User talk:FloraC/Proposed standard ear-training waveform

Version Stable 0

Hello. You can get the semisine (repeated parabola) by integrating a sawtooth function. Note the difference between the amplitude spectrum and the power spectrum. (-6 dB/oct is known as stop-band of 1st order LPF, but here we need a pure integral...)

sawtooth: 1, 1/2, 1/3, ... in amplitude but 1, 1/4, 1/9, ... in power, and described as -6 dB/oct.

semisine: 1, 1/4, 1/9, ... in amplitude but 1, 1/16, 1/81, ... in power, and described as -12 dB/oct.

Your additive synthesis is interesting. I too sometimes experiment with separating sawtooth waves into odd and even harmonics and shifting the frequencies of each.

440Hz sawtooth = 440Hz square + 880Hz sawtooth = 440Hz square + 880Hz square + 1760Hz sawtooth = ... --Dummy index (talk) 14:49, 12 April 2022 (UTC) --Dummy index (talk) 16:02, 12 April 2022 (UTC)

Glad my essay got some attention. You're right about the integral method. We discussed about this on Discord. Just that I have no idea how to do it in a consumer level synth. None of my synths ships with -6 dB/oct filters, either. FloraC (talk) 17:57, 12 April 2022 (UTC)