cursor2audio_freq_v2.m

function freq = cursor2audio_freq_v2(cursor, fb_cal)
%Cursor is the neural control signal.  From old work, this would be: 
%sumE2-sumE1
%
%Auditory feedback calculation
%freq = a*exp(b*(cursor_trunc-cursor_min))
% freq_min = a =  a*exp(0)
% freq_max = a*exp(b*(cursor_max-cursor_min))
% b = log(freq_max/a)/(cursor_max-cursor_min)
%param: cursor_min, cursor_max
%%
%If params.fb.target_low_freq == 1, then we negate cursor and cursor_min,
%cursor_max
%
%This is because cursor is E2-E1.  Our target is gonna be positive.  

%%
% %Debug:
% cal.fb.target_low_freq = 1; 
% cal.fb.freq_max = 19000; 
% cal.fb.freq_min = 6000; 
% 
% cal.fb.cursor_min    = -1;
% cal.fb.cursor_max    = 1;
% cal.fb.cursor_range  = 2; 
% 
% cal.fb.a = cal.freq_min;
% cal.fb.b = (log(cal.freq_max) - log(cal.a))/cal.cursor_range; 
% 
% cursor = -4:0.1:2; 

%%
%Handle target -> freq:
if fb_cal.settings.target_low_freq == 1
    %This means cursor up makes auditory freq go down:
    cursor      = -cursor; 
    cursor_min  = -fb_cal.cursor_max;
    cursor_max  = -fb_cal.cursor_min;
else
    %This means cursor up makes auditory freq go up:
    cursor_min  = fb_cal.cursor_min;
    cursor_max  = fb_cal.cursor_max;
end

%%
cursor_trunc    = max(cursor, cursor_min); 
cursor_trunc    = min(cursor_trunc, cursor_max); 
freq = fb_cal.a*exp(fb_cal.b*(cursor_trunc-cursor_min));
freq = double(freq); 
% h = figure;
% plot(cursor, freq);