Accordion sound?

[campidelli]

Hi everyone, I am trying to design an accordion sound using the lib but I am not there yet.

You can see the demo and sketch here https://hackaday.io/project/197399-arduino-esp32-standalone-accordion/log/232583-now-it-is-polyphonic

I really would like to have something like this https://www.youtube.com/watch?v=3mif1_ttTgk but I don't if I can achieve that with the Mozzi library. Do you?

Any tips?

Cheers!

[Mozzi-users]

Hi,

it looks like you've made some progress!  It also looks like there are a lot of examples and discussion online of accordion sound synthesis.  What elements in the video are you wondering about doing with Mozzi?  If you can break down the synthesis process into basic components as they do in that video and elsewhere, you can probably put them together with Mozzi, cpu permitting... it looks like a couple of additive waveforms, envelopes, some modulation (I'm probably missing plenty here)...

[campidelli]

Hi!! Thanks a lot for your reply, I thought I was the only one using this forum haha.

I would be very happy to achieve what he did in the video from 1:00 to 1:23, that sound is good enough for me. In these 23 seconds he:

1. Selects a waveform

2. "Throws the Spectrum down" of that waveform. Having a look at the Massive Synth manual, they say this about the Intensity and Spectrum:

3. He then applies a comb filter with damping and feedback set to the maximum. Again, from their manual, they say:

So. theoretically, I could:

1. Create an oscillator with any waveform, let's say the Oscil<SAW8192_NUM_CELLS, MOZZI_AUDIO_RATE> oscillator(SAW8192_DATA);

2. Apply some sort of filter to diminish the waveform spectrum. I have tried with the ResonantFilter<LOWPASS> but I couldn't achieve a good result;

3. And finally create a comb filter. I tried to use the AudioDelayFeedback to do it but I couldn't hear any difference, I simply don't know how to use it.

Anyway, this looks simple, but since I know nothing about Synths and the names are so different from the video and the library, I am kind of lost.

I appreciate your help!

If you want to check it out, this is the Github repo https://github.com/campidelli/arduino-accordion/tree/using-mozzi

Cheers!

--
You received this message because you are subscribed to the Google Groups "Mozzi-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mozzi-users...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/mozzi-users/2334160e-1319-4f48-b97d-ddcd33f4949fn%40googlegroups.com.

[Mozzi-users]

Hey again,

>tried with the ResonantFilter<LOWPASS> but I couldn't achieve a good result

what did it sound like?  No effect? (Maybe the initial waveform doesn't have enough high frequencies to notice it?)  Breaking up? (cpu load?)

>AudioDelayFeedback to do it but I couldn't hear any difference

Maybe if it's not modulating (changing the delay time) it might be hard to notice unless you compare the sound directly... could try more feedback?

Also in the video he says he adds some chorus and reverb - your detuned oscillators would add some chorus already, though you could add modulation and longer delay time to the delay for additional chorus effect.

Or you could try Mozzi's AudioDelayFeedback <256, ALLPASS>  (https://sensorium.github.io/Mozzi/examples/#09.Delays) for a more pronounced comb filter with a short reverby character...

>I simply don't know how to use it

Which bits are unclear?  Do the examples need more explanation?

I think it's do-able, so ask more if you need to...

Tim

[campidelli]

Hi Tim, thanks a lot for replying to my email.

I ended up using another library that fits better my purpose of creating a standalone instrument. If you want to know more about it, please check it at https://hackaday.io/project/197399-arduino-esp32-standalone-accordion/log/232631-found-the-synth-library-amy

Cheers,

Bruno

To view this discussion on the web, visit https://groups.google.com/d/msgid/mozzi-users/735bd222-9711-4a3a-a478-07fa4509af16n%40googlegroups.com.

[tomco...@live.fr]

Snif :'( I really liked the way you implemented voices in structures for Mozzi thought!

AMY seems also to be a very interesting project! Looks like the API is a bit higher level than Mozzi where all elements of synthesis are broken down into small pieces (oscillators, filters etc) and not as full voices like in AMY. Probably easier to work with if you do not want to implement the full synthesis algorithm yourself. (Note that Mozzi is also made for standalone instruments :P )

Good luck for that very nice project!