How would you create a sitar sound in Rack?

Plugins & Modules
81

Indeed, it’s often much about those ‘perfect imperfections’. And in the end it’s alll about perception, which is subjective. And even that is fluid and contextual, depending on physical and mental state(s).

Another physics analogy: The Fourier transform suggests/claims that the spectrum of any sound is defined by the sum of perfect wholenumber multiples of it’s fundamental frequency at the right amplitude (and phase)…where in truth it’s more complicated then that…especially when the time (and space) dimension is added.

Anyway…

On emulating the Sitar using FM/PM…or just to explore…

I started my replies by suggesting FM/PM and you actually did end up using FM/PM (Vult Opulus). In FM/PM you can indeed get pretty convincing end expressive plucked strings (although it’s quite a challenge to emulate the continous resonations of the sympathetic strings).

With FM/PM (as with AM) it’s about sum and difference frequencies of all frequencies in the spectrum of both carrier and modulator, where in FM/FM, the ‘negative’ frequencies reflect back into the spectrum at inverted phase.

Also…FM?PM is very sensitive to exact ratio’s. All sorts of inharmonics may be introduced when deviating for integer ratios. But…this might be just what you need/want in some use cases. Especially because this happens only at pretty controllable/predictable places in the spectrum.

With that in mind you can use additive oscillator(s) as a (added) modulator(s) and/or carrier(s), to control where and how much/many. And thus livening up (parts of) the final spectrum. Be gentle with any FM/PM feedback and complex signals, since this soon results in harder to control/predict complectity in the spectrum.

One more thing: don’t forget you can use STATIC (not following V/Oct or notes) carriers/modulators to emulate ‘fixed’ frequency/bandwidth resonation/resonators.

Depending on the desired control over the additive spectrum / harmonics we have many options:

About the modlator (‘sender’) and carrier (‘receiver’) roles…

All these (or any oscillator) can (of course) be used as modulators where we just need the output signal. But not all oscillators offer PM (or linear FM). There are several options to use any oscillator as a PM carrier, even if the module does not have it’s own input/parameter for phase/frequency. E.g. Sckitam WaveguideDelay

VCV Rack 2 - Phase Modulation using Waveguide Delay

3 Likes
82

Thanks. There is a wealth of info in your post.

83

Yes, but it was @fractalgee who suggested the Opulus and @pgatt who suggested the Rescomb 2. The parameter space is so large for sound synthesis that it very difficult for me to start from parametric principles and design a sound. The specific advice of others is invaluable, as is the low level parametric advice.

84

I’m not trying to take credit for anything. There’s no ‘competition’ or ‘selfpromotion’ element involved from my part. As an analyst by nature and profession, I tend to generalize to the max to gain conceptual multipurpose knowledge and then specialize back into application where applicable.

Your Sitar (and Tabla) post (as well as similar challenges or questions) inspire(d) me (and others) to (further) explore some known and unknown territories.

In this case, for me, specifically in the domain of physical modelling, plucked string, exciter and resonator. All stuff I was (and am) a lot less familiar with as compared to more ‘traditional’ synthesis techniques like subtractive, additive, FM, PM, AM, sync and waveshaping.

My main takeaways from of all this fiddling with plucked strings and sympathetic strings were in the (combined) use of:

  • tuned delay
  • feedback circuits
  • (controlled) (high) (self) resonance

Also the use (resonant) filters in feedback circuits:

  • comb filter (comb filtering effects can be simply achieved by mixing a signal with a delayed (static eddect) or detuned (periodic effect) copy of itself)
  • bandpass filters (mainly to create formants of various amplitude and bandwidth)
  • lowpass filters (for a ‘natural’ spectral decay/dampening of high-frequencies-first)
  • allpass filters (to diffuse the spectrum by phase shifting bands, e.g. as used in many reverbs)

Speaking of filters. For those who just can find the ‘right’ filter for the job. Or who are just curious what knobs there are to twiddle WITHIN a filter design. Maybe twiddle some knobs on this combo:

Or maybe go all the way down the digital-filter-by-delay-mix-and-feedback rabbit hole with Jakub Ciupinski? Concepts that might open yet another universe of possibilities…

VCV Rack Hacks | Jacob’s Ladder

Or use this mutiple-pass-feedback approach to your advantage:

VCV Rack Hacks | Supercharge your VCF

1 Like
85

Yeah, nor was I trying to enforce “credit”. It was more of me trying to explain how I personally take inspiration from information sources.

86

We all build on the shoulders of giants.

1 Like
87

What distinguishes the sound of the sitar. I think there are a few things, some of which have already been discussed here, some of which less so:

• The contact of the string with the bridge. Much longer contact zone than other instruments, which affects the tone dramatically. Different traditions of sitar playing shape their bridges differently to give different sounds, the Ravi Shankar school is more of an open jawari, the Vilayat Khan school is more closed.

