Translation function of melda eq band frequency
-
- KVRAF
- 2623 posts since 20 Oct, 2014
Hi,
I would like to connect Renoise's note follower to a band of the melda eq, so 12 notes are one octave in band.
Already did that with Renoise EQ and a bunch of other EQs.
Can you please provide the proper translation function for a band of the eq?
Greetings.
I would like to connect Renoise's note follower to a band of the melda eq, so 12 notes are one octave in band.
Already did that with Renoise EQ and a bunch of other EQs.
Can you please provide the proper translation function for a band of the eq?
Greetings.
-
- KVRAF
- Topic Starter
- 2623 posts since 20 Oct, 2014
I mean the "simple" formula that makes it possible to use linear note values to logarithmic frequency of a band... Is it understandable what I mean?
-
- KVRAF
- 10375 posts since 2 Sep, 2003 from Surrey, UK
I don't quite understand.
Perhaps you could explain how you connect the Note Follower output to another EQ and what data passes between them?
Or do you just mean this formula? 8.1758125 * 2 ^ (NoteNumber/12), which gives:
MIDI note 60 is "Middle C" is 261.626 Hz, Also see:
https://newt.phys.unsw.edu.au/jw/notes.html
Perhaps you could explain how you connect the Note Follower output to another EQ and what data passes between them?
Or do you just mean this formula? 8.1758125 * 2 ^ (NoteNumber/12), which gives:
Code: Select all
0 8.1758125
12 16.351625
24 32.70325
36 65.4065
48 130.813
60 261.626
72 523.252
84 1046.504
96 2093.008
108 4186.016
120 8372.032
128 13289.7724
https://newt.phys.unsw.edu.au/jw/notes.html
-
- KVRist
- 206 posts since 17 Mar, 2013
If I understood correctly what the issue is,
He's asking for the formula of the frequency parameter in the melda eq. This formula controls how the vst parameter's value gets interpreted by the plugin, i.e. which value results to which frequency. Each plugin/plugin dev choses their own formula that they think is the most convenient. If you know the formula you can effectively bypass the interpretation by the plugin and get the specific frequency you wanted by supplying the corresponding parameter value.
So instead of
set parameter value -> get some frequency inside the plugin
it's
set desired frequency -> set the exact parameter value for that frequency
This is useful if you want to fully automate the process, let's say, make an eq band always follow your note's pitch.
This can be done with modulator's pitch tracking, but it's based on audio pitch detection, which isn't always ideal when precise pitch data is available, i.e. in a midi track. (I really expected this to be implemented with MCharacter) The melda modulators' pitch tracker also has a pitch offset from your current detection which could generally make this work, but then a base pitch must be supplied first, and when I checked it seemed that pitch basically gets reset through normal operation of the daw.
He can do it by making a huge lookup table for each value that the plugin shows to the daw as the string value of that parameter value.
Basically, the plugin shows "1khz" but the daw is seeing and working with, let's say 0.5 out of 0 to 1, and you increment from zero to 1, recording the value the plugin shows. This wont be particularly fun to make as you'll have to manually handle changes in the displayed form of the frequency, for example 1000hz turning into 1khz, 20-100hz having decimal places etc. The table will be huge, a couple of thousand entries for sure (could then later be increased with interpolation).
So basically the pitch tracker modulator should have a function of midi tracking only. (not what he asked for, just my suggestion)
Unrelatedly, MWobler's midi follow is not compatible with micro tuning
He's asking for the formula of the frequency parameter in the melda eq. This formula controls how the vst parameter's value gets interpreted by the plugin, i.e. which value results to which frequency. Each plugin/plugin dev choses their own formula that they think is the most convenient. If you know the formula you can effectively bypass the interpretation by the plugin and get the specific frequency you wanted by supplying the corresponding parameter value.
So instead of
set parameter value -> get some frequency inside the plugin
it's
set desired frequency -> set the exact parameter value for that frequency
This is useful if you want to fully automate the process, let's say, make an eq band always follow your note's pitch.
This can be done with modulator's pitch tracking, but it's based on audio pitch detection, which isn't always ideal when precise pitch data is available, i.e. in a midi track. (I really expected this to be implemented with MCharacter) The melda modulators' pitch tracker also has a pitch offset from your current detection which could generally make this work, but then a base pitch must be supplied first, and when I checked it seemed that pitch basically gets reset through normal operation of the daw.
He can do it by making a huge lookup table for each value that the plugin shows to the daw as the string value of that parameter value.
Basically, the plugin shows "1khz" but the daw is seeing and working with, let's say 0.5 out of 0 to 1, and you increment from zero to 1, recording the value the plugin shows. This wont be particularly fun to make as you'll have to manually handle changes in the displayed form of the frequency, for example 1000hz turning into 1khz, 20-100hz having decimal places etc. The table will be huge, a couple of thousand entries for sure (could then later be increased with interpolation).
So basically the pitch tracker modulator should have a function of midi tracking only. (not what he asked for, just my suggestion)
Unrelatedly, MWobler's midi follow is not compatible with micro tuning
-
- KVRAF
- Topic Starter
- 2623 posts since 20 Oct, 2014
yes, what pone writes, but without a lookup table, but a perfect formula/function.
I have a note follower device that gives each note a number from 0-127, so linear. Now I want to connect this number to a band of melda eq, so it can exactly fit the frequency of the note. But since the band parameter is not linear, a conversion / translation formula is needed, which I will write into Renoise's formula device.
So I would like to know this formula - Thanks!
I have a note follower device that gives each note a number from 0-127, so linear. Now I want to connect this number to a band of melda eq, so it can exactly fit the frequency of the note. But since the band parameter is not linear, a conversion / translation formula is needed, which I will write into Renoise's formula device.
So I would like to know this formula - Thanks!
-
- KVRAF
- Topic Starter
- 2623 posts since 20 Oct, 2014
Any further questions regarding this?
Just like here: http://forum.renoise.com/index.php/topi ... /?p=336564
Just like here: http://forum.renoise.com/index.php/topi ... /?p=336564
-
MeldaProduction MeldaProduction https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=176122
- KVRAF
- 14339 posts since 15 Mar, 2008 from Czech republic
Ok, so MIDI note to frequency is
f = C5 * mpow(2.0, (note - (12 * 5)) * (1.0 / 12.0))
where C5 = 1391.915051294
I assume the formula is the same to what DarkStar wrote:
f = 8.1758125 * 2 ^ (note/12)
I just like higher numbers for numerical stability, but it most likely doesn't matter at all
.
Then you need to convert it to the scale of the eq, hence into logarithmic scale in 0..1 (since automation parameters are in 0..1):
y = (log10(f) - log10(20)) / (log10(20000) - log10(20))
(don't accuse me of the log10...
bad choice at the time, but actually ln should probably work the same way)
f = C5 * mpow(2.0, (note - (12 * 5)) * (1.0 / 12.0))
where C5 = 1391.915051294
I assume the formula is the same to what DarkStar wrote:
f = 8.1758125 * 2 ^ (note/12)
I just like higher numbers for numerical stability, but it most likely doesn't matter at all
Then you need to convert it to the scale of the eq, hence into logarithmic scale in 0..1 (since automation parameters are in 0..1):
y = (log10(f) - log10(20)) / (log10(20000) - log10(20))
(don't accuse me of the log10...
-
- KVRAF
- Topic Starter
- 2623 posts since 20 Oct, 2014
Vojtech, thanks
