AES6 Wow & Flutter Meter — Python & Standalone Analyzer (2σ Method)

[alvaro-oliver]

Simulated 0.8Hz 0.05% mixed with 4Hz 0.1%
View attachment 460121
same with Mkanes DIN W&F
View attachment 460122


Those who do not think w&Flutter matter much can listen to this...these are not high levels 2S 0.19% according to Mkane and about 0.1% according to the new 2S app

08-005 4-01.wav

drive.google.com

And a good test record on a good DD drive Denon 51F sounds like this,

merolmez 2025 51f shure filter wf 3150.wav

drive.google.com

and it looks like this
View attachment 460125

Thanks for sharing the file! I was curious about those peaks in the synthetic signal — when I opened it in Audacity, I noticed it's clipping. That should definitely be fixed to get an accurate measurement.

 

[alvaro-oliver]

uses ONE channel only and 1khxz high pass filter ,looks more resonable, stil I cannot understand how peak speed cariation is between 3125 abd 3185, shakenspin is nowhere near those values
View attachment 460063

View attachment 460064

Anoter TT and different record..
View attachment 460066

Spoiler: more

another record again
View attachment 460089

View attachment 460090

That’s all related to artifacts from the instantaneous frequency method. it's super sensitive to small phase shifts and noise in the waveform. Your shaknspin uses a gyroscope to measure angular velocity directly, so it’s not affected by these kinds of artifacts.

That said, the actual W&F value filters out most of that noise by design, so it still gives a reliable reading despite the wild-looking trace.

 

[Thomas_A]

Hi Thomas, thanks for sharing. Your measurements look great.

The difference you're seeing between 33.30 and 33.33 RPM across both files can be explained in a few ways. Personally, I’ve noticed that speed and stability tend to improve on my turntable after a few minutes of warm-up, but the W&F measurements remain quite consistent overall.

Regarding the 1 Hz peak, feel free to send me the file if you’d like me to cross-check it using Multi-Instrument.

Best regards,
Álvaro

Since the Axis speed can be adjusted, the better speed was probably after I made adjustment. That said there is some variation which is unavoidable with belt drives. I suspect the 1.1 and 2.2 Hz derives from the test record; while I can reasonably adjust the 0.55 Hz by centering the record, other issues are difficult to adjust for.

 

[alvaro-oliver]

So the synthetic test signal gives +/- 0.0990 % for the script and 0.189 % with MKane's software. Something not computes here.

We don't know if Mkane's software is actually using the 2-sigma method, which makes a big difference. But even if it is, it's important to keep in mind that no two W&F analyzers are likely to give exactly the same result. Standards like DIN, IEC, and AES6 define what should be measured, but not how. And when they do hint at design choices, they refer to analog filter circuits from decades ago, not to modern DSP-based methods.

That means the actual implementation is always up to the designer. The key is to get results that are consistent and comparable, not necessarily identical.

Synthetic signals are great for verifying a method, but even then, slight differences are expected and with real world tones, even more so.

 

[Thomas_A]

We don't know if Mkane's software is actually using the 2-sigma method, which makes a big difference. But even if it is, it's important to keep in mind that no two W&F analyzers are likely to give exactly the same result. Standards like DIN, IEC, and AES6 define what should be measured, but not how. And when they do hint at design choices, they refer to analog filter circuits from decades ago, not to modern DSP-based methods.

That means the actual implementation is always up to the designer. The key is to get results that are consistent and comparable, not necessarily identical.

Synthetic signals are great for verifying a method, but even then, slight differences are expected and with real world tones, even more so.

Ok, yes. Perhaps due to a clipped test signal. Would also like to know if the Mkane 2sigma value is presented as ± or not.

 

[somebodyelse]

We don't know if Mkane's software is actually using the 2-sigma method, which makes a big difference.

We could just ask @pkane

 

[Thomas_A]

Is that the Multitone software by @pkane? As a macintosh user I am limited to Python (meaning the polar plot script) and the NAK T-100 app, none of these have the 2-sigma method. Would be good to get some info on Multitone software and perhaps @stereoplay could compare his software as well. I’ve been looking for a 2-sigma app but not been succesful. This script solves it but I would be nice with some more comparisons.

 

[JP]

Happy to see this thread. Have only glanced, but a nice feature would be to be able to analyze low frequency signals, circa 100Hz, as that’s generally in the neighborhood of DD motor FGs.

 

[Thomas_A]

Just a question, would this thread be better placed in the turntable, cartridge etc forum?

 

[JP]

Yes

 

[pkane]

We could just ask @pkane

Multitone Analyzer is using 2-sigma method.

 

[pkane]

Is that the Multitone software by @pkane? As a macintosh user I am limited to Python (meaning the polar plot script) and the NAK T-100 app, none of these have the 2-sigma method. Would be good to get some info on Multitone software and perhaps @stereoplay could compare his software as well. I’ve been looking for a 2-sigma app but not been succesful. This script solves it but I would be nice with some more comparisons.

MTA will run under Wine on Mac OS, although audio input/output configuration may be difficult. Without audio devices, you can still analyze captured waveforms from a file.

