Quilcom SIM-CHAU

[Spogg]

When I made my Quilcom SIM-GONG I had no idea that Martin would come up with a truly amazing feat of ASM coding which would create 256 resonant bandpass filters which barely touch the CPU.
To avoid too much repetition here’s the topic for background:

viewtopic.php?f=3&t=91211#p229505

This huge bank of filters was what I was trying to fake using resonant comb filters, since I considered that multiple bandpass filters wouldn’t be possible due to CPU load. I found the sound of Martin’s filter system to be considerably more convincing than combs so I decided to create a SIM which focussed on this particular type of gong. The original SIM-GONG produces tuned or bell-like gongs better than Martin’s system which excels at the more chaotic cymbal-like sounds. So I decided not to make a version 2 of the SIM-GONG which works in a very different way (and has been well received as-is anyway).

You might need to read the User Guide for this one!

Download:
https://www.dropbox.com/s/7gqwu4ob1tgp6 ... 0.zip?dl=0

And here’s my video:
https://youtu.be/E4B-MpQBR00

[martinvicanek]

Great sounds, Spogg! It is clever to split off the lowest (32) modes and treat them separately, as these give the sound an individual character (while the upper (224) modes are more statistical in nature). One thing I thought could be beneficial to add is an option to pan each mode randomly to create a mighty wide stereo texture.

[Spogg]

Great sounds, Spogg! It is clever to split off the lowest (32) modes and treat them separately, as these give the sound an individual character (while the upper (224) modes are more statistical in nature). One thing I thought could be beneficial to add is an option to pan each mode randomly to create a mighty wide stereo texture.

Thanks Martin!
A large gong played close would have a spacial component and not just a point source (mono). To achieve that I used a reverb set to minimum feedback so as to make use of the comb and all pass filters and have minimal tail length. Of course that creates more of a room ambience than panning each mode, but I thought the pseudo-stereo sound was acceptable. To pan each individual mode (and not each filter module) would require you to modify the filters of course.

[Tepeix]

Very cool and beautiful !)
Get me time to understand which note i have to press.
I like also playing with timbre.
Maybe an option to play different midi note could be cool, with maybe a random little de-tuning ?

Also i was seeing a video about spring reverb are used in a percussive way hitting the pedal.
And was thinking you must also be close to get this sound with this technique.

[Spogg]

Thank you Tepeix!

I chose middle C because I figured everybody would have a middle C!
You can use the B below for damping if you don’t have a modwheel.

I don’t think Random de-tuning would sound right for a Chau because it’s a fixed object so I think the modes would also be fixed, apart from the intermodulation produced by Martin’s amazing filters.

Regarding reverb I’m already making good progress on the Quilcom GONGVERB!

[Duckett]

Just stopping by to say- most payware (let alone freeware) developers would not have the steadfastness to try for as quality a modeled sim as possible for the less exotic PM-candidate-instruments, when making rompler instruments is so very much less work in 2022... though it might be a "grumpy old man" thing to say, seeing the steady proliferation of "do THIS-get RICH instead of work for a living!"-based "developers", YT channels, social media accounts, etc. in the realm of audio/music production does make me nostalgic for the mid-2000's... sure, there were garbage plugins and slimy people then too, but the SNR was higher in some ways IMO.

Also, sticking my nose into why gongs are so complex (why the standard PM "plate" model is insufficient) has me investigating Kolmogorov complexity https://www.youtube.com/watch?v=EFsLKtiz-qs; since there is plenty of discussion within most DSP communities basically about interpreting ideal (mathematical objects,unlimited processing speed/scale) to real (physically real objects, sharply defined boundaries/limits), and while one could spend a lifetime investigating Shannon information theory, Kolmogorov complexity is, to my mind, the most natural /fruitful extension of investigation

[Spogg]

Very interesting comments Mr Duckett!

There’s certainly a place for Romplers of course. If you want the sound of a real instrument you get it. But by using synthesis, once you have a model set up it can usually be adjusted away from the target sound and behaviour to create something impossible in the real world, or at least different.

I watched the video you linked but my mind wandered as it always does when I see mathematical formulae.
I got the impression it was saying the simplest solution is the best, like Occam’s Razor maybe. But I probably misunderstood.

[Duckett]

...by using synthesis, once you have a model set up it can usually be adjusted away from the target sound and behaviour to create something impossible in the real world, or at least different.

That's why I'm a fan of your efforts; while truly amazing things with scripting have been done (There's a fellow over on the REAPER forums named EvilDragon who's truly gifted at getting the most out of Kontakt), a sample is, at heart, a static entity, the starting place- and, if your samples are information-dense (keeping environment reverb tails, etc.), there's less room to take things into new territory that don't sound like flavors of noise.

Sorry the vid was perhaps a bit too mathy/sleep -inducing, but the short version of why I like it is: Kolmogorov complexity measurements have a shorter path to applicability in a wider range of real-world applications, and, for instance, could be used to pre-constrain the amount of code needed for "good-enough-for-human-ears"-accurate simulation of a given string/membrane/cylinder etc.

Seems like scanned synthesis might also benefit from further developments in the field.. but, I'm just rambling on.

For me, ideas tend to bubble under the surface for a while, before "the stuff under the hood" finishes working out why exactly it is actually interesting