0
$\begingroup$

I am working on trying to apply a low and high pass filter to an audio file that contains a set of exhalations over a microphone. The inhalations have been cut out of the file, and the exhalations are stitched together in the file. I am attempting to replicate this paper where they have set 10 and 150 Hz cutoffs on their microphone data. I have attached the relevant portion of the paper at the end of the post.

I am currently following the code linked at this DSP stack exchange post, but am unclear on what an appropriate measure for the 'order' parameter would be when using the Butterworth bandpass filter. Currently, with an order of 5, it seems to filter out all of the audio from the .wav file... what is the effect of the order on the input data?

The end goal is to perform a spectral analysis on the filtered data. The unfiltered file can be found here.

Thank you in advance for any help. enter image description here

Improve this question
$\endgroup$
0

1 Answer 1

Reset to default
1
$\begingroup$

From the paper

we used a microphone that has a low-pass filtered with a cutoff frequency at 10 Hz and a high-pass filtered at 150 Hz and is amplified by 20 dB

This makes no sense whatsoever. If you lowpass filter audio at 10 Hz, you have nothing left. I'm guessing it's a typo. Probably it's supposed to be 10 kHz.

A bandpass from 10Hz to 150 Hz also makes no sense since the analysis frequencies are much higher. Example:

The nasal sounds were calculated for each of the nasal cavities and a 2000- to 4000-Hz frequency interval was used for evaluation.

Improve this answer
$\endgroup$
3
  • $\begingroup$ Yup. Time to contact the authors for a clarification. $\endgroup$
    – TimWescott
    Aug 4, 2021 at 15:01
  • $\begingroup$ Thanks for the response. We reached out to the authors but are not hearing back. Do you have any recommendations on how to proceed? It seems like there are inherent flaws in the methodology but we would still like to try and reproduce portions of the paper. $\endgroup$
    – Kabir Jolly
    Aug 10, 2021 at 16:37
  • $\begingroup$ I'd go with with 10 kHz. All of the mentioned frequency ranges should work well with this. Frankly, the article sounds very heuristic and not particularly scientific. Describing an A/D converter as a " a very fast digital voltmeter" seems more appropriate for a high school class rather than an academic environment. $\endgroup$
    – Hilmar
    Aug 10, 2021 at 18:38

Your Answer

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

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