 

[Thomas_A]

MTA will run under Wine on Mac OS, although audio input/output configuration may be difficult. Without audio devices, you can still analyze captured waveforms from a file.

Thanks. I'm quite a novice with command line though Terminal but I can at least try see if I get Wine going. It's an old mac with Catalina, so I am hesitant it will work.

 

[alvaro-oliver]

Happy to see this thread. Have only glanced, but a nice feature would be to be able to analyze low frequency signals, circa 100Hz, as that’s generally in the neighborhood of DD motor FGs.

Hi JP, glad you brought up the motor frequency generator — I’ve actually been working on that.

The captures I posted earlier are from a Reloop RP‑4000 turntable, which outputs a 50 Hz FG signal at 33⅓ RPM. As you’ll notice, there's no obvious peak near that frequency in the unweighted spectrogram.

However, here’s a synchronized capture showing:

• Blue: Audio signal at 3.15 kHz (reference tone), demodulated to extract instantaneous deviation and converted to RPM.
• Brown: FG signal from the turntable, measured using a reciprocal frequency counter that outputs RPM directly.
• Green: Gyroscope signal from a TDK ICG‑20660, converted from °/s to RPM.

And here’s the W&F and spectral analysis of those three signals, using the same method as in the script shared earlier:

All three signals are low-pass filtered at 20 Hz due to the FG's resolution limits, but there’s strong spectral correlation between them. A few observations that stood out:

• The gyroscope and FG match each other remarkably well.
• A ~2.2 Hz peak consistently appears only in the audio signal across all my tests. I’m planning to retest with a new test record (just arrived today) — it might be something specific to the Analogue Productions disc.
• Sensitivity differs by method: audio consistently shows the highest W&F, followed by the gyroscope, and then the FG — always in that order. Interestingly, the ratio between them also stays quite stable, which suggests that all three methods provide consistent measurements, just with different levels of sensitivity.

That’s what I can share for now — I’ve got a lot of material piled up and I’m hoping to publish some of these findings once I upgrade my improvised lab setup and run more structured tests.

Cheers,
Álvaro

 

[alvaro-oliver]

Just a question, would this thread be better placed in the turntable, cartridge etc forum?

Sorry if I posted in the wrong section — I’m new to the forum. I don’t think I have the ability to move the thread myself, but maybe a mod or admin could help with that?
Thanks!

 

[Balle Clorin]

Thanks for sharing the file! I was curious about those peaks in the synthetic signal — when I opened it in Audacity, I noticed it's clipping. That should definitely be fixed to get an accurate measurement.

View attachment 460220

Thank I thought is was just ok , I can fix that

Here is a new file made by combining 2 virtin files - 088Hz 0.05% and 4Hz 0.1%-this time without a time shift, still get those strange peaks. The virtin file may have clipping inherently .CooleditPro give waring of some possible clips

I used to be able to use the Wow and flutter visualisation in Audacity, but after getting a new PC and new installation of Audacity I cannot get the w%F addin to work ,even it addins points to it, I would be grateful for advice on how to get a demodulator/visualizer to work.

virtin mix 08-005 4-01.wav

drive.google.com

 

[Thomas_A]

Thank I thought is was just ok , I can fix that

Here is a new file made by combining 2 virtin files - 088Hz 0.05% and 4Hz 0.1%-this time without a time shift, still get those strange peaks. The virtin file may have clipping inherently .CooleditPro give waring of some possible clips

I used to be able to use the Wow and flutter visualisation in Audacity, but after getting a new PC and new installation of Audacity I cannot get the w%F addin to work ,even it addins points to it, I would be grateful for advice on how to get a demodulator/visualizer to work.

virtin mix 08-005 4-01.wav

drive.google.com

View attachment 460600

View attachment 460599

The 2-sigma is hidden by spectrum?

 

[Balle Clorin]

Barely readable as 0.3

 

[alvaro-oliver]

Thank I thought is was just ok , I can fix that

Here is a new file made by combining 2 virtin files - 088Hz 0.05% and 4Hz 0.1%-this time without a time shift, still get those strange peaks. The virtin file may have clipping inherently .CooleditPro give waring of some possible clips

I used to be able to use the Wow and flutter visualisation in Audacity, but after getting a new PC and new installation of Audacity I cannot get the w%F addin to work ,even it addins points to it, I would be grateful for advice on how to get a demodulator/visualizer to work.

virtin mix 08-005 4-01.wav

drive.google.com

View attachment 460600

View attachment 460599

Hi Balle, I see what you're trying to achieve — but you can’t simply mix two reference tones. Any phase difference between them will result in unpredictable modulation artifacts, both in frequency and amplitude.

I've created a reference tone for you, with modulation at 4 Hz (0.1%) and another at 0.8 Hz (0.05%). In the measurement results, you can clearly see how the unweighted spectrum reveals both modulations and their respective amplitudes.

modulated_tone.wav

Hope this helps!

Álvaro

 

[alvaro-oliver]

I've been trying to use Multitone. Can someone explain the procedure for measuring wow & flutter from an audio file?