• The resonance of the gourd. The gourds are chosen to have very specific resonating properties which are influenced by their thickness and uniformity.

• The sympathetic strings

• The bending. On guitar you can bend maybe one or two half steps? On sitar you can bend (meend) up 5-6, so that allows for both long glides but also very complex ornamentations and related techniques e.g. “gamaks”.

2 Likes
88

So root is C4 unless you transpose it, and the octave range is plus or minus 0,1, or 2 from there. Each octave change (probability determined by the corresponding parameter) is up or down one octave at a time, so octave range just constrains how far it can drift from C4 as root. If you are at C2 you won’t have an octave change that goes down, but could have one that goes up. Does that answer your question?

1 Like
89

I think so. Watching the output with FOUR-VIEW and Proteus RANGE set to 2, I see notes ranging from C2 to to C6. I see octave changes of up to 2 per step, which does not seem to agree with what you said above, but maybe I misunderstood. Specifically:

Thanks for explaining.

90

That’s definitely unexpected behavior. I will have a look.

91

I used a pair of comparators and s&h modules to track the highest and lowest notes output by Proteus over time. Below is a screen capture that shows that with a RANGE=2 and no transposition, the low note output was G1 and the high note was C7. I would imagine that Proteus may be much like Meander’s fBm range that is only loosely enforced. Does that statement sound correct… or almost correct?

image

92

No, it shouldn’t do that and I have not noticed anything that far off by ear but haven’t measured it that way. Possible that a mutation could cause it if you are +2 octaves and a mutated note was allowed to be +2 from the current octave. I would consider that a bug, but something like that might be happening. Would be curious if you had mutations on during this test.

Edit: if a mutation of +2 happened, and subsequently octave shift up, it’s not going to compress that note down when it moves everything up an octave. So that could definitely happen and I don’t know if I would consider that a bug. Have to think about it. Maybe a strict “enforce octave limit” could be a menu option.

93

I have “Mutation can transpose notes by octave” set to Off. I have MODE set to 0 I am occasionally sending a NEW trigger via a slow clock. “Lock CV In Behavior” is set to 0-2, but I’m not sure what that does. MUTATE is set to 67%.

I don’t think this is worthy of being a bug. I just discovered it while trying to understand Proteus. Initially I would occasionally hear what I I thought was a C7 but I had to set up the comparator to capture that event that might not occur for several minutes.

Is C7 in general playable and played on the sitar? What is the lowest note playable on the sitar?

Thanks.

94

After many minutes of playing, the lowest note played is now C#1. I suspect it will eventually hit C1 so the low range and high range would be acting similarly

95

From the base Sa you can go up about an octave and half, and down a full octave. That base Sa depends on the instrument but is usually close to C# (I keep mine between C# and D!). I will have to check which C# that is. I would guess C#4 but will have to check to be sure.

96

It may just be an academic point. Both C1 and C7 sound fine and natural to me, but then I am not sitar knowledgeable. It probably does not matter too much in this use case.

And I could set the RANGE to 1 to better fit the sitar. I am running that experiment now.

Edit: With RANGE set to 1, played notes ranges from C2 to G6 (perhaps approaching C7 asymptotically with time).

97

Yeah there’s no reason your synthetic sitar shouldnt do things the real one can’t as long as it sounds good!

2 Likes
98

I think I am settling in on what I feel to be my best effort at creating a synthetic sitar that sounds reasonably close to an acoustic sitar. The primary change in this version is that I reduced the Proteus RANGE to 1 to better limit the pitch range to the playable sitar range. I also added in a Chronoblob delay in the mixer to sort of double up the synthetic tabla emulation.

As a reminder, this uses the Vult Opulus Sitar2 preset along with two instances of the Vult Rescomb2 resonant comb filter to give 2 octaves of resonant sympathetic strings. Scale is Ahir Bhairav, produced by Seaside Modular’s Proteus module and Jawari along with plucked drone.

Slow articulated sitar with Hold on Tied Notes and String Bends Range1 -1.vcv (15.3 KB)

Slow articulated sitar with Hold on Tied Notes and String Bends Range1 -1-2
Slow articulated sitar with Hold on Tied Notes and String Bends Range1 -1-21902×807 237 KB

2 Likes
99

Here is a cross-post to my sitar patch using the beta build of Seaside Modular Tala tabla machine module. If you download the patch, be sure you have this Seaside Modular beta build or higher installed.

100

remember the corral electric sitar? I think the sympathetic strings don’t affect the sound. It’s mostly a piece of plastic under the strings that buzzes.

Joe sound played it a lot. It think he played it on a lot of sessions for others, actually.

1 Like