This
interpolate (0,0) (1,1) (2,5) - Wolfram|Alpha
shows an example which might be useful.
You will probably get good results using 4 or 5 of your points. Just experiment and see how well it works in practice.
Hi,
Since I'm a musician and dont know that much about mathematic things I need your help!
I'm looking to find a way to get a formula (third degree equation?) based on the following data.
I need a formula because I want to use it in a max/msp environment because the input X needs to change to a Y but its not linear.
The VolumeSlider can only get 0.0 - 1.0 in but needs to move the slider to the corresponding dB place (-70 till +6dB).
(image of Ableton VolumeSlider)
I have some 'snapshots' of corresponding data and plotted them in a graphic thing. (see screenshots)
(table 1)
VolumeSlider (0.0 - 1.0) VolumeMeter (-70 - +6 dB) 0.000 -70 0.035 -60 0.104 -48 0.187 -36 0.238 -30 0.302 -24 0.400 -18 0.550 -12 0.700 -6 0.850 0 1.000 6
But because of the conversion IN the VolumeSlider I also need an OUT conversion. Which is a flip of the above data. (I subtracted -6 dB of all the VolumeMeter because of some sort of offset in maxmsp function window. Couldn't specify a range between minus and some plus above thing. Random).
(table 2)
VolumeMeter (70 - 0 dB) VolumeSlider (0.0 - 1.0) 0 1.000 6 0.850 12 0.700 18 0.550 24 0.400 30 0.302 36 0.238 42 0.187 54 0.104 66 0.035 76 0.000
But I want some sort of formula based on these two tables so I can use it fluently. I need some sort of "best curve fitting" thing of "trendline" based on these two tables.
Hopefully you guys can help me out!
This
interpolate (0,0) (1,1) (2,5) - Wolfram|Alpha
shows an example which might be useful.
You will probably get good results using 4 or 5 of your points. Just experiment and see how well it works in practice.
thanks for giving me a great website for using this stuff. nonetheless it gives me an idea of a good representation of a formula.
here are some results.
but how can i use this thing. it gives me a 'bumb' between the 0.8 and 1.0?
(click here for original data from wolfram)
Glad to see my post was sufficient to get you started.
The point about using fewer points is that will get a simpler polynomial and the results will I suspect be indistinguishable to our ears.
Edit: I missed your comment the 'bump'.
You could add a suitable point at 0.9.
Just play around with it. Add points, leave points out. Find out how few points you can have and still get a good
result.