|Home · Standalone · rtcmix~ · iRTcmix · Tutorials · Reference|
an open-source digital signal processing and sound synthesis language
|about · links · contact|
STRUMFB -- extended Karplus-Strong ("plucked string") physical model
with distortion and feedback
quick syntax:STRUMFB(outsk, dur, AMP, PITCH, FEEDBACK_PITCH, squish, FUND_DECAY, NYQ_DECAY, DISTORTION_GAIN, FEEDBACK_GAIN, CLEANLEVEL, DISTLEVEL[, PAN])
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.
p0 = output start time (seconds) p1 = duration (seconds) p2 = amplitude (absolute, for 16-bit soundfiles: 0-32768) p3 = pitch (Hz or oct.pc *) (see note below) p4 = feedback pitch (Hz or oct.pc *) (see note below) p5 = squish (0-10) p6 = fundamental decay time (seconds) p7 = Nyquist decay time (seconds) p8 = distortion gain (0-100 (or more!)) p9 = feedback gain (0-10 -- values > 1.0 are very 'fed-back') p10 = clean signal level (0.0-1.0) p11 = distortion signal level (0.0-1.0) p12 = pan (0-1 stereo; 0.5 is middle) [optional; default is 0] p2 (amp), p3 (pitch), p4 (feedback pitch), p6 (fundamental decay), p7 (Nyquist decay), p8 (distortion gain), p9 (feedback gain), p10 (clean signal level, p11 (distortion signal level) and p12 (pan) can receive updates from a table or real-time control source. * If the value of p3 or p4 field is < 15.0, it assumes oct.pc. Use the pchcps scorefile convertor for direct frequency specification below 15.0 Hz
STRUMFB is an updated version (updated from the older STRUM family of instruments) of the Karplus-Strong "plucked string" algorithm, with extensions allowing distortion and feedback based upon work by Charles Sullivan. The Karplus-Strong plucked-string algorithm is a subtractive synthesis system featuring a burst of white noise, a recirculating delay line, a lowpass filter, an allpass filter, and a snazzy recursion (see Roads, 1997).
The basic idea is that a burst of noise is pushed through a delay line, which splits its output, sending one half as output and the rest of it back into itself after going through a lowpass and allpass filter setup. The result is a burst of rich sound that gradually loses its higher harmonics as it decays (as does, interestingly enough, a plucked string).
Charlie added an additional delay with waveshaping distortion to produce a simulation of a distorted electric guitar with feedback.
Usage NotesAs noted above, the "PITCH" parameter (p3) can be in Hz or oct.pc form. The decision is based upon the value being < 15.0 (< 15.0 will be interpreted as oct.pc).
The "FEEDBACK_PITCH" parameter (p4) sets up a delay corresponding to the desired pitch for the feedback sound. This does not mean that you will get this pitch as a result. Instead, the feedback will generally align with some harmonic of this pitch. However, not even that is guaranteed. This is a very non-linear parameter. Altering this pitch as a note evolves via the pfield control system can produce interesting changes in the output, kind of like Jimi Hendrix leaning into and away from his amplifier (essentially he was altering the length (pitch) of the feedback "delay line" between the amp and his guitar). This parameter can also be in Hz or oct.pc form, like p3 above.
The "squish" parameter (p5) tells how "squishy" is the item being used to pluck the string. Values are integers ranging from 0 to 10 The lower the value, the harder the plucking object (0 is like plucking with a steel pick). The higher, the more "fleshy" (fat fingers!)
The "FUND_DECAY" parameter (p6) sets the time for the decay of the fundamental frequency in the synthesis algorithm. Usually this is the same as the duration of the note (p1), but shorter times can give a 'damped' effect, where longer times can yield a more sustained string sound. If p6 is > p1, it's generally a good idea to apply an amplitude envelope of some kind to prevent clicks at the end of the note.
The "NYQ_DECAY" parameter (p7) attempts to control the decay rate of the highest, fastest-decaying partials of the sound (i.e. the decay time for frequencies at the top of the frequency range, the Nyquist frequency). This can work up to a point, but realistically our implementation of the algorithm doesn't allow for much independence between the fundamental decay and Nyquist decay. Theoretically the ratio of the "funddecay" and "nyqdecay" parameters effect the brightness of the plucked note, but this effect isn't too noticeable. The "squish" parameter does a better job of altering the plucked timbre.
"FEEDBACK_GAIN" (p9) is a very sensitive parameter. You can set this to virtually any value you want, but high values don't necessarily produce much change. Often very low values (< 0.01) are necessary for subtle feedback sounds.
Simularly, "DISTORTION_GAIN" (p8) can be set arbitrarily (we can go waaaaay 'beyond 11'), but at some point it won't make much difference.
The "CLEANLEVEL" (p10) and "DISTLEVEL" (p11) can be set to mix between the 'straight' Karplus-Strong sound and the distorted sound, if desired.
STRUMFB can produce either mono or stereo output.
Sample Scoresbasic use:
rtsetparams(44100, 2) load("STRUMFB") dur = 7 freq = 6.08 decaytime = 1 nyqdecaytime = 1 distgain = 10 fbgain = 0.05 cleanlevel = 0 distlevel = 1 amp = 20000 squish = 2 start = 0 fbfreq = 7.00 STRUMFB(start, dur, amp, freq, fbfreq, squish, decaytime, nyqdecaytime, distgain, fbgain, cleanlevel, distlevel) start = 8 fbfreq = 7.01 STRUMFB(start, dur, amp, freq, fbfreq, squish, decaytime, nyqdecaytime, distgain, fbgain, cleanlevel, distlevel) start = 16 fbfreq = 6.01 STRUMFB(start, dur, amp, freq, fbfreq, squish, decaytime, nyqdecaytime, distgain, fbgain, cleanlevel, distlevel)
slightly more advanced:
rtsetparams(44100, 2) load("STRUMFB") dur = 7.5 amp = 10000 pitch = cpspch(6.08) pitch2 = cpspch(7.00) freq = maketable("line", "nonorm", 1000, 0,pitch, 4,pitch, 6,pitch2, 8,pitch) reset(200) decaytime = 1 nyqdecaytime = 1 distgain = 10 fbgain = 0.05 fbpitch = 7.00 squish = 2 cleanlevel = 0 distlevel = 1 STRUMFB(0, dur, amp, freq, fbpitch, squish, decaytime, nyqdecaytime, distgain, fbgain, cleanlevel, distlevel)
rtsetparams(44100, 2, 256) load("STRUMFB") dur = 7 pitch = 7.07 decaytime = 1 nyqdecaytime = 1 distgain = 10 fbgain = 0.05 fbpitch = 7.00 cleanlevel = 0 distlevel = 1 amp = 10000 squish = 2 viblow = 4 vibhigh = 7 vibdepth = 5 // Hz vibseed = 0 lfreq = makerandom("low", frq=10, viblow, vibhigh, vibseed) vibenv = maketable("line", 1000, 0,0, 1,1, 2,0) vib = makeLFO("sine", lfreq, vibdepth) * vibenv freq = cpspch(pitch) + vib env = maketable("line", 1000, 0,1, 19,1, 20,0) reset(500) STRUMFB(0, dur, amp * env, freq, fbpitch, squish, decaytime, nyqdecaytime, distgain, fbgain, cleanlevel, distlevel)
See Alsomaketable, makerandom, makeLFO, MSITAR, STRUM, STRUM2