This function implements a chorus. It is mono, taking in data from the left channel, and presenting the mono output on both the left and right channels.
As with a vocal choir, there are a number of voices, all ever so slightly out of time and pitch from one another. This function attempts to replicate this effect by pitch shifting and delaying the input signal, to create a slightly out of time signal, which can be mixed back in with the original. This function adds two such voices, each with a different, predetermined delay. The easiest way to accomplish the pitch shifting is to add in a time varying delay, so at any point, the samples are being read back at a slower or faster rate. But, since its moving back and forth, there is no buffer boundary to deal with. Any time-varying signal can be used, but in this case, we are using a sinusoidal signal, as it introduces the least amount of added frequency content. The two voices generated by this chorus effect use the opposite sides of the same sinusoid, so as one voice is pitching up, the other is pitching down. This sinusoid is created by taking values from a look-up table stored in program memory. The frequency of the sinusoid is set by the pot (MOD1), and is variable from .084Hz to 10.84Hz.
The depth of the low frequency oscillator (LFO), and the resulting pitch shifting of the choir, can be modified by the rotary encoder (MOD2). Rotations to the right increase its amplitude (from 0ms to 6ms). The rotary encoder increments a value from 0-255 at the rate of 1 per click (code lines 350 and 357). As a result, it takes quite a few turns to go from one end to the other, however, this also gives the user very fine control over the resolution.
Open Music Labs Wiki; program: chorus-16b-sine.asm
; UID = 000042 - unique id to eliminate conflicts between variables
; 16b address space
; mono data in on left channel, identical stereo out
; pot (MOD1) controls lfo frequency
; rotary encoder (MOD2) controls lfo depth
; program overview
;
; data is read in from the codec and stored to sram. this data is then read
; back out at a time varying delay. the average delay time is a fixed number
; set at the beginning of the code, and is varied with a sinusoidal lfo. the
; lfo is generated via interpolating a 16b 512s half sinewave lookup table.
; a 32b number is used to index into this lookup table, to allow for very
; slow lfo rates. this lfo is then multiplied by a 16b amplitude signal,
; and the data is fetched from the sram, at that location. the adc is
; oversampled 256 times and deadbanded before updating the lfo rate. the
; rotary encoder is used to increment the amplitude of the lfo. there are
; two voices, each swept with the same lfo signal, but in opposite
; directions. the average delay for each is independent.
; constants
;
.equ delay1_000042 = $0321 ; chorus average delay time for voice 1
; (1/44.1 ms per unit), min value $0100
.equ delay2_000042 = $0444 ; chorus average delay time for voice 2
; (1/44.1 ms per unit), min value $0100
; register usage - may be redefined in other sections
;
; r0 multiply result lsb
; r1 multiply result msb
; r2 accumulation lsb
; r3 accumulation mlb
; r4 right/left lsb out/accumulation mhb
; r5 right/left msb out/accumulation msb
; r6 left lsb in
; r7 left msb in
; r8 adc accumulator fractional byte
; r9 adc accumulator lsb
; r10 adc accumulator msb
; r11 rotary encoder counter
; r12 lfo rate lsb
; r13 lfo rate msb
; r14 null register
; r15 switch sample counter
; r16 temporary swap register
; r17 temporary swap register
; r18 sine wave buffer/multiply msb
; r19 sine wave buffer/multiply msb
; r20 multiply swap register
; r21 multiply swap register
; r22 sinetable lookup address lsb
; r23 sinetable lookup address mlb
; r24 write address lsb
; r25 write address msb
