 	-2,0			; these operations shift the resulting fractional value to the range 0.0 to 0.999...
sof 	-2,0			; only -2.0 is exact, but it changes sign of shifted value
sof 	-2,0
sof 	-2,0
sof	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof	-2, 0
sof 	-1, 0			; 15 shifts
wrax	k2, 0			; write result as coefficient for second sample read
or	0x7FFFFF
rdax	k2, -1
wrax	k1, 0			; write result as coefficient for first sample read

; Read from first pointer:
rmpa	1			; read memory for first sample read
mulx	k1			; multiply by K1	
wrax	temp, 0			; and store in temp, while clearing the acc

; Get second pointer:
or	fladel_139 < 8		; fladel^ + 139	;form second pointer
rdax	tri, 0.03125		; add triangle waveform again	
wrax	addr_ptr, 0		; establish address for upper interpolation sample
rmpa	1			; read second interpolation sample
mulx	k2			; multiply by K2
rdax	temp, 1			; add temp (first value*K1)
wrax	fladout, 1		; write the result to the flanger delay output
rda	fladel, 1			; add the input to the flanger delay
wrax	flaout, 0			; write result to flaout and clear acc

rdax	flaout, 1			; add flaout
wrax	dacr, 1			; write to right output, save accumulator
wrax	dacl, 0			; write to left output, clear acc

[frank]

The problem is, even if pot1 is set to 0 at the beginning, flanger goes itself from low to high frequency and just then I can control it by myself. Is it possible to make it starting directly from pot value?

I assume you mean pot0 as pot1 does not appear in the code.

Try doing a skp run at the start of the code and initialize tri to a value based on the pot

[gary00k]

I assume you mean pot0 as pot1 does not appear in the code.

Yes, my bad

Try doing a skp run at the start of the code and initialize tri to a value based on the pot

Unfortunately, that didn't change anything.

Code: Select all

;pot0 = flange rate

mem	fladel	512

equ	fladel_138  fladel+138	; mid pointer to flange delay
equ	fladel_139  fladel+139	; mid pointer to flange delay +1

equ	tri	reg0		; triangle made from ramp
equ	flaout	reg1		; mixed flanger output
equ	fladout	reg2		; flange delay output
equ	k1	reg3		; first coefficient for linear interpolation
equ	k2	reg4		; second coefficient for linear interpolation
equ	temp	reg5		; general temp register

; Inits
skp	run, 2
sof	0, -0.0000195		; ( -0.13 + ( 0 * 0.13 ) ) * 0.00015
wrax	tri, 0

; Inputs control
rdax	fladout, 0.6		; get fraction of flange output	
rdax	adcl, 0.5			; add inputs * 0.5 each
rdax	adcr, 0.5	
wra	fladel, 0			; write flange delay input, clear acc

; Flanger control
sof	0, -0.13
rdax	pot0, 0.13
sof	0.00015, 0		; scale to filter
rdax 	tri, 0.9999		; filter triangle waveform so that pitch bend does not get too great at high rates
wrax	tri, 0			; write triangle wave to register, clear accumulator

; Formulate delay pointer,
or	fladel_138 < 8		; fladel^ + 138	;get midpoint address pointer
rdax	tri, 0.03125		; add triangle wave modulation, scaled to fit delay range
wrax	addr_ptr, 1		; establish address for lower interpolation sample

; Develop fraction of tri:
and	0x0000FF		; mask off integer portion of address, leaving a fractional value in the lowest acc byte
sof 	-2,0			; these operations shift the resulting fractional value to the range 0.0 to 0.999...
sof 	-2,0			; only -2.0 is exact, but it changes sign of shifted value
sof 	-2,0
sof 	-2,0
sof	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof	-2, 0
sof 	-1, 0			; 15 shifts
wrax	k2, 0			; write result as coefficient for second sample read
or	0x7FFFFF
rdax	k2, -1
wrax	k1, 0			; write result as coefficient for first sample read

; Read from first pointer:
rmpa	1			; read memory for first sample read
mulx	k1			; multiply by K1	
wrax	temp, 0			; and store in temp, while clearing the acc

; Get second pointer:
or	fladel_139 < 8		; fladel^ + 139	;form second pointer
rdax	tri, 0.03125		; add triangle waveform again	
wrax	addr_ptr, 0		; establish address for upper interpolation sample
rmpa	1			; read second interpolation sample
mulx	k2			; multiply by K2
rdax	temp, 1			; add temp (first value*K1)
wrax	fladout, 1		; write the result to the flanger delay output
rda	fladel, 1			; add the input to the flanger delay
wrax	flaout, 0			; write result to flaout and clear acc

rdax	flaout, 1			; add flaout
wrax	dacr, 1			; write to right output, save accumulator
wrax	dacl, 0			; write to left output, clear acc

[frank]

Try doing a skp run at the start of the code and initialize tri to a value based on the pot

Unfortunately, that didn't change anything.

