View Full Version : Turntable Speed Analysis Tool
Ammonite Audio
03-11-2012, 16:59
Some time ago, there were various annular plots, published here and elsewhere, showing true speed accuracy of various turntables. Does anyone know if the software tool that was used for these plots is available; and if so, where from, how much etc? I'd like to use it on my TD124 and Kenwood & LAD direct drive decks.
Many thanks
Wasn't there a humungous thread on PFM about this? I think Simon SQ may be able to help here?
sq225917
03-11-2012, 17:40
Hugo, pop over to Pinkfishmedia.net and look in the DIY sub-forum, the thread is called 'turntable speed stability' Paulr is the gent you need to speak to.
Mike_New
05-11-2012, 08:29
Hi Shuggie,
About a year ago now or maybe a bit more, there was an ADD in HFW
concerning a lazer speed accuracy/measuring device.
This was in the form of a center weight which emitted a lazer beam at exactly once per revolution controlled by a crystal oscillator, and was acurate to about 0.01% if I recall.
The further away from the center you placed the projection screen (white paper in a dark room?) the more accurate would be the readout.
The only problem is that it cost about $600.00!! at the time.
Ammonite Audio
05-11-2012, 08:54
Hugo, pop over to Pinkfishmedia.net and look in the DIY sub-forum, the thread is called 'turntable speed stability' Paulr is the gent you need to speak to.
Thanks - it seems that I need a test record with a 3k tone, then I can send the files to Paul for analysis. Some of those plots are quite telling!
Stratmangler
05-11-2012, 09:05
Thanks - it seems that I need a test record with a 3k tone, then I can send the files to Paul for analysis. Some of those plots are quite telling!
Does the whole speed testing thing fall down when you consider that the turntable on the cutting lathe probably suffers from the same speed stability issues that you're trying to measure?
sq225917
05-11-2012, 09:33
Stratmangler, the limit of the resolution does eventually become the test record itself. Which is why we tended to use the same record. Shuggie drop me your address via pm and I'll send you the test record once it comes back from Mark.
The level of information provided by a strobe, any strobe, pretty much pales into insignificance when you compare it with the FM demodulation or wavelet analysis that has been performed on the PFM thread. The Fm demod' gives instantaneous speed accuracy measurements at whatever the sampling rate of the D/A process was.
Ammonite Audio
05-11-2012, 09:38
What test record is it? I don't mind buying one.
Mike_New
05-11-2012, 09:56
I would suggest that measuring the speed of any TT to the accuracy that would be indicated by such mathematical analysis, must by it’s very nature be somewhat academic.
The reason being that this exhaustive analysis can only be done on the output from the cartridge at the time of capture of the signal, and must by it’s very nature reflect the conditions only during the signal capture process. Also how does one measure the accuracy of the test track, do the suppliers of the test record provide this data for each recording, to the accuracy necessary for meaningful error correction?
I would think that an accurately produced strobe disk, or indeed the strobe on the SL platter itself is all you really require in the real world.
I think you misunderstand the nature of the measurement that is being made. I have a very accurate strobe and it is only good for setting the basic speed - this measurement gives very much more detail. Whilst the record used does introduce anomalies of its own, those are a constant and, if the same record is used, do not impede comparative measurement.
Stratmangler
05-11-2012, 10:17
Stratmangler, the limit of the resolution does eventually become the test record itself. Which is why we tended to use the same record.
So what you're saying is that all the data that was obtained using a different test record is useless?
What test record is it? I don't mind buying one.
It's this one http://www.analogueseduction.net/test-records-turntable-set-up/analogue-productions-the-ultimate-analogue-test-lp-180g.html
http://cache1.bigcartel.com/product_images/44230771/300.jpg
So what you're saying is that all the data that was obtained using a different test record is useless?
No
Mike_New
05-11-2012, 11:08
Hi H,
I cannot agree with you, whilst you say that the test record used does
introduce anomalies, how can you say that they are constant?? Surely this is a contradiction in itself. The fact of the matter is that unless you have an accurate data readout of the error mapping of the test record then your analysis can only be as good as the first guess of it’s accuracy.
