(03-07-2025, 01:24 PM)Jens Noriʇzsɔɥ Wrote: Do you have a practical example of application of such octave bands. The Wikipedia article is a bit short on references and, right now, I have no idea what it is good for.
Anything should be possible in phyphox, however, we likely need some guidance in this regard (also: linear or log scale…?). Until then, you need to/could zoom into the graphs of “Spectrum” and “History” to select a window of interest.
There should be an acoustic department at a local university that has more information. It has been a long time since I worked with this, during my studies we had hardware filters for the different octaves, electronic/digital filters came later. I think the octave filters just take an interval around the octave centre [f_0/sqrt2, sqrt2 f_0] and sum all "intensities" in that interval. But I need to check that. I will try and write up an analysis in phyphox that will do that with the FFT data.
It was a long time ago I worked with this and I do not have my notes as I am on a sabbatical at another university.
(03-07-2025, 04:14 PM)Jens Noriʇzsɔɥ Wrote: Oh, something like this “Bandpass Amplitude”? Sebastian once did his magic…
Yes, kind of. Instead of taking the average in a specific band, I need to add the amplitudes and use the sum. I can not figure out how to do that in a proper way. Defining 10 different bands (octaves) seems quite easy. The final step will be to adjust the amplitudes to the A-filter used in sound analysis, by multiplying with weights.
Have you had any progress on this? I do not yet understand the math behind it. What amplitudes need to be added? Are the A-filters done in Fourier Transform space (by multiplication)?
(04-04-2025, 04:07 PM)Jens Noriʇzsɔɥ Wrote: Have you had any progress on this? I do not yet understand the math behind it. What amplitudes need to be added? Are the A-filters done in Fourier Transform space (by multiplication)?
I have managed to write an experiment that gives the SPL in different octaves. I need to tidy up the code before sharing.
But I need to check with a sound level meter that gives the different octave bands.
It worked when I did a test with a single tone.
(04-04-2025, 04:07 PM)Jens Noriʇzsɔɥ Wrote: Have you had any progress on this? I do not yet understand the math behind it. What amplitudes need to be added? Are the A-filters done in Fourier Transform space (by multiplication)?
I have managed to write an experiment that gives the SPL in different octaves. I need to tidy up the code before sharing.
But I need to check with a sound level meter that gives the different octave bands.
It worked when I did a test with a single tone.
I enclose the code, which I tested on Pink noise and single tones. The values of each octave band is just an integration of the levels in the specific octave. Filter values are the summation of all octave bands with different correction values, where A-filter mimics human hearing.
The sensitivity of the microphone and loudspeaker limited the tests. With pink noise, I got a flat curve for the highest octaves as expected, while the lower octaves showed lower values due to the response curve of the microphone, which drops off at lower frequencies.
Given time, I will try to find a way to calibrate the values and find suitable experiments.
