I am trying to implement a method for measuring the response of a loudspeaker in a room using a log sine sweep stimulus.

To check my method: 1. I create the stimulus using:

enter image description here

  1. Make recording length long enough so that the measurement falls towards zero (or below the noise floor)

  2. Window the signal, and zero pad the stimulus to equal the same length as the measurement signal.

  3. Take FFT of both stimulus and measurement

  4. Multiply the measurement and stimulus by the inverse filter:

enter image description here

  1. Shift the FFT results to order them in the ‘natural way’

  2. Divide the FFT of the measurement signal by the reference signal

  3. Perform inverse FFT of the division to get back to the time domain. This will be the impulse response.

I can plot this and obtain something resembling an impulse response, however there are 2 ‘mirrored’ response signals here. Resulting in what looks like 4 impulse responses. I would expect to have a symmetrical impulse response, however this has thrown me. At all points in my method I use the full symmetrical signal and do not discard any data.

Please can someone confirm if I have missed a step or misunderstood along the way?

  • $\begingroup$ Can you please edit the question for clarity? If there is a good reference (and preferably open) to the method, then add that and simply mention the exact point where you seem to be having problems. Also, it would be good to post representative input and output signals (or at least their plots). I suspect that your chirp may be folding. $\endgroup$ – A_A Apr 25 '19 at 7:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.