This is an offshoot of my tabla question. How would you go about creating a sitar string sound in Rack using free modules and no 3rd party samples? Here is my best, mostly using Surge XT Wavetable VCOs and Jawari.
I am curious to hear and see how others would model the sitar sound specifically; not necessarily using this patch. Like adding Karplus Strong in somehow?
Pseudoraga Sitars and Tabla-1.vcv (84.8 KB)
I kind of accidentally created a Sitar-esque patch using
2023-03-08.vcv (1.0 MB)
It this case it was a combination of envelope mod on a Surge XT Filter (OB-Xd 24 dB model) and the Surge Xt Wavetable VCO (Harmonizer Morph 1 wavetable). In upper registers it has a pan drum kind of sound, the middle octaves have that signature Sitar timbral bloom.
Sitar would be brighter and more sustained, and probably could do with a hint of resonator added in.
A while back, I tried to physically model a sitar using Rings⌠well, 20 of them to simulate all the resonating strings. It didnât sound anything like a sitar, but I still had fun.
20rings.vcv (8.6 KB)
edit: I used VCV Chords, not a free module.
Vult Opulus has 2 Sitar presets, need some tweaking of envelopes but can sound sort of sitarishâŚ
Surge string VCO would be my pick.
Not bad for an accident.
Surge XT String VCO and Resonator (Kriskros setting) are producing some very nice sitar-like sounds. Thanks for the suggestion.
I will try as many suggestions as are posted here, more or less.
Ah, Vult Opulus sitar sounds very good! Thanks for suggesting it.
You are right, doesnât sound like a sitar but sounds interesting. Thanks.
Does this patch have an Air Windows pre-release module in it?
Here is my take on using Surge XT Strings as sitar-ish sounds with Ohmer Modules QuadPercs at tabla-ish sounds. Jawari Ahir Bhairav is used in the quantizers.
Pseudoraga Surge XT Quantized Strings Sitars and Tabla-2.vcv (13.2 KB)
Here is my take on using Vult Opulus as sitar-ish sounds with Ohmer Modules QuadPercs as tabla-ish sounds. Jawari Ahir Bhairav is used in the quantizers.
About emulating that Sitar. I guess, If starting from scratch, I would go for FM/PM. Some V/Oct toned operators to do the strings. ButâŚhow to emulate those sympathic resonating strings? We could try to stick to FM and introduce some fixed frequency operators for that.
OrâŚwe go the Physical Modeling wayâŚ
With some quick fiddling I got pretty nice results from TyrannosaurusRu Pluck as the plucked string playing into Audible Instruments Resonator in Sympathetic Strings mode.
Some fiddling with the Resonator Frequency (I settled for 47 and a bit), the Structure (I settled around 0.33), the brightness (approx .88), the damping (approx 0.78) and position (approx 0.27).
Finally mixing Odd and Even and the raw output of Pluck into Bogaudio MIX4. Odd and Even a bit to L/R and at some -7.5 dB and Pluck in the middle at 0 dB.
I tend to use Squinky Labs Colors as the Excitation noise source, because you can easily set and modulate the noise color.
The pitch is only going to Pluck, because the Sympathetic string just resonate at fixed frequencies. Using VCV CV Mix to mix the pitch (at 100%) with the output of an envelope (at some 10%) that is triggered at random to emulate the pitch bending.
Some long Reverb from Valley Plateau to put the whole thing in its virtual space.
Pretty satisfactory for a quick and dirty first try. Itâs plucking and droning away happily, randomly triggered from Grayscale Permutation (6hp) with a Audible Instruments Bernoulli Gate in the âpitch bendâ gate path.
For those who are interested. Hereâs a quick and dirty video and the patch.
About sitar physics in general (as I understand it):
You need one monophonic âinstrumentâ that is tuneable, plucked (so short attack) and sometimes pitch-bent.
A second instrument is/are the sympathetic strings that have a slow attack and release, and a limited number of fixed (non-bending, as @kwurqx mentioned) pitches. They only turn on when the plucked note of corresponding frequency is played, so not all plucked notes will set off a sympathetic string.
Something Iâm not sure of is how much a plucked note will set off an additional sympathetic string that is tuned an octave higher or lower than the plucked noteâs fundamental.
Would it work with a vocoder, like Frozen Wastelandâs âMr. Blue Skyâ?
Itâs been a while since Iâve tried using any vocoder, but I think you want the plucked note going into the MOD input, and up to 16 monophonic (triangle waves?) drones (each tuned to one of the sympathetic strings) going into the various CAR IN inputs.
Arpeggios? Uh . . . I guess.
I have an electric gitar that is almost certainly badly out of tune after not playing it in several years.
I also need to implement string bend. I last tried that for electric guitar sound in Rack a couple of years ago. It is really difficult to attain a natural sound.
I donât think so. But arenât there numerous âtuned resonatorâ modules? I think thatâs more what you are describing.
Vult Rescomb may be a good fit.
Hmm⌠Wikipedia sez: âAnalog vocoders typically analyze an incoming signal by splitting the signal into multiple tuned frequency bands or ranges.â
Doesnât that sound like your description? I guess the only thing to do is to try it.
On another tack â one could have a V/Oct CV that is quantized, then âbentâ before it hits the main VCO. Have a parallel path with âunbentâ CV going to the sympathetic string VCO. Then . . . ( 
okay, it needs more work).
Thatâs what I did two years ago for guitar. It works, but is very touchy to get right.
Maybe I can find my V1 patch to see what I did.
[quote=âk-chaffin, post:15, topic:19631, full:trueâ]I also need to implement string bend. I last tried that for electric guitar sound in Rack a couple of years ago. It is really difficult to attain a natural sound. [/quote]
Yeah, Iâve been thinking about simulating a pedal steel guitar tone. It sounds so pure, yet has a distinctive timbre. For any arpeggios during a slide, all the bends have to track each other. Waiddaminnit â thatâs probably a simple addition of the same voltage to every CV.