; r26 sinetable lookup address mhb
; r27 sinetable lookup address msb
; r28 temporary swap register
; r29 lfo depth
; r30 jump location for interrupt lsb
; r31 jump location for interrupt msb
; t rotary encoder edge detect indicator
; program starts here first time
; intialize registers
ldi r30,$04 ; increment z pointer to new jump location
clr r14 ; clear null register
ldi r29,$0e ; initialize lfo depth
reti ; finish with initialization and wait for next interrupt
; program starts here every time but first
; initiate data transfer to codec
sbi portb,portb0 ; toggle slave select pin
out spdr,r5 ; send out left channel msb
cbi portb,portb0
adiw r25:r24,$01 ; increment write address
wait1_000042: ; check if byte has been sent
in r17,spsr
sbrs r17,spif
rjmp wait1_000042
in r7,spdr ; recieve in left channel msb
out spdr,r4 ; send out left channel lsb
wait2_000042: ; check if byte has been sent
in r17,spsr
sbrs r17,spif
rjmp wait2_000042
in r6,spdr ; recieve in left channel lsb
out spdr,r5 ; send out right channel msb
;write left channel datat to sram
out portd,r24 ; set address
sts porth,r25
out portg,r14 ; pull ce low,we low,and set high bits of address
ldi r17,$ff
out ddra,r17 ; set porta as output for data write
out ddrc,r17 ; set portc as output for data write
out porta,r6 ; set data
out portc,r7
sbi portg,portg2 ; pull we high to write
out ddra,r14 ; set porta as input for data lines
out ddrc,r14 ; set portc as input for data lines
wait3_000042: ; check if byte has been sent
in r17,spsr
sbrs r17,spif
rjmp wait3_000042
in r17,spdr ; recieve in right channel msb
out spdr,r4 ; send out right channel lsb
wait4_000042: ; check if byte has been sent
in r17,spsr
sbrs r17,spif
rjmp wait4_000042
in r17,spdr ; recieve in right channel lsb
; vco generation
movw r17:r16,r31:r30 ; store z register
;get sample 1
add r22,r12 ; increment sinetable address
adc r23,r13
adc r26,r14 ; r14 is cleared above
adc r27,r14
movw r31:r30,r27:r26 ; move to z register for data fetch
lsl r30 ; adjust pointer for 16b fetch
rol r31
andi r31,$03 ; limit value to 10b (512 samples x 2 bytes)
ori r31,$48 ; set to memory address location where table is stored
lpm r18,z+ ; get sine value lsb, increment z register
lpm r19,z ; get sine value msb
sbrc r27,$01 ; flip sign for half of the values
rjmp interpolate_000042
neg r18
adc r19,r14 ; r14 is cleared above
neg r19
interpolate_000042: ; multiply sample 1 by distance
movw r21:r20,r23:r22 ; get distance from sample 1
com r20 ; invert distance
com r21
mulsu r19,r21 ; (signed)Ah * (unsigned)Bh - multiply high bytes
movw r5:r4,r1:r0 ; store high bytes result for later
mul r18,r20 ; (unsigned)Al * (unsigned)Bl ; multiply low bytes
movw r3:r2,r1:r0 ; store low byets for later
mulsu r19,r20 ; (signed)Ah * (unsigned)Bl - multiply middle bytes
sbc r5,r14 ; r14 is cleared above - subtract sign bit
add r3,r0 ; accumulate result
adc r4,r1
adc r5,r14 ; r14 is cleared above
mul r21,r18 ; (unsigned)Bh * (unsigned)Al - multiply middle bytes
add r3,r0 ; accumulate result
adc r4,r1
adc r5,r14 ; r14 is cleared above
;get sample 2
adiw r27:r26,$01 ; set to next sample
movw r31:r30,r27:r26 ; move to z register for data fetch
lsl r30 ; adjust pointer for 16b fetch
rol r31
andi r31,$03 ; limit value to 10b (512 samples x 2 bytes)
ori r31,$48 ; set to memory address location where table is stored
lpm r18,z+ ; get sine value lsb, increment z register
lpm r19,z ; get sine value msb
sbrc r27,$01 ; flip sign for half of the values
rjmp interpolate1_000042
neg r18
adc r19,r14 ; r14 is cleared above
neg r19
interpolate1_000042: ; multiply sample 2 by distance
sbiw r27:r26,$01 ; reset address
movw r31:r30,r17:r16 ; restore z register
mulsu r19,r23 ; (signed)Ah * (unsigned)Bh - multiply high bytes
add r4,r0 ; accumulate result
adc r5,r1
mul r18,r22 ; (unsigned)Al * (unsigned)Bl ; multiply low bytes
