Quantization of onset detection

[mohriner]

I just started using the "Note onset detector: Note onsets" plugin.  This is a very powerful application, and kudos to its authors.  When I export the results as text files to examine in another programming environment, I notice that they're all quantized.  Although it looks like there are several significant digits to each time instant, they seem to move by increments of 25 milliseconds, which is perceptually significant.  Inverted to beats per minute, it's 1.5 bpm.  Is this a consequence of the fft window size, or a feature of the plugin?  Can I get higher resolution somehow?

Thanks,

Mitch

[cannam]

Mitch,

For this plugin at least, the quantization of the results is entirely dependent on the step size used when running the plugin: this is a basic limitation of this plugin, it isn't capable of returning results any more precisely than that.  Although the default step size is supposed to be 11.6ms, not 25 -- for example, it is 512 samples at 44100Hz -- so I'm a little puzzled about your results.

The step size can be changed from the Advanced section of the transform parameters in SV -- though note that the internal workings of the plugin are somewhat tuned for a particular step size, so if you change the step size you may need to adjust the sensitivity, and may find the results differ substantially.  (And be sure to also adjust the block size so that it remains double the step value.)


Chris

[grulbal]

I tried to change the value of the step size, but I have always the same results... I'm using the onset detection and the beat tracker, but even if I change the window size and the window increment in the advance parameter of the transform parameters in SV I have always the same results, with a step size of 11,6ms. So I don't know if I have to change another value somewhere in order to have a window step of 640ms (512 data points with my Fs=800Hz).
I hope you can explain why I have this results, and how can I change it.

Cheers.

Aymeric