Haha the easter bunny has delivered.

[ustk]

@d-healey Yeah I was just playing a wee bit :)
ATM I don't really see many use cases for me anyway if it can't be real-time, but we'll see when it's ready...

EDIT: except if it can extract the data, like the harmonics!

[d.healey]

@ustk Even if it did work realtime it would only be suitable for GPL projects.

[ustk]

@d-healey Yeah, unfortunately. But I am very interested in a way to extract the harmonic dataset from a non-real-time sound

[d.healey]

@ustk

But I am very interested in a way to extract the harmonic dataset from a non-real-time sound

You might like this program (not open source) https://www.klingbeil.com/spear/

[ustk]

@d-healey Please, Dave, stop finding more tools, I can't follow anymore
Thank you I'm checking this! But in the end, I need a Hise-embedded solution for analysing input recordings (non real-time)

[d.healey]

@ustk said in Haha the easter bunny has delivered.:

But in the end, I need a Hise-embedded solution for analysing input recordings (non real-time)

For use in HISE or in your compiled projects?

[ustk]

@d-healey Compiled project.
Since the sampler's sample editor can show the partials already, the analysis might not be far from being ported as an API, or even is already possible and I don't know...

[ustk]

@ustk In the end I just want a list of partials, let say the first 8th ones, and their level... Like for additive synthesis but it's not what I plan to do.

[d.healey]

@ustk

Compiled project

Since Loris is GPL it can only be embedded in a GPL project so you might be out of luck.

[ustk]

@d-healey oh I thought it was just the real-time capability that was GPL only, not everything...

[d.healey]

@ustk The entire library, as far as I know there is no specific real-time stuff and based on what Christoph was saying earlier it wouldn't be fast enough for real-time processing.

Now you have another great reason to use the GPL for your software :)

[ustk]

@d-healey

[ustk]

@d-healey Since I just need a partial list it might not be the simpler tool anyway.
And if the sampler can draw them it should be possible. I reckon there's a 2DSpectrum thingy in the FFT already so getting a data object should be pretty simple to add for "who is not me" :)

EDIT: In fact, I'm sure it's already there because we can get all bins from an FFT, I just have to somehow filter out the one I need...

[d.healey]

@ustk I use audacity and then export the list to csv, open in a spreadsheet, and filter the data. Might be too slow though if you have a lot of samples

[ustk]

@d-healey Yeah but remember it has to be doable in a compiled project in the end ;)

[Lindon]

@ustk cant you use Loris in HISE to generate the SDIF file and then some other (not open source) process to read the partials from the SDIF?

[Lindon]

@Lindon what format is the SDIF file?

[ustk]

@Lindon Don't you know what an SDIF file is, Lindon? Because... well... I don't... :)

WikiWiki:
"SDIF, "Sound Description Interchange Format" is a standard for the well-defined and extensible interchange of a variety of sound descriptions."

[Lindon]

@ustk its the partials analysis produced by loris analyse - essentially the sines and their envelopes that (when combined) re-generate your original audio...

[ustk]

@Lindon A nice idea indeed, if it's authorised by the license