On Sat, 15 Feb 2020 00:59:29 +0100 Robin Gareus <robin@xxxxxxxxxx> wrote: >On 2/15/20 12:19 AM, Will Godfrey wrote: > >> This is what I use for near constant power panning. Any good? >> float t = (float)(GlobalPar.PPanning - 1) / 126.0f; >> pangainL = cosf(t * Pi/2); >> pangainR = cosf((1.0f - t) * Pi/2); > >Why cosine? You might want to use sinus curve as deflection >`t = sin(t * M_PI/2)`, but since signal-power is proportional to the >square of the signal [1], for equal power pan you want square root: > > gainL = sqrtf (t) / DB3 > gainR = sqrtf (1.f - t) / DB3 > >where #define DB3 1.4125375 // 10^(3/20) > >> Where GlobalPar.PPanning is in the range 0 - 127 >> It actually gives a 3dB hump in the middle > >You should attenuate the signal by 3dB to prevent clipping, or maybe >only 2.5dB or perhaps 4.5dB, 4.2dB is also not unheard of: > >http://admiralbumblebee.com/music/2019/12/08/Daw-V-Daw-Pan-Curves.html > >Cheers! >robin > > >[1] Power is P = U * I >plug in Ohm's Law: R = U / I >P = U^2 / R >_______________________________________________ Most interesting. Thanks for the link. A fascinating read! I don't actually know where the law I use originally came from, but I'll be reconsidering. -- Will J Godfrey http://www.musically.me.uk Say you have a poem and I have a tune. Exchange them and we can both have a poem, a tune, and a song. _______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx https://lists.linuxaudio.org/listinfo/linux-audio-user
