Close in jitter?

[The Smokester]

Thanks again, Amirm. I was trying to find a figure somewhere that showed a measurement of turntable speed versus time. I remember it being a smooth, continuous, undulating curve. Anyway, I can't find it but in the process of searching I see this topic has been beaten into the ground. Many thanks again.

 

[Don Hills]

... I was trying to find a figure somewhere that showed a measurement of turntable speed versus time. ...

Something like this?
http://archimago.blogspot.com.au/2014/08/measurements-technics-sl-1200-m3d-wow.html

 

[amirm]

Something like this?
http://archimago.blogspot.com.au/2014/08/measurements-technics-sl-1200-m3d-wow.html

Ah, I didn't know folks were doing timing measurements of turntables that way....

 

[The Smokester]

Something like this?
http://archimago.blogspot.com.au/2014/08/measurements-technics-sl-1200-m3d-wow.html

Sort of. Not sure exactly what's in the low pass filter, though. I saw the graph I want but can't find it again. I want the rotational speed of a turntable versus time. A time series of the speed. The graph I saw was smooth...Not chaotic although I know that in a bad TT it can be chaotic as Amirm alludes above.

 

[Fitzcaraldo215]

Changing the graph will simply not change result. I am afraid that any optimism that good, presumably meaning high priced turntables, overcome these issues is misplaced. I think even Michael Fremer's Stereophile measurements more than amply confirm that. So, once again, there is pretty damning confirmation that the whole vinyl thing is about something other than accuracy or fidelity, as if we did not already know.

 

[The Smokester]

Changing the graph will simply not change result. I am afraid that any optimism that good, presumably meaning high priced turntables, overcome these issues is misplaced. I think even Michael Fremer's Stereophile measurements more than amply confirm that. So, once again, there is pretty damning confirmation that the whole vinyl thing is about something other than accuracy or fidelity, as if we did not already know.

Can you please point us to the Fremer Stereophile article?

 

[Fitzcaraldo215]

Can you please point us to the Fremer Stereophile article?

He has recently been publishing graphs related to timing error. Please refer to any turntable review of his over the last 2-3 years at their website.

 

[The Smokester]

He has recently been publishing graphs related to timing error. Please refer to any turntable review of his over the last 2-3 years at their website.

Those look like what Don showed...Not enough technical detail for me to know what I'm looking at.

 

[Fitzcaraldo215]

Sorry. But, the evidence, no matter how you slice or display it, looks pretty damning to me.

 

[Don Hills]

Those look like what Don showed...Not enough technical detail for me to know what I'm looking at.

They make sense to me, the vertical scale is frequency of the test tone, instead of RPM. It's the same thing in different units.
The yellow trace is the deviation of the test tone from nominal over time as read from the disc. The green line is the same data with the noise filtered out. The plugin shows the results expressed according to the DIN / IEC standards for wow and flutter.

 

[amirm]

Sort of. Not sure exactly what's in the low pass filter, though. I saw the graph I want but can't find it again. I want the rotational speed of a turntable versus time. A time series of the speed. The graph I saw was smooth...Not chaotic although I know that in a bad TT it can be chaotic as Amirm alludes above.

The graph I showed from HiFi news uses a very high resolution FFT to show deep detail in very small bandwidth. That is why it looks rough. Using lower frequency resolution will smooth them out.

 

[Blumlein 88]

You can download a free version of Platter Speed for Android. Doesn't have the DIN wow and flutter ability without a purchase. Tried it with a digital 3150 hz signal. It is picky enough if you try and handhold your phone you will get wiggles in the graph you don't get if you sit your phone on a table or other stable object. It can be effected by noise especially low frequencies as a train passed (I am just lucky that way to have that happen the third time I ran it). With the train it showed a rough graph and avg value of 3152.4 hz. Without noise and steady positioning of the phone, it has a blip when it starts recording on the graph and otherwise shows a straight flat line for for raw and for notch filtered. Don't know if this is different in a new update or just on Android, but rather than filtered it shows notched which is described as notched out at +/- .55 hz. I would get an avg value of 3150.0 everytime. Max notched and raw deviations ran between .00% and .04% rather at random if you kept repeating the test. Far cleaner than what is being shown for a TT. Not surprising.