Code: Select all

;pot0 = flange rate
; Inits
skp	run, 2
sof	0, -0.0000195		; ( -0.13 + ( 0 * 0.13 ) ) * 0.00015
wrax	tri, 0

I do not see the pot being used in the init code, you need to read it and use it.

[gary00k]

Okay, I thought it was simply about entering the value I want to start with.
Still without change.

Code: Select all

;pot0 = flange rate

mem	fladel	512

equ	fladel_138  fladel+138	; mid pointer to flange delay
equ	fladel_139  fladel+139	; mid pointer to flange delay +1

equ	tri	reg0		; triangle made from ramp
equ	flaout	reg1		; mixed flanger output
equ	fladout	reg2		; flange delay output
equ	k1	reg3		; first coefficient for linear interpolation
equ	k2	reg4		; second coefficient for linear interpolation
equ	temp	reg5		; general temp register

; Inits
skp	run, 4
sof	0, -0.13
rdax	pot0, 0.13
sof	0.00015, 0
wrax	tri, 0	

; Inputs control
rdax	fladout, 0.6		; get fraction of flange output	
rdax	adcl, 0.5			; add inputs * 0.5 each
rdax	adcr, 0.5	
wra	fladel, 0			; write flange delay input, clear acc

; Flanger control
sof	0, -0.13
rdax	pot0, 0.13
sof	0.00015, 0		; scale to filter
rdax 	tri, 0.9999		; filter triangle waveform so that pitch bend does not get too great at high rates
wrax	tri, 0			; write triangle wave to register, clear accumulator

; Formulate delay pointer,
or	fladel_138 < 8		; fladel^ + 138	;get midpoint address pointer
rdax	tri, 0.03125		; add triangle wave modulation, scaled to fit delay range
wrax	addr_ptr, 1		; establish address for lower interpolation sample

; Develop fraction of tri:
and	0x0000FF		; mask off integer portion of address, leaving a fractional value in the lowest acc byte
sof 	-2,0			; these operations shift the resulting fractional value to the range 0.0 to 0.999...
sof 	-2,0			; only -2.0 is exact, but it changes sign of shifted value
sof 	-2,0
sof 	-2,0
sof	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof	-2, 0
sof 	-1, 0			; 15 shifts
wrax	k2, 0			; write result as coefficient for second sample read
or	0x7FFFFF
rdax	k2, -1
wrax	k1, 0			; write result as coefficient for first sample read

; Read from first pointer:
rmpa	1			; read memory for first sample read
mulx	k1			; multiply by K1	
wrax	temp, 0			; and store in temp, while clearing the acc

; Get second pointer:
or	fladel_139 < 8		; fladel^ + 139	;form second pointer
rdax	tri, 0.03125		; add triangle waveform again	
wrax	addr_ptr, 0		; establish address for upper interpolation sample
rmpa	1			; read second interpolation sample
mulx	k2			; multiply by K2
rdax	temp, 1			; add temp (first value*K1)
wrax	fladout, 1		; write the result to the flanger delay output
rda	fladel, 1			; add the input to the flanger delay
wrax	flaout, 0			; write result to flaout and clear acc

rdax	flaout, 1			; add flaout
wrax	dacr, 1			; write to right output, save accumulator
wrax	dacl, 0			; write to left output, clear acc

[gary00k]

Any aditional suggestions, please?

[frank]

Any aditional suggestions, please?

You will have to go through the code line by line and find where you need to make the changes. Use something like Excel to calculate the values each sample period and graph them to see what is going on and what need to change.

[gary00k]

I was afraid of that but ok, tell me please, does "fladel_139 < 8" means to shift fladel_139 value 8 places left?

[frank]

I was afraid of that but ok, tell me please, does "fladel_139 < 8" means to shift fladel_139 value 8 places left?

Yes, shift of 8 bits left

[gary00k]

Okay. One more thing: is there a easier way to shift ACC value by X bits? I mean to simplify this:

Code: Select all

and	0x0000FF
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof 	-2,0
sof	-2, 0
sof 	-1, 0

to something like this:

Code: Select all

and	0x0000FF < 15

Leave the last byte and shift it 15 bits left. Or maybe some trick with log/exp to rise the value after mask?
You know, this multiplications uses at least 15 lines of code.

[frank]

The log might help (the "and" with the <15 will not work) but I have never done it. I would try something like:

mask with the AND 0x0000FF
take log of result
add 0x710000
take exp