add r2,r0 ; accumulate result
adc r3,r1
adc r4,r14 ; r14 is cleared above
adc r5,r14
mulsu r19,r22 ; (signed)Ah * (unsigned)Bl - multiply middle bytes
sbc r5,r14 ; r14 is cleared above - subtract sign bit
add r3,r0 ; accumulate result
adc r4,r1
adc r5,r14 ; r14 is cleared above
mul r23,r18 ; (unsigned)Bh * (unsigned)Al - multiply middle bytes
add r3,r0 ; accumulate result
adc r4,r1
adc r5,r14 ; r14 is cleared above
;set lfo depth
movw r19:r18,r5:r4 ; move lfo signal to multiply register
mov r21,r29 ; move lfo depth to multiply register
mulsu r19,r21 ; (signed)ah * (unsigned)b
movw r5:r4,r1:r0
mul r21,r18 ; (unsigned)b * (unsigned)al
add r4,r1
adc r5,r14 ; r14 is cleared above
;create first voice
;add lfo to delay
mov r2,r4 ; make a backup copy of lfo value for second voice
mov r28,r5
movw r17:r16,r25:r24 ; move current location to read address
subi r16,low(delay1_000042) ; remove delay time
sbci r17,high(delay1_000042)
clr r21 ; prepare to add in lfo time
tst r5 ; check if lfo time is negative
brpl lfoadd_000042 ; add in lfo time if positive
ser r21 ; subtract lfo time if negative
lfoadd_000042: ; add in lfo time
add r16,r5 ; add in lfo time
adc r17,r21
;get left channel sample 1 from sram
out portd,r16 ; set address
sts porth,r17
nop ; wait setup period of two cycles
nop
in r18,pina ; get data
in r19,pinc ; get data
;multiply sample 1 by distance
mov r20,r4 ; get distance from sample 1
com r20 ; invert distance for sample weighting
mulsu r19,r20 ; (signed)ah * b
movw r5:r4,r1:r0
mul r18,r20 ; al * b
add r4,r1
adc r5,r14 ; r14 is cleared above
mov r3,r0
;get left channel sample 2 from sram
subi r16,$ff ; set to next sample
sbci r17,$ff ; done this way because there is no addi or adci
out portd,r16 ; set address
sts porth,r17
nop ; wait setup period of two cycles
nop
in r18,pina ; get data
in r19,pinc ; get data
;multiply sample 2 by distance
com r20 ; reset sample 2 distance
mulsu r19,r20 ; (signed)ah * b
add r4,r0 ; accumulate result
adc r5,r1
mul r18,r20 ; al * b
add r3,r0 ; accumulate result
adc r4,r1
adc r5,r14 ; r14 is cleared above
movw r7:r6,r5:r4 ; store voice 1
;create second voice
;add lfo to delay
mov r4,r2 ; restore lfo value for second voice
mov r5,r28
com r4 ; invert lfo for second voice
com r5
movw r17:r16,r25:r24 ; move current location to read address
subi r16,low(delay2_000042) ; remove delay time
sbci r17,high(delay2_000042)
clr r21 ; prepare to add in lfo time
tst r5 ; check if lfo time is negative
brpl lfoadd1_000042 ; add in lfo time if positive
ser r21 ; subtract lfo time if negative
lfoadd1_000042: ; add in lfo time
add r16,r5 ; add in lfo time
adc r17,r21
;get left channel sample 1 from sram
out portd,r16 ; set address
sts porth,r17
nop ; wait setup period of two cycles
nop
in r18,pina ; get data
in r19,pinc ; get data
;multiply sample 1 by distance
mov r20,r4 ; get distance from sample 1
com r20 ; invert distance for sample weighting
mulsu r19,r20 ; (signed)ah * b
movw r5:r4,r1:r0
mul r18,r20 ; al * b
add r4,r1
adc r5,r14 ; r14 is cleared above
mov r3,r0
;get left channel sample 2 from sram
subi r16,$ff ; set to next sample
sbci r17,$ff ; done this way because there is no addi or adci
out portd,r16 ; set address
sts porth,r17
nop ; wait setup period of two cycles
nop
in r18,pina ; get data
in r19,pinc ; get data
;multiply sample 2 by distance
com r20 ; reset sample 2 distance
mulsu r19,r20 ; (signed)ah * b
add r4,r0 ; accumulate result
adc r5,r1
mul r18,r20 ; al * b
add r3,r0 ; accumulate result
adc r4,r1
adc r5,r14 ; r14 is cleared above
;divide both voices by 2 and add them
asr r7 ; divide voice 1 by 2
ror r6
asr r5 ; divide voice 2 by 2
ror r4
add r4,r6 ; add the two voices
adc r5,r7
;check rotary encoder and adjust lfo depth
; although rotary encoder is externally debounced, it is done here again.
