Re: [music-dsp] Time-domain noisiness estimator

> Am 22.02.2016 um 17:01 schrieb Dario Sanfilippo : > > I'll try studying autocorrelation more and see if I can implement a new > algorithm or combine it to the one I already have. Dario, you need to be careful with polyphonic material plus noise. The

Re: [music-dsp] Time-domain noisiness estimator

Thanks, Corey, Steffan and Risto. I'll try studying autocorrelation more and see if I can implement a new algorithm or combine it to the one I already have. ZCR alone doesn't seem to work unless for specific contexts. Would you agree on that? For example, a 10kHz sine and a 2kHz bandwidth noise

Re: [music-dsp] Time-domain noisiness estimator

On February 22, 2016 at 12:13:39 pm +01:00, Corey K wrote: > I don't have any links on the use of autocorrelation in this context, and I > don't even know if it would work. My basic thought, however, was that the > autocorrelation of white noise should be zero at all

Re: [music-dsp] Time-domain noisiness estimator

These properties are true, if you have only noise or only signal. In case of a mixture, also the described properties mix and this “torpedoes” that approach. So, an FFT with a subsequent processing like floor estimation (connect a line thru all floors between peaks) and peak estimation (connect

Re: [music-dsp] Time-domain noisiness estimator

I don't have any links on the use of autocorrelation in this context, and I don't even know if it would work. My basic thought, however, was that the autocorrelation of white noise should be zero at all time lags other than 0. Pitched signals, on the other hand, should have peaks at multiples of

Re: [music-dsp] Time-domain noisiness estimator

Thank you so much for the explanation, Ethan. Best, Dario On 21 February 2016 at 23:47, Ethan Duni wrote: > Not a purely time-domain approach, but you can consider comparing sparsity > in the time and Fourier domains. The idea is that periodic/tonal type > signals may be

Re: [music-dsp] Time-domain noisiness estimator

Not a purely time-domain approach, but you can consider comparing sparsity in the time and Fourier domains. The idea is that periodic/tonal type signals may be non-sparse in the time domain, but look sparse in the frequency domain (because all of the energy is on/around harmonics). Similarly,

Re: [music-dsp] Time-domain noisiness estimator

Hello. Corey: I'm honestly not so familiar with auto-correlation; I'm aware that it is implemented for pitch-detection but I didn't know about those other features; would you have a reference or link to a document I could check out? Evan: I get your point; in my case I was following more of a

Re: [music-dsp] Time-domain noisiness estimator

wouldn't using varying ZCR be defeated by frequency modulated or bell tones? One could also craft a very noisy signal with a perfectly periodic ZCR. James McCartney > On Feb 19, 2016, at 04:49, Dario Sanfilippo > wrote: > > Hello everybody. > > Following on a

Re: [music-dsp] Time-domain noisiness estimator

Noise is an elusive concept. One way of thinking about it is that the real signal can be decomposed into a sum of two theoretical signals, comprising desirable and undesirable information. What isn't signal is noise; and what isn't noise is signal -- so our definitions or models for these must

[music-dsp] Time-domain noisiness estimator

Hello everybody. Following on a discussion about cheap/time-domain spectral centroid estimators, I thought it could have been interesting to also discuss time-domain noisiness estimators. I think that a common approach is the FFT-based spectral flatness algorithm. In the time-domain,