1) Write a PPM function which allows for variable attack and release times

  - The function should take an audio signal, attack time, release time, and Fs as inputs and return the filtered signal (i.e. function vppm = ppm( x, attackTime, releaseTime, fs ))

  - Remember the PPM extracts the envelope of a signal on a sample-per-sample basis as opposed to block-wise analysis like in the peak envelope function

  - plot your vppm over the original audio signal (what happens when you change the attack/release times?)

  - how might you use this as a feature?

2) Weighted RMS according to Broadcasting standards

  - Generate a couple seconds of white noise

  - Write an RMS function which weights the signal according to broadcasting standards. As a reference use http://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1770-3-201208-I!!PDF-E.pdf

      - In order to do this you must:

            write a 2nd Order IIR filter

            Filter the signal twice according to the filter coefficients in the reference (the coefficients are for 48kHz)

            compute block-wise RMS over the new filtered signal

there are a few ways to do this but you might have a function for computing the block-wise RMS and another function for the 2nd order IIR and a 3rd function which calls both taking the original signal as input and returning the weighted RMS values