Tesla Coil Music

[Transgarp]

For a month since I have been trying to soften the musical sound of my Tesla Coil
I have made progress since then, but not enough and the vast majority of my close friends find the sound abominable
I understand that the natural sound of lightning is not soft and much aggressive
So do you know any modules or way to configure existing ones to soften the sound?

For the comparison I use the same musical piece, Genesis Supper's Ready
In order of sound quality, from worst to best depending on how I feel:

EVS Universal MIDI 1.5 Interrupter

EVS Pro MIDI 2.0 Controller (midi_duty_max 250, midi_max_pw 150)

TMAX MidiStick (max on time per note : 100us,  min on time : 2us,  max duty cycle : 10%,  note holdoff : 60 us)

[Max]

Three ideas

1. Don‘t let the tesla coil play alone. Give it a few key voices and let all the other instruments play on normal speakers. In that case, even the most basic interrupter can serve for a nice show:

Many Synthesizers come with some crappy default sounds. However, there are very high quality „soundfonts“ (thats how a set of audio samples for playing MIDI files are called) available for free. Check out my PC setup for details: https://github.com/MMMZZZZ/Syntherrupter#pc-midi-setup

2. Read the manuals of your high-end interrupter(s)*. TMax‘ MIDIStick is - similar to my Syntherrupter - capable of rather complex modulations - most notably something called ADSR or in general envelope. Normal interrupters play notes at a constant volume; the sound of the note does not change while it‘s on. Now think of a piano: during the attack (when hitting a key), the volume rises over a finite amount of time up to a peak, and then decreases slowly with time. This volume shape is what‘s called envelope. In a simplified form one can approximate most instruments by dividing that envelope in 4 sections: Attack, Decay, Sustain and Release.
To give an example of what ADSR can do, listen to the following audio (really noticeable from 0m37s onwards). The short lower notes are not sharp on/off. You can hear them rising and falling quickly in volume giving them a softer sound (like a short note on a cello). Additionally you got a slight reverb. That‘s not the room or the metal, it‘s the tesla coil itself, it‘s part of the envelope.

* I guess TMax has written some decent documentation on how to do this; I don‘t know. I can only point you to my own documentation for my Syntherrupter but that likely doesn‘t help much.

3. Decent MIDI files. For years one had to modify MIDI files and strip 90% of the data from it to get single or dual channel MIDI files that‘d play on classic MIDI interrupters. However, modern interrupters like your MIDIStick or my Syntherrupter can process a lot of the other information, too. Nuances (increases/decreases in volume during the piece) f.ex. Therefore forget about the 2kB MIDI files with 1 Track and nothing else and play around with richer files. Good MIDIs tend to be in the 30kiB-80kiB range (obviously depends on the length). With those files you usually got meaningful other voices, too (see point 1).

(4. Pulse skipping / long ontimes. May not apply to your coil and only makes sense for lower notes. The thing is, long ontimes, like >>300us, tend to be less „sharp“ in their sound.)


Kind regards,
Max

[Intra]

do you know any modules or way to configure existing ones to soften the sound?

You should set ontime from microseconds to milliseconds with respect of current limiting and cooling of each part making heat. Try freewheeling and waterblocks\watercooling for primary. Midi could be the same. It's not important. It's about sound quantization noise on small ontimes.

in-driver\in-controller ADSR could help but by other way.