; pin1 is sampled on a transition from high to low on pin0. if pin1 is
; high, a left turn occured, if pin1 is low, a right turn occured.
dec r11 ; check if time to sample rotary encoder
brne adcsample_000042 ; continue if not
ldi r17,$40 ; adjust sample frequency to catch all rising edges (1.5ms)
mov r11,r17
lds r17,pinj ; get switch data
sbrs r17,$00 ; check if pin0 is low
rjmp edge_000042 ; check if pin0 was low on previous sample
clt ; clear state register if back high
rjmp adcsample_000042 ; finish off
edge_000042: ; check for falling edge
brts adcsample_000042 ; do nothing if edge was already detected
set ; set t register to indicate edge detected
sbrs r17,$01 ; check if pin1 is high
rjmp increment_000042 ; increment desired delay if right rotation
subi r29,$01 ; decrement lfo depth register
brcc adcsample_000042 ; check if underflow occured
clr r29 ; set lfo depth to min
rjmp adcsample_000042 ; finish off
increment_000042: ; increment desired delay register
ldi r16,$01 ; increment lfo depth register
add r29,r16
brcc adcsample_000042 ; check if overflow occured
ser r29 ; set lfo depth to max
adcsample_000042: ; sample adc for lfo rate
lds r17,adcsra ; get adc control register
sbrs r17,adif ; check if adc conversion is complete
rjmp done_000042 ; skip adc sampling
lds r16,adcl ; get low byte adc value
lds r17,adch ; get high byte adc value
add r8,r16 ; accumulate adc samples
adc r9,r17
adc r10,r14 ; r14 is cleared above
ldi r17,$f7
sts adcsra,r17 ; clear interrupt flag
dec r15 ; countdown adc sample clock
brne done_000042 ; get delay time if its been long enough
deadband_000042: ; set the low value of the delay
lsr r10 ; divide adc value by 16
ror r9
ror r8
lsr r10
ror r9
ror r8
lsr r9 ; r10 is now empty
ror r8
lsr r9
ror r8
movw r17:r16,r9:r8 ; move adc sample to temporary register
ldi r21,$80 ; add in offset of min lfo rate ($0080)
add r16,r21
adc r17,r14 ; r14 is cleared above
sub r16,r12 ; find difference between adc sample and current lfo rate
sbc r17,r13
brsh check_000042 ; check for deadband if positive
neg r16 ; invert if negative
adc r17,r14 ; r14 is cleared above
neg r17
check_000042: ; check if difference is greater than deadband
cpi r16,$10 ; check if difference is less than 1 adc lsb
cpc r17,r14 ; r14 cleared above
brlo empty_000042 ; do nothing if less than 1 adc lsb
movw r13:r12,r9:r8 ; move adc sample to lfo rate register
add r12,r21 ; add in offset of min lfo rate ($0080)
adc r13,r14 ; r14 is cleared above
empty_000042: ; empty accumulation registers and finish off
clr r8 ; empty accumulation registers
clr r9
clr r10
switchsample_000042: ; check rotary switch
lds r16,pinj ; get switch data
andi r16,$78 ; mask off rotary switch
lsr r16 ; adjust switch position to program memory location
lsr r16
ldi r17,$02
add r16,r17
cpse r16,r31 ; check if location has changed
clr r30 ; reset jump register to intial state
mov r31,r16
done_000042:
reti ; return to waitingLast edited 2010-08-21 by guest · 7 revisions