Oscillator-bank channel vocoder.
in RTcmix/insts/jg
VOCODESYNTH(outsk, insk, dur, AMP, nfilts, CFREQLO/FTABLETRANSP, CFREQMULT/FTABLEFORMAT, transp, filtbw, windowsize, smoothness, threshold, attack}, release, hisigmix, hifreq, inputchan, PAN, WAVETABLE[, SCALINGTABLE, CFREQTABLE])
CAPITALIZED parameters are pfield-enabled for table or dynamic control (see the maketable or makeconnection scorefile commands). Parameters after the [bracket] are optional and default to 0 unless otherwise noted.
| Param Field | Parameter | Units | Dynamic | Optional | Notes |
|---|---|---|---|---|---|
| p0 | output start time | seconds | no | no | |
| p1 | input start time | seconds | no | no | |
| p2 | duration | seconds | no | no | |
| p3 | amplitude multiplier | relative multiplier of output signal | yes | no | |
| p4 | number of filters | - | no | no | |
| p5 | if p4 > 0, lowest filter center frequency | Hz/oct.pc | no | no | |
| if p4 == 0, transposition of function table* | oct.pc | yes | no | * reference to pfield table-handle | |
| p6 | if p4 > 0, center frequency spacing multiplier | >1 | yes | no | |
| if p4 == 0, function table format* | 0: octave.pc, 1: linear octave; Hz if > 15.0 | yes | no | * reference to pfield table-handle | |
| p7 | amount to transpose carrier oscillators | Hz/oct.pc | no | no | |
| p8 | filter bandwidth proportion of center frequency | > 0 | no | no | |
| p9 | power gauge window length | seconds, try 0.01 | no | no | |
| p10 | smoothness – how much to smooth the power gauge output | 0-1, try 0.5 | no | no | |
| p11 | threshold – below which no synthesis for a band occurs | 0-1, try 0.0 | no | no | |
| p12 | attack time – how long it takes the oscillator for a band to turn on fully once the modulator power for that band rises above the threshold | seconds, try 0.001 | no | no | |
| p13 | release time – how long it takes the oscillator for a band to turn off fully once the modulator power for that band falls below the threshold | seconds, try 0.01 | no | no | |
| p14 | amount of high-passed modulator signal to mix with output | amplitude multiplier, try 0.0 | no | no | |
| p15 | cutoff frequency for high pass filter applied to modulator | Hz, try 5000 | no | no | This pfield ignored if p10 is zero. |
| p16 | input channel | - | no | no | |
| p17 | pan | 0-1 stereo; 0.5 is middle | yes | no | |
| p18 | wavetable to use | reference to pfield table-handle | yes | no | |
| p19 | table giving the carrier scaling curve | reference to pfield table-handle (as [freq, amp] pairs) | yes | no | |
| p20 | table giving list of center frequencies | reference to pfield table-handle | yes | no | if p4 is zero |
Parameters labled as Dynamic can receive dynamic updates from a table or real-time control source.
Author: John Gibson, 8/7/03.
VOCODESYNTH performs a filter-bank analysis of the input channel (the modulator), and uses the time-varying energy measured in the filter bands to trigger wavetable notes with corresponding frequencies (the carrier).
This is a bit different from the other VOCODE-family instruments like VOCODE2 and VOCODE3 in that it does a resynthesis of the signal. This instrument is similar in some respects to PVOC, but it is a channel vocoder using a bank of band-pass filters instead of an FFT analysis (like with phase vocoders).
If parameter p4 (“nfilts”) is > 0, then p5 (“CFREQLO/FTABLETRANSP”) is interpreted as the lowest center frequency of the filters. This can be specified in Hz or oct.pc, < 15.0 is the point at which it will be interepreted as oct.pc. p6 (“CFREQMULT/FILTMULT”) will be interpreted as a multiplier of p5 – it will space the center frequencies of all the filters (up to “nftilts”, p4) as a multiple of each succesive center frequency. For example, p6 = 2.0 will make a stack of octaves; p6 = 1.5 will make a stack of perfect (Pythagorian) fifths. To get stacks of an equal tempered interval (in oct.pc), you can use the cpspch convertor:
However, if p4 == 0, then p20 (“CFREQTABLE”) has to be present. It is interpreted as a reference to a table containing the oct.pc center frequencies of all the filters used. Parametet p6 (“CFREQMULT/FTABLEFORMAT”) then serves as a flag for how the table referenced in p20 will be interpreted. If p6 == 0, then the table will be interpreted as containing oct.pc pitch values. If p6 == 1, then the table will be interpreted as linear octave values. NOTE: This only applies if the values in the table ar < 15.0; otherwise they will be interpreted as direct frequency values regardless of the setting of p6. The number of filter bands is determined by length of function table.
An example of this second method for center frequency specification:
Make a table containing the desired center frequencies:
numbands = 5
cftable - maketable("literal", "nonorm", numbands, 8.00, 8.07, 9.00, 9.07, 10.02)
or specify the table as direct frequencies:
numbands = 9
cftable - maketable("literal", "nonorm", numbands, 100, 200, 300, 400, 500, 600, 700, 800, 900)
Set p4 to 0, and pass the “cftable” pfield-handle in through parameter p20. This table will be transposed by the setting given in p5 (“CFREQLO/FTABLETRANSP”).
Parameter p9 (“windowsize”) determines how often changes in modulator power are measured. p10 (“smoothness”) operates upon the output of p9. It has more of an effect for longer p9 values.
Parameters p14 (“hisigmix”) and p15 (“hifreq”) allow some of the original signal to appear in the output. This is helpful in capturing sharp noise transients (such as consonants in speech) that aren’t captured well by standard FFT analysis. p15 is ignored if p10 is set to 0.
p18 (“WAVETABLE”) is a reference to a waveform table that will be used by the internal wavetable oscillators.
p19 (“SCALINGTABLE”) is a table used for scaling the carrier notes, interpreted as [frequency, amplitude] pairs.
VOCODESYNTH produces both stereo or mono output.
one example:
rtsetparams(44100, 2)
load("VOCODESYNTH")
rtinput("mysound.aif")
inskip = 0.0
dur = DUR()
amp = 4.0
env = maketable("line", 1000, 0,0, .1,1, dur-.1,1, dur,0)
numbands = 25
lowcf = 300
interval = 0.025
cartransp = 0.00
bw = 0.009
winlen = 0.001
smooth = 0.98
thresh = 0.0001
atktime = 0.001
reltime = 0.01
hipassmod = 0.0
hipasscf = 2000
carwavetable = maketable("wave", 10, 20000, "sine")
scale1 = 0.5
scale2 = 1.0
scalecurve = maketable("curve", "nonorm", 100, 0,scale1,1, 1,scale2)
spacemult = cpspch(interval) / cpspch(0.0)
VOCODESYNTH(0, inskip, dur, amp * env, numbands, lowcf, spacemult, cartransp,
bw, winlen, smooth, thresh, atktime, reltime, hipassmod, hipasscf,
inchan=0, pan=1, carwavetable, scalecurve)
cartransp += 0.001
spacemult += 0.002
VOCODESYNTH(0, inskip, dur, amp * env, numbands, lowcf, spacemult, cartransp,
bw, winlen, smooth, thresh, atktime, reltime, hipassmod, hipasscf,
inchan=0, pan=0, carwavetable, scalecurve)
maketable, CONVOLVE1, LPCIN, PVOC, SPECTACLE, SPECTACLE2, SPECTEQ, SPECTEQ2, SPECTACLE, VOCODE2, VOCODE3