Ploki

how do i go about doing that? I want to some arbitrary frequencies to what we perceive as natural.

remap-linear-frequency-plot-to-percepted-pitch

do you mean remap frequency to the western equal tempered scale? use [ftom] for that

otherwise maybe a look at emmanual joudans zsa dscriptors: 

for more complex modelling of pitch perception, which is a combination of frequency, harmonics, overall intensity, noisiness etc of a sound event

do you mean remap frequency to the western equal tempered scale? use [ftom] for that
otherwise maybe a look at emmanual joudans zsa dscriptors: http://www.e--j.com/
for more complex modelling of pitch perception, which is a combination of frequency, harmonics, overall intensity, noisiness etc of a sound event


It's related to my last post where I asked about how I divide frequencies to get fractal something brownian tree something.

https://cycling74.com/forums/help-with-fractal-kind-numbers/

now I want them to be on a "pitch scale", i.e. if I generate a middle frequency between 50 and 100hz, i get 75Hz. (and then generate two new, between 50 and 75 and 75 and 100).

Meaning a middle between 50 and 100 (an octave) would be a tritone. But a middle between 75 and 100 is something else. Doesn't need to be conformed to 12TET at all, just that "middle splits" are on a more logarithmic level. So, I need something to remap the results? I guess.

I used "m to f" and "m to f" sequentially to get a curve of how max interprets pitches but it doesn't look like there's a constant function.

It's related to my last post where I asked about how I divide frequencies to get fractal something brownian tree something.
https://cycling74.com/forums/help-with-fractal-kind-numbers/

I used "m to f" and "m to f" sequentially to get a curve of how max interprets pitches but it doesn't look like there's a constant function.


just convert frequency to midi before inputting to the algorithm you're using, then convert the results back afterwards (mtof).

it doesn't matter what the tuning system that you use is, equal temperament is fine-- its just important to convert to a log scale, but it doesn't matter what the base of the scale is, so you get the geometric mean, rather than the arithmetic mean

just convert frequency to midi before inputting to the algorithm you're using, then convert the results back afterwards (mtof).
it doesn't matter what the tuning system that you use is, equal temperament is fine-- its just important to convert to a log scale, but it doesn't matter what the base of the scale is, so you get the geometric mean, rather than the arithmetic mean


that's all fine and well, if [m to f] and [m to f] would support "float" notes. (3.5 midi notes), instead it jumps and comforms to 12TET.

edit: i was mistaken, looks like it works with floats and I really can use [ftom] and [mtof] for my scaling. Now to just think about what I wanted to do in the first place

>Now to just think about what I wanted to do in the first place

I.... I'd advise you to spend a good bit of time doing this. Spending time doing specification is by FAR the biggest problem solver in a system; "Why do I want X to happen, and how?"

"...looks like it works with floats and I really can use [ftom] and [mtof] for my scaling..."

yes I think the name is misleading, given in MIDI there is no such thing as a fraction of a MIDI note number-- they should be called ftop (frequency to *pitch*) and ptof, to indicate fractional 12-tet pitches

what you described in your second post still does not make a difference: you convert hertz to linear using [ftom] or an expression which does the same what ftom does.

in case you want to display hertz later, you would first use ftom, then do your calculations to find the new members, then use mtof again.

this is the only way how to properly work on frequencies in hertz on the linear layer - convert hertz to linear and back.

Remap linear frequency plot to "percepted pitch"