FFT average analyser?

[_Morph_]

Hi!

I wonder if anybody knows about an OSX plugin that analyses not the momentary spectrum, but the RMS average throughout a song. Rather than showing the peaks, it would show the average RMS level in the whole spectrum range.

Thanks!

[Alexey Lukin]

You can switch Ozone's spectrum analyzer to Infinite averaging mode, and it will show you the RMS average.

[masteringhouse]

Someone once told me that if you stick your head in the oven and your feet in the freezer the average is that you are at a comfortable temperature.

I have no idea how that pertains to the OP but I was reminded for some reason.

[Alexey Lukin]

Long averaging is often useful for accurate analysis of stationary noise spectra. Short-term spectral estimates are noisy (hmm... variant) themselves, while longer averages nicely converge to the "true" spectrum (be it white noise or a pink one).

[masteringhouse]

Originally Posted by Alexey Lukin

Long averaging is often useful for accurate analysis of stationary noise spectra. Short-term spectral estimates are noisy (hmm... variant) themselves, while longer averages nicely converge to the "true" spectrum (be it white noise or a pink one).

So the purpose would be for measuring noise or hum?

[Axon]

This sort of thing can be useful with comparing non-stationary signals that were originally equal but put through different signal processing. I do it to evaluate vinyl vs CD comparisons, place constraints on peoples' cable comparison recordings, etc. Obviously some interpretation finesse is needed when considering nonlinear distortion.

My first answer would be AudioLeak. My second is to write it yourself, which is what I did.

To be honest, I think there are enough caveats to configuring such tools for non-stationary inputs, and interpreting their results, that anybody who uses such tools probably ought to be able to write them (and understand precisely what is going on under the hood).

[_Morph_]

Alexey, thank you very much!
This seems to be exactly spot on!

masteringhouse, we all have something to learn...

Axon, I was actually provided a self-coded program by a fellow engineer, but it's PC only. Hence the search for a Mac version. Unfortunately, I'm not much of a coder, but thanks for the heads up.

[masteringhouse]

Absolutely, never stop learning. I was curious as to the purpose for its use. Two very legitimate reasons as opposed to say using an average FFT curve to make aesthetic decisions regarding EQ for a track.

[Alexey Lukin]

Yes, it is quite useful for measurement of noises and for registration of long tones buried in noise.

For example, measurement of sound amplification systems in concert halls is often done with a noise stimulus, and averaging helps figuring out the accurate frequency response.

[Thor]

There is also the Dolby Media Meter, but that will only give you a Dolby LKFS value.

AudioLeak is less polished, but provides several measurement weightings to choose from, and is inexpensive.

Thor