 

[amirm]

As an interesting aside, someone had that app on his iphone at VPI event. He put it on a quartz controlled VPI turntable and the two disagreed with speed on the second decimal place. Not sure which one was wrong, the table or app.

 

[Blumlein 88]

As an interesting aside, someone had that app on his iphone at VPI event. He put it on a quartz controlled VPI turntable and the two disagreed with speed on the second decimal place. Not sure which one was wrong, the table or app.

That would only be a 100 ppm difference. Good quality DACs can vary by that and sometimes more. If the VPI readout and iPhone were that close that is not bad at all.

You can get some pretty nice strobe tuning apps for musical instruments patterned after the Peterson mechanical strobe tuners. Those you can then calibrate by playing a tone from other gear like your very well timed DAC. Those sometimes disagree with the inexpensive clip on tuners many use for stringed instruments. Every musician assumes my app is wrong, but I don't know about that.

 

[The Smokester]

The graph I showed from HiFi news uses a very high resolution FFT to show deep detail in very small bandwidth. That is why it looks rough. Using lower frequency resolution will smooth them out.

Amirm, In my mind, an Fourier transform is a list of complex numbers which I can retransform back into the original data. What you are showing is the square of the magnitudes.

 

[The Smokester]

They make sense to me, the vertical scale is frequency of the test tone, instead of RPM. It's the same thing in different units.
The yellow trace is the deviation of the test tone from nominal over time as read from the disc. The green line is the same data with the noise filtered out. The plugin shows the results expressed according to the DIN / IEC standards for wow and flutter.

Thank you for your patience. There is no limit to the depths of my naivety in this. So there is a 4252.5 Hz tone on an LP I guess. Is that oscillating behavior of the yellow curve about the mean frequency then reflecting changes in the instaneous speed of the turntable? Or is it a measurement artifact?

 

[amirm]

Amirm, In my mind, an Fourier transform is a list of complex numbers which I can retransform back into the original data. What you are showing is the square of the magnitudes.

The imaginary component of the complex numbers show the phase which we don't care about. What is shown is just the "real" component which represents the amplitude of the basis function. We don't need to maintain the phase because we don't need to invert back to time domain.

 

[The Smokester]

The imaginary component of the complex numbers show the phase which we don't care about. What is shown is just the "real" component which represents the amplitude of the basis function. We don't need to maintain the phase because we don't need to invert back to time domain.

I guess what I was trying to see was some original data. What Don has referenced may be what we need.

Looking at the figures in this link which Don supplied above:

http://archimago.blogspot.com.au/2014/08/measurements-technics-sl-1200-m3d-wow.html

The data required depends on the answer sought.

 

[Don Hills]

Thank you for your patience. There is no limit to the depths of my naivety in this. So there is a 4252.5 Hz tone on an LP I guess. Is that oscillating behavior of the yellow curve about the mean frequency then reflecting changes in the instaneous speed of the turntable? Or is it a measurement artifact?

Actually, it's a 3150 Hz tone. Play the disc at 33 RPM and that's what you get. Play it at 45 RPM and you get 4252.5 Hz tone... Yes, the curves represent the instantaneous frequency as read from the disc over time (60 seconds in the tests shown), NOT the actual tone waveform.

The raw curve is noisy because the read from the disc is "jittery". Most of the yellow curve is turntable speed variations (wow). The irregular parts of the waveform are caused by other effects. For example, if the record is not perfectly flat, the cartridge will "bounce" up and down. Because the cantilever in the cartridge is not parallel to the record, when the cartridge goes down towards the record, the stylus tip will move slightly towards the direction of rotation. The movement is subtracted from the platter speed, so there will be a momentary drop in the frequency read. The opposite happens when the cartridge goes up. Sideways movements have a similar but lesser effect.

Filtering out this "noise" produces the green curve which better represents the actual speed variation of the platter.

 

[Cosmik]

This is how audiophilia works...

People espouse science and objectivity, and acknowledge the existence of the 'placebo effect'. But in parallel they clearly believe that if audiophiles love vinyl, 'there must be something in it'. And so a science-based forum still finds itself discussing vinyl year after year...

If people really believed in science, vinyl would not warrant any more attention than 78s.