questions about the output about cava

[karlstav]

@livibetter let me just say that I really appreciate all the feedback. I had also forgot all i learned in signal processing (it was so hard!), but I re learned some of it when working on cava.

You are absolutely right, there seems to be some "noise" in the fifo processing. I will look through the code, it's over a year since I wrote that part so it might take a while.

Now over to the frequency bands. They are distributed in the same scale as with musical notes:

f=440^(n/12)

where n is the number of semitones up or down from a A4.

In cava the cutoff frequencies are calculated by:

fc[n] = 10000 * pow(10, -2.37 + ((((float)n + 1) / ((float)bands + 1)) *
                                             2.37));

So yes, a log scale, per my calculations the middle frequency would be 743Hz. So it looks about right with 1k a bit to the right of the center. I don't remember the details, but as i write in the readme it's supposed to go from 50-12,000Hz

[livibetter]

@karlstav Thanks for explanations. I hope you can locate and eliminate those three pesky peaks soon.

[karlstav]

turns out i was putting the samples from the temporary read buffer into the shared buffer the wrong way

While the shared buffer contains 2048 samples the temp buffer contains 128 samples. So each of those 128 samples would be inserted backwards. Should be fixde now.