Axe-Fx III Firmware Release Version 12.08 Public Beta 3

Status
Not open for further replies.
Furthermore, "dumb" autocorrelation methods like this are mostly useless for determining "pitch". They may find a frequency but are extremely prone to octave errors. A guitar note often has a 2nd harmonic that has more energy than the fundamental. Play an A at the 2nd fret on the G string. The 2nd harmonic is usually 10-20 dB greater than the fundamental. This method will not find the fundamental.

And you assumed I am using "dumb" autocorrelation ?

One of the biggest mistakes an engineer can make is thinking they know more than they do. This happens in any discipline but some engineers seem especially prone to this. Humility is a virtue. I'd probably hire someone like you if you were looking for a job as you seem motivated. But you need to understand your limitations. No matter how much you think you know, someone out there knows more about a particular subject than you do.

That I agree. I am aware of the Dunning-Kruger effect and never did I at any point claim that I know more than anybody. I do understand my limitations. And no I am not looking for a job.

Yes humility is a virtue, which apparently you do not show.
 
https://patents.google.com/patent/US4429609A/en?oq=4429609

At 44.1 kHz sample rate this is around 11 us. The period of a sine wave at 440 Hz is 22 ms. 11 us of error may not seem like a lot but it's actually 5 cents of error. This is unacceptably high for most applications. The error gets progressively worse as you go up in frequency. The error at 820 Hz will be ten times worse than the error at 82 Hz. The basic problem is that the autocorrelation is quantized. This isn't a problem if you are using a significant number of bits to represent the signal as you can interpolate the autocorrelation (which is what we do) but with a one-bit representation you cannot interpolate.

I do not use pure sine waves for testing. I test using high second harmonics (higher than fundamental). There are tests that
use various phase offsets and harmonics, even with missing fundamental and with injected noise.

Moreover, I also test using REAL guitar recordings.

The default is:

_2nd_harmonic = 2.0f; // Second harmonic multiple
_3rd_harmonic = 3.0f; // Second harmonic multiple
_1st_level = 0.3f; // Fundamental level
_2nd_level = 0.4f; // Second harmonic level
_3rd_level = 0.3f; // Third harmonic level

And the results:

TEST_CASE("Test_middle_C")
Average Error: 0.001236 cent(s)
Min Error: 0.00009617 cent(s)
Max Error: 0.00231753 cent(s)

TEST_CASE("Test_middle_A")
Average Error: 0.001998 cent(s)
Min Error: 0.0 cent(s)
Max Error: 0.00552 cent(s)

TEST_CASE("Test_low_E")
Average Error: 0.00003447 cent(s)
Min Error: 0.00003447 cent(s)
Max Error: 0.00003447 cent(s)

TEST_CASE("Test_E_12th")
Average Error: 0.0000359 cent(s)
Min Error: 0.00003447 cent(s)
Max Error: 0.0001258 cent(s)

TEST_CASE("Test_E_24th")
Average Error: 0.00018 cent(s)
Min Error: 0.00003447 cent(s)
Max Error: 0.000606 cent(s)

TEST_CASE("Test_A")
Average Error: 0.00012 cent(s)
Min Error: 0.00012 cent(s)
Max Error: 0.00012 cent(s)

TEST_CASE("Test_A_12th")
Average Error: 0.000001413 cent(s)
Min Error: 0.000120 cent(s)
Max Error: 0.00012 cent(s)

TEST_CASE("Test_A_24th")
Average Error: 0.000264 cent(s)
Min Error: 0 cent(s)
Max Error: 0.00108 cent(s)

TEST_CASE("Test_D")
Average Error: 0.000189 cent(s)
Min Error: 0.0000207 cent(s)
Max Error: 0.000339 cent(s)

TEST_CASE("Test_D_12th")
Average Error: 0.000733 cent(s)
Min Error: 0.0000207 cent(s)
Max Error: 0.00164 cent(s)

TEST_CASE("Test_D_24th")
Average Error: 0.00151 cent(s)
Min Error: 0.0000207 cent(s)
Max Error: 0.009556 cent(s)

TEST_CASE("Test_G")
Average Error: 0.0000601 cent(s)
Min Error: 0.0000601 cent(s)
Max Error: 0.0000601 cent(s)

