Now add to it a 100 Hz waveform that is -80 dB relative to the 200 Hz This sounds like the old "octave problem". > the slightly inharmonic physical or audible frequency content of interest). the mathematical fundamental frequency, completely unrelated to > some super low sub-harmonic related to the true least common multiple not necessarily the local maxima you want, or possible even To zero, you start to have an "octave problem". if the energy of the odd harmonics gets close a strong 2nd harmonic will not change that as long as there is someĮnergy at the odd harmonics. the ASDF or AMDF (or an auto-correlation derivedįrom it) will be minimum (or max for autocorrelation) at what looks like the > Most techniques similar to auto-correlation will lock onto the strongest I do not entirely understand why some weaker harmonic should be the one toĬare about in a pitch-detection problem. > might recognize as the musical "pitch" under some conditions). > harmonics is the one you care about (e.g. > not exact multiples of the missing fundamental and one of the weaker > However AMDF or ASDF does care if the "harmonics" are close to but > is derived from ASDF) does not care about a missing fundamental > the AMDF or ASDF (or auto-correlation which application to pitch detection > And one more time: FORGET ABOUT USING FFT for pitch detection !!!Įngineering is the art of making what you want from things you can get. Something funny with the link as it appeared. The requested URL /p� itch was not found on this server > trying to mislead general public asking questions here. > It seems that some people are unable to learn, or they are deliberatly PID loops were obsolete since about 1975, but > Didn't I tell everybody here that pitch detection problem is solved ?ĭon't hold your breath. Others here have mentioned using long zero-padded fft's as one alternativeįor certain types of pitch detection/measurement. The slightly inharmonic physical or audible frequency content of interest). Some super low sub-harmonic related to the true least common multipleĮ.g. Techniques similar to auto-correlation will lock onto the strongest Might recognize as the musical "pitch" under some conditions). Harmonics is the one you care about (e.g. Not exact multiples of the missing fundamental and one of the weaker However AMDF or ASDF does care if the "harmonics" are close to but >is derived from ASDF) does not care about a missing fundamental >the AMDF or ASDF (or auto-correlation which application to pitch detection as long as there is some short term periodic behavior (iĬall that property "quasi-periodic"), you can use AMDF or ASDF orĪutocorrelation with some interpolation to get the period length to whatever Is derived from ASDF) does not care about a missing fundamental (it *would*Ĭare if all of the odd harmonics were missing, which sometimes creates an The AMDF or ASDF (or auto-correlation which application to pitch detection > microphone), and for nearest semitone versus 0.1 cent pitch accuracy. > quick tap of the bottom key on a piano as filtered through a cheap telco > slightly inharmonic plink with a completely missing fundamental (a > consisting of mostly the fundamental frequency (flute) versus a short > to what accuracy? It's a very different problem for a sustained tone What instrument, for how long a sound, and pitch recognition/detection
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |