I have created a little classifier that listens for certain keywords. Sometimes it will not react to a keyword, but if I shout the word, then the classifier will be able to pick it up.
The classifier uses MFCC as features, which is basically a Short-time Fourier Transform with some extra stuff. This tells me that when I shout, the frequencies that the classifier are detecting are actually different from the non-shout.
So I'm wondering: does the volume of my shout change the frequencies in my speech, or am I somehow enunciating differently when I shout?