TEST_CASE("Test_G_12th")
Average Error: 0.000206 cent(s)
Min Error: 0.0000601 cent(s)
Max Error: 0.00021 cent(s)

TEST_CASE("Test_G_24th")
Average Error: 0.000345 cent(s)
Min Error: 0.0000601 cent(s)
Max Error: 0.00668 cent(s)

TEST_CASE("Test_B")
Average Error: 0.000643 cent(s)
Min Error: 0.00000166 cent(s)
Max Error: 0.0015 cent(s)

TEST_CASE("Test_B_12th")
Average Error: 0.00436 cent(s)
Min Error: 0.00000166 cent(s)
Max Error: 0.0104 cent(s)

TEST_CASE("Test_B_24th")
Average Error: 0.00332 cent(s)
Min Error: 0.00000166 cent(s)
Max Error: 0.0295 cent(s)

TEST_CASE("Test_high_E")
Average Error: 0.00076 cent(s)
Min Error: 0.0000344 cent(s)
Max Error: 0.00308 cent(s)

TEST_CASE("Test_high_E_12th")
Average Error: 0.00282 cent(s)
Min Error: 0.0000345 cent(s)
Max Error: 0.00805 cent(s)

TEST_CASE("Test_high_E_24th")
Average Error: 0.01705 cent(s)
Min Error: 0.000286 cent(s)
Max Error: 0.0404 cent(s)
 
I wasn't referring to that but to the hi cut in the axe fx reverb.
That's the main difference I hear in your samples: less highs in Space Designer.
Well as I wrote multiple times already, my concern is NOT about the EQ curve, but about the artificial sounding modulation and slight flanging effect in the reverb decay which I never ever heard in any real spring reverb. But everyone else here seems to think that a real spring reverb show the same artifacts, so I'll keep my mouth shut on this subject.
 
Well as I wrote multiple times already, my concern is NOT about the EQ curve, but about the artificial sounding modulation and slight flanging effect in the reverb decay which I never ever heard in any real spring reverb. But everyone else here seems to think that a real spring reverb show the same artifacts, so I'll keep my mouth shut on this subject.
It's not about shutting up when there's issues not being confirmed by others. You may want to post in the "bugs" section.

I've seen others including Cliff himself confirm at first not happening on their end and then later on confirming there is something going on.

Also ...it may be an issue just on your end and all the great Fractal elite folks here are all about helping you figure it out if that's the case.
 
  • Like
Reactions: jon
It's not about shutting up when there's issues not being confirmed by others. You may want to post in the "bugs" section.

I've seen others including Cliff himself confirm at first not happening on their end and then later on confirming there is something going on.

Also ...it may be an issue just on your end and all the great Fractal elite folks here are all about helping you figure it out if that's the case.
The only "help" I got so far is the proposal to change the EQ curve which I know is not the problem for me. And I don't think this is a "bug" (quite obviously other people here really like how the Spring Reverb sounds) so I don't see the point in opening a thread in the bugs section for this.
 
Well as I wrote multiple times already, my concern is NOT about the EQ curve, but about the artificial sounding modulation and slight flanging effect in the reverb decay which I never ever heard in any real spring reverb. But everyone else here seems to think that a real spring reverb show the same artifacts, so I'll keep my mouth shut on this subject.
I was just asking if you tried that before dismissing it, a more aggressive high-cut could help mask that modulation you don't like, besides making it sound closer to Sound Designer
 
Unlikely. There will probably be another beta before a release version.

I've been slaving away all week on the pitch shifter. It's very, very good now. I'm now using a wavelet approach for the pitch detection. This provides even faster detection than before and increases the low frequency detection by an octave which makes shifting chords smoother.

I still want to do an overhaul of the Chorus using the techniques that were used in the Flanger. I also have some ideas on reverb (although the reverb is pretty great already).

Is there any advantage with the current algorithm vs the new version you are working on? The single note tracking in beta 3 is pretty amazing!
 
Is there any advantage with the current algorithm vs the new version you are working on? The single note tracking in beta 3 is pretty amazing!

A quote from Cliff on May 11th:

"I extended the range of the pitch detector by another octave today and that reduces the warbling considerably when playing chords with dissonant intervals. A side-effect what I did is that it further reduces the detector latency. "

Shit's about to go crazy.
 
Status
Not open for further replies.
Back
Top Bottom