how does harmonic distortion add in generator work. does it add a sine wave with level and phase or jitter sine wave ?

Hi
I notice with the generator can hear phase shifts very good and this give some strange feeling in ears at 27 and - 27 degree more as at 0. for example in ABX a blind test i detect that in 7 out of 8 try

Can you confirm that the 2. wave add a clean sine tone without distortion at the double frequency as the base frequency ?
Because Toole say phase is not much important hearable, so can test with REW generator if people can hear or not.

I do test wth second and -18 db when make louder can hear even more

It adds the harmonics at the selected levels and with the selected relative phase. Your test is not a valid way to examine the audibility of phase response. Experimenting with all-pass filters might be more meaningful.

John Mulcahy said:

It adds the harmonics at the selected levels and with the selected relative phase. Your test is not a valid way to examine the audibility of phase response. Experimenting with all-pass filters might be more meaningful.

Click to expand...

there can create such test tones without phase shift and put it thru a allpass filter later. but when i create such tones with phase shift direct in generator give more control over the phase shift. with a allpass filter it is hard to get a value. that is small. or what allpass filter i need, to get at diffrence between 75 hz and 150 hz of 27 degree ?. I also can hear 17 degree diffrence when level of second harmonic is -12 db. this is also in real world with bass sounds

also the audio scientist tell that in bass humans can hear phase problems much more as in mid or high freq. And the generator test with 1 harmonic simulate when play bass instruments. when do this test with 40 hz so that 2. harmonic is 80 hz, with 27 degree can hear as 2 tones not as a single tone.

The problem with phase shift it give bad feeling in ear which increase when longer hear. also when stop hearing this strange feeling stay some time. hear longer time speaker with changed polarity. the longer you hear the more it sound worse. same is with phase problems too. but i notice a workaround.bass EQ with harmonic EQ. only melda plugins can do that. reduce room modes, such way give more nice phase and sound better. all is cumulative. if the speaker itself have worse phase and add eq make it even more worse and reach the level that brain dont like it. phase linear in bass have problems with transient

When you sum harmonics you create a signal with a different amplitude envelope, your test is for audibility of distortion, not phase.

John Mulcahy said:

When you sum harmonics you create a signal with a different amplitude envelope, your test is for audibility of distortion, not phase.

Click to expand...

Of course, the waveform changes. An ideal speaker adds no phase shift between the fundamental and the harmonic (0° phase). A poor-quality speaker, on the other hand, has uneven group delay: it emits the fundamental tone with a delay, while the harmonic is already present. My test is designed to simulate this by changing the phase of the second harmonic. This way, you can hear the difference in sound quality and in terms of auditory fatigue. It’s not necessary to correct everything; it’s enough to correct just enough so that the brain perceives it as good and no listening fatigue occurs.
One way to measure this and compare speakers would be to use an FSAF signal from the generator with the fundamental at -6 dB and the second harmonic at -18 dB and 0 phase to display the distortion. However, the point is if many humans can hear phase shifts or not.

Translated with DeepL.com (free version)
Dr. Klippel wrote

Phase distortion corresponding with a group
delay variation of 0.4 …2 ms within a critical band change the timbre and roughness
of the sound

Click to expand...

with 75 hz 0.4 ms are ~27 degree phase shift. I can hear easy and lower than 0.4 ms and i am 60 years old. so i think many younger people can hear better than me. I think Dr. Klippel mean with critical bands the bark scale.
bands are in bass. 0-100 100-200

In my test it use neighbour bands. I can also reduce frequency to 48 and 2. harmonic is 96 hz. cause problems too.