The Sitar - synthesizing one.

[deastman]

I seem to recall seeing a standalone application which gave sitar-like overtones to a live audio input, mainly intended for guitar. This would have been around 2000 or so.

The problem with physical models of stringed instruments is the user interface...

[himalaya]

There was a user here who posted an excellent Sitar preset done in Cytrus some time ago.
I think he was subsequently employed by IL to make more presets, so perhaps you should ask them. I would imagine that preset being part of Cytrus library by now, if not then maybe its available for Cytrus users to download.

[loose snare]

(if you want to be fancy and a simple filter envelope isn't sufficient..)


i'm likely to attempt this at some point in the next year. sampled sitars are awful.

i've heard one sitar model around.. forget who but it was also awful, basically a karplus-strong.

my main obstacle is information.. (had a quick look at pdf, thx for link) i can't model something when i don't know what it is, and all i really know is bar frets/resonator strings inside frets.. knowing how they are connected to the body, what their pitches are, the distance between strings et c. is necessary to correctly couple them and produce the twang.

sharp pictures and diagrams would really help.
"

.... well I can tell you that a major factor in the sitar sound is the unique shape of the bridge for the main (playing, drone and chikari) strings. A typical stringed instrument has a thin bridge, let's say, 3mm and slightly rounded so that the string is effectively resting on a single point, but the sitar has a bridge which is flat and wide where the string rests - it is maybe 60mm (a rough guess) long (ie measured along the string). Also it is not really 'flat' but actually has a slight (parabolic I think) curve to it which slopes down away from the string as you go towards the fretboard end of the instrument.

This means that when a string is plucked its vibration will effectively lengthen and shortens the string as the vibration causes the string to come into contact with different parts the bridge along its subtly sloping surface. That's not to say the whole 60mm is used up (that would change the pitch too much) but the slope of the curve and the length allows this effect to occur in the same manner as you start fretting up the fretboard - which will of course pull the string down and change the angle of string meeting bridge (far more so in a sitar with its high action than a typical western acoustic guitar).

I hope that makes sense - if you want a sketch let me know! But you can see how wide the bridge is here, and the above explain why I hope

http://www.google.com/imgres?imgurl=htt ... 1&ct=image

As for the body - it's a hollowed out gourd. The neck is also hollow and is really part of the body too. Sometimes a second gourd is attached to the neck.

The smaller separate bridge for the sympathetic strings is also curved in a similar way to the main bridge - obviously these strings are quieter and are never played directly (well actually sometimes they are gently strummed or slowly plucked during the beginning parts of a raga as a sort of embellishment, but not generally - and they are always open - they have no frets).

They are tuned before a performance to the scale of the particular piece being played. Those swelling resonance strings on certain notes coupled with the 'smudged' pitch is what gives the sitar its characteristic 'meeeoooow' sound I guess........

Presumably the variance in pitch for any given note increases with amplitude .....(?) ... And presumably the pitch variance is more on the sharp side than the flat - although I'm not too sure TBH just thinking out loud....

All I can add though is that although I'm making it sound like a very fuzzy, inexact instrument, it's not really - in the hands of a good player (which also means a good tuner!) the pitches are very exact. The pitch is very much determined by the player and the pressure on and sideward bend of the strings (a technique also used to play entire melodies, runs as well as just passing notes) so that although it is fretted it is kind of played with the attention, technique and mindset similar to a fretless instrument - if that makes any sense!

That's a very brief description. Probably you know all this except maybe the bridge detail.

HTH!

[xoxos]

thank you..

what would help is if someone with a sitar and some appreciation of synthesis could create a schematic.. that page is the first insight i've had into the hollow neck...

there are so many strings and elements and i'm half a world away. i have no confidence that i have even identified a portion of the "acoustic circuit components."

the 'meow' swell, from waveguide experiments (which are informative to conduct)... i'll leave it at "i didn't hear the coupling algorithm in sekhtar."

i am confident that dispersion filtering and coupling with correct pitches would produce a better swell than i have heard (i think perry cook's meowed properly).