SirSickSik Contributions

[SirSickSik]

NEW MODULE

CONV

"audio2pitch"
a converter for getting the frequency/pitch of a mono audio source.

Can be used with the note-scaling and pitchshifter to make an auto-tuner.

[SirSickSik]

NEW MODULE

HARMONY

"allEqtempScales"

generic scaling module.
Input expects 1v/note, so with 19 notes in an octave, next octave starts at 20.
Output is scaled back to 12v/octave to be used with the oscillators with a "pitch" input.

"key" sets the key for the scale that's being generated, jumps with "interval" size through the available notes in the octave. This parameter is normalised to the "notes" parameter, so it always fits the total amount of notes per octave.

"Notes" sets the total amount of notes per octave (this means without scaling. western temperament = 12 notes) This can go up to even 128 notes per octave! Famous scales are 12, 19, 22, 31, 53.

"Use" sets the amount of notes which are active in the scale (western temperament = 7 notes). This parameter is normalised to the "notes" parameter, so it always fits the total amount of notes per octave.

"Interval" sets the interval by which the active notes are selected (western temperament = interval 7). This parameter is normalised to the "notes" parameter, so it always fits the total amount of notes per octave.

[Blindsmyth]

NEW MODULE

CONV

"audio2pitch"a converter for getting the frequency/pitch of a mono audio source.

Can be used with the note-scaling and pitchshifter to make an auto-tuner.

epicness! looking forward to using this for a guitar synth!

[Blindsmyth]

@SirSickSik
tried out the audio2pitch with my guitar.

it works for lower notes but is really erratic at higher pitches. Also latency is quite big.
I also had to tune the oscialtor up 5 semitones to have the right pitch.
So right now it is for me not really usable. Are you planning on tweaking this a bit more?

This is the setup:

[toneburst]

I've gotta say, I'm not surprised. Pitch-detection for guitars is really hard to do. Even the expensive guitar synths by big companies like Roland aren't that reliable, according to the reviews I've read.

a|x

[Blindsmyth]

I've gotta say, I'm not surprised. Pitch-detection for guitars is really hard to do. Even the expensive guitar synths by big companies like Roland aren't that reliable, according to the reviews I've read.

a|x

That's true but this roland things are also polyphonic, so that's also more a matter of picking up each string indivdually. And of course quick playing techniques like hammer ons, trills etc are not so easy to translate.

Monophonic pitch detecetion on the other hand is not too tricky I think. Each cheap digital multifx unit has like a monophonic guitar synth.
On the axoloti there is this yin object by @joren, wich did a pretty decent job on the guitar but wasn't yet finished and crashed my axoloti patches.

Creating new objects - Guidlines for operations on large buffers?

Hi, I am porting a monophonic pitch detection algorithm to an Axoloti object. The object does more ore less what it needs to do: it detects pitch in a buffer of 1024 samples. There are however a few things unclear to me: 1) What is the best way to execute operations on large buffers? I need 1024 samples, to I currently create an array and fill it with the incoming samples. After 1024/16=64 blocks of audio I do some pitch estimation calculation. Since audio starts to stutter i think this take…

So i think monophonic should be absolutely doable

[toneburst]

On the axoloti there is this yin object by @joren, wich did a pretty decent job on the guitar but wasn't yet finished and crashed my axoloti patches.

I was really intrigued by this, but @joren never replied to my PMs about it. I guess he abandoned it. Shame..

a|x

[SirSickSik]

you might add some extra lowpass filtering to remove more high frequencies and maybe an extra highpass filter at 20hz to remove sub-frequencies.
I used zero crossing technique for getting the pitch and to be able to get the right pitch, I needed a longer buffer, otherwise there was this contstant offset. Though there are some ways to improve this.
at the moment I'm coding a random oscillator which, hopefully, just stays around or under 4% cpu..

[Blindsmyth]

Ok thx the filtering helped a lot!
I think I will build a pickup just for the low E string on my guitar and use this to create a clean sine sub for my guitar synth/sampler:

Guitar Sampler/Synthesizer

It's been three months of intense patching with Axoloti and finally I've got my sequenced sampler patch ready to go on the road and hopefully on stage this summer! Thanks to all of you that helped me, the input of this community is really great! It is kind of a sampler that records little samples of my playing, triggers them from a step sequencer that I patched, plays the samples back in different octaves and/or reverse and puts it through filter and amp envelopes. In the end everything goe…

[SirSickSik]

NEW MODULE

OSC

"randOsc"
oscillator with randomisable waveshape.
Uses up to 64 steps and features both a seeded random as well as a triggered-randomise option for both waveshape-amplitude and waveshape-amplitude-update (where the shape changes to other values).
The whole array goes through a filter (yes, this becomes a smoothly faded array->morphed waveshape).
A mix-knob (SeRa) allows to set the mix between seeded and randomised waveform.
Multiple knobs and inputs are given to modulate the oscillator.

[JLS]

Hi

randOsc is great module thanks

little bug - smooth not working (see code and range value smooth is 0 to 0)

[SirSickSik]

hmm guess I accidentally changed the knob into a selector... fixed it

[mongrol]

Dear SSS,
While you're being so outrageously prolific, have you thought about enhancing some of the GPIO interfacing objects? It would be great to have better options or wrappers for exporting and importing Axo data out over i2c, UART etc.

[toneburst]

hmm guess I accidentally changed the knob into a selector... fixed it

The inlet/outlet tabs in the Editor are still quite glitchy in my experience. It's not hard to do things you didn't intend when working in these areas.

a|x

[SirSickSik]

yes, I think it happened when I editted the desciption boxes just before I saved it to the community.

[SirSickSik]

I haven't even touched these things yet haha
after 4 months, my axoloti is still a bare-bone circuitboard. I'm still thinking about the best layout and options for controls (eg. thinking about adding a multiplexer so I can choose between using a slider, knob, button or encoder for each control) that is directed towards the use of my modules. So first I'm gonna create all the modules I want and after that I'll focus on the internal/external control of those.

[SirSickSik]

new module coming up
4 voice poly oscillator
https://www.facebook.com/SirSickSik/videos/10157205287360626/

[SirSickSik]

NEW MODULE

DISP

"LFO_xy_scopeLowRes"
trying out whether I could a very low resolution xy scope using the "button-row displays"-thingies.
Takes in positive fractional signals and shows the xy position in a 16 by 16 steps matrix

[SirSickSik]

NEW MODULE:

OSC

"syncModOsc"
multi-modded oscillator featuring a big bunch of options for morphing your sound

audio modulations:
-linear and exponential FM
-PM
-AM
-softFM (phase modulation by pseudo-derivative of sync-input)

audio syncing:
-hardsync
-softsync
-sample&hold wave and add to original at sync
-reverse play direction
-shift phase by pseudo-derivative of sync input at sync moment

as far as it was usable, all parameters have external inputs for audio, mod-width and off/on sync-options.

while fully modded with several K-rate modules (including tempoSyncedLFO) and a sine-oscillator in the same patch, the cpu-use is 10%

[SirSickSik]

NEW MODULES:

PATT

'rndEuclidPatt"
pseudo-random pattern generator
-randomise pattern externally or manually
-"repeat" sets euclidian stepsize where "gates" are generated
-"width" sets the width of these gates, being added to the sequence.
-4 euclidian patterns are combined with the random pattern to generate a looping CV-pattern

"sRndEuclidPatt"
a simplified version of the above, featuring only 1 euclidian pattern and a random pattern.