guitar to synth without midi pickup

[stratum]

Does anyone know how this might possibly be working?

EHX KEY9 Electric Piano Machine demo:

https://www.youtube.com/watch?v=av4X84lChEE

[Miles1981]

You also have JamOrigin selling a guitar to MIDI app for smartphones or workstations. So not really amazing to see that you can put it in a pedal with a "simple" synth after

[stratum]

You also have JamOrigin selling a guitar to MIDI app for smartphones or workstations. So not really amazing to see that you can put it in a pedal with a "simple" synth after

I guess you are right, given that we are living in a time where apps like melodyne are also found. I felt myself a bit outdated:-)

[stratum]

So lets rephrase the question. How do we reliably detect pitches in a polyphonic sound? Perhaps with this? https://en.wikipedia.org/wiki/Cepstrum

This part was interesting:

The independent variable of a cepstral graph is called the quefrency. The quefrency is a measure of time, though not in the sense of a signal in the time domain. For example, if the sampling rate of an audio signal is 44100 Hz and there is a large peak in the cepstrum whose quefrency is 100 samples, the peak indicates the presence of a pitch that is 44100/100 = 441 Hz. This peak occurs in the cepstrum because the harmonics in the spectrum are periodic, and the period corresponds to the pitch. Note that a pure sine wave should not be used to test the cepstrum for its pitch determination from quefrency as a pure sine wave does not contain any harmonics. Rather, a test signal containing harmonics should be used (such as the sum of at least two sines where the second sine is some harmonic (multiple) of the first sine).

So existence of multiple harmonics looks like a reliable hint for existence of a given pitch, only if we also could distinguish between harmonics due to the top E string from those of the bottom E, simultaneously playing.

[Miles1981]

There is also the fact that on a string, you have a decreasing amount of frequencies, so if there is a bump, it's more than likely that you ahve a new note.
cepstrum is indeed a tool often used in audio signal processing and taught usually in the firts three years.

[stratum]

There is also the fact that on a string, you have a decreasing amount of frequencies, so if there is a bump, it's more than likely that you ahve a new note.

This looks like a solution at first, but the situation is complicated by the fact that relative strengths of frequencies may not really be a reliable hint given the various EQ's one might impose upon a given signal. The case might be simpler for the EHX pedal in the video, as it is directly connected to a guitar, so in spite of different guitars and pickups, the algorithm would still be in a known "ballpark", so the chances that somebody might insert an EQ with a peak is rather low.

cepstrum is indeed a tool often used in audio signal processing and taught usually in the firts three years.

I guess such courses are only found in developed countries where there is strong demand for specialization and an existing market for their applications.

[Miles1981]

I guess such courses are only found in developed countries where there is strong demand for specialization and an existing market for their applications.

Probably, my school was quite intense.
Anyway, cepstrum is indeed easy to do, and you can "transform" it in a time-frequency transform, just like FFT gives SFT as well. You still get latency out of such a tool or lack of precision, depending on the compromise you choose.

[stratum]

Probably, my school was quite intense.

It's known since Adams Smith's time that smaller markets wouldn't afford specialization and without specialization there would be less division of labour and less economic efficiency. I don't know if that has anything to do with intensity of education, though, since in the end they would be educating an human being there would be a limit to that (the japanese might disagree )

Anyway, cepstrum is indeed easy to do, and you can "transform" it in a time-frequency transform, just like FFT gives SFT as well. You still get latency out of such a tool or lack of precision, depending on the compromise you choose.

Perhaps that was the reason for the usage of midi pickup, who knows. Or perhaps such algorithms have became popular only because cpu's that would execute them became cheaper. Back in the day when I had seen the line6 patent about oversampling I couldn't understand what was worthy of patenting in it - it just seemed like the idea became feasible only because it was the right time for it because of hardware availability. To be fair, I couldn't implement it to the same level of audio quality so there was something "nonobvious" about it, at least for me.