I'm interesting in knowing more about different kinds of data representations that are used for speech data. I know of FFT, but not any others. More precisely, I have the 3 following questions.

  1. What data representations for speech data?
  2. What is considered the most 'raw' representation?
  3. How has the use of different representations evolved over time?

Simply answering 1 or 2 of the questions would also be great. Thanks.


1 Answer 1


I have seen people use MFCC coefficients for speech related analysis.It is a done using filter banks and logarithmic scales for representing speech.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.