They are a constant between comparisons made. I don't see any contradiction - the anomalies do not vary from play to play - they are constant. If you take a measurement - make a change - and measure again - any errors introduced by the record are duplicated and can easily be ignored. Irrespective of the theory, in actual fact it works very well and is much more useful than traditional wow and flutter measurement (which is what it is really, rather than outright speed measurement). To be honest, it doesn't sound like you have read the pfm thread (it is now very long) - all these issues, and much more, are covered by it.
Stratmangler
05-11-2012, 12:23
To be honest, it doesn't sound like you have read the pfm thread (it is now very long) - all these issues, and much more, are covered by it.
Therein lies the problem - the thread on PFM is too long to wade through.
Has anyone done a précis of the thread, because probably 90% (I'm guessing) of the thread has no useful information.
Yes, it is very long, I completely appreciate that. I wouldn't say 90% is useless info - but I guess it depends what you consider useful and useless.
I have submitted data for my Technics SL-1210 using Simon's record. Fascinating, but the test record does have anomalies and you have to look at the plots with those in mind.
I would say that my plot vindicates the Technics motor and Mike New bearing as being an excellent combination.
Well, one would really have to do the measurements with and without the bearing to be able to draw any conclusions.
Indeed, I probably more than most abused Paul's good will on that thread as I found it an interesting exercise on the PL-71 which in the end correlated with an improvement in SQ and it helped identify an issue with my SP-10 after I repaired one of the motor drive phases....
Well, one would really have to do the measurements with and without the bearing to be able to draw any conclusions.
Amen to that! Too often people assume their 'upgrade' is a great improvement without comparing to the previous state. Maybe it was good all along.
Be good to test my Anni, I'll have to check out the thread.
As a related aside, someone on vinylengine tested his 3 Pink Triangle Anni decks (one DIY, one original, one upgraded by Funk Firm) by simply recording the same track on each. Some resonances inherent in the working of the deck were apparent simply by looking at the audio - the resonances seemed not to be there in the vector'd deck with current feedback PSU K-Drive (compared to the two using servo feedback). Can't discount differences in set-up but still gave a clue to something, whatever it was!
Thinking out loud, could those records with Time Code on be used for a similar thing? They're used as a way of controlling software via a real record deck, e.g. for DJ-ing MP3s http://www.thedjshop.co.uk/digital-dj-midi-controllers/replacement-time-code-vinyl
Too often people assume their 'upgrade' is a great improvement without comparing to the previous state.
Umm, well there's the little matter of it sounding hugely better than before. I didn't need a measurement to establish that.
Umm, well there's the little matter of it sounding hugely better than before. I didn't need a measurement to establish that.
Perhaps the bolts were previously loose from years of use and putting in the new bearing meant everything was tightened up properly again?
I'm only speaking generally, not implying that Mike's bearings aren't an improvement.
There are too many variables to say that changing it was an improvement without going back and changing it back with only that one variable, even if tested by ear.
Perhaps the bolts were previously loose from years of use and putting in the new bearing meant everything was tightened up properly again?
There was nothing wrong with the bolts. An inspection of the two bearings quickly demonstrates the massive difference in engineering standards. The sonic differences are not small.
There was nothing wrong with the bolts. An inspection of the two bearings quickly demonstrates the massive difference in engineering standards. The sonic differences are not small.
Might well sound different to you but the one measurement afterwards is meaningless.
"....yeah, but this one goes up to 11".
That you find it sound better is obviously the important point. Just don't bother with the test unless you just want a certificate for its current measurement.
Just don't bother with the test unless you just want a certificate for its current measurement.
I have an engineering mind and am capable of correlating some measurements with what I hear. Seeing the test results is always interesting, but the listening is all the proof I need. No certificate needed here, we're talking about an appreciation of music after all.
Might well sound different to you but the one measurement afterwards is meaningless.
"....yeah, but this one goes up to 11".
That you find it sound better is obviously the important point. Just don't bother with the test unless you just want a certificate for its current measurement.
Nat - the idea is to compare the modded Technics speed stability to other decks - not to the un-modded techie - although a comparison would be interesting.
Mike_New
05-11-2012, 22:12
We are after all, only trying to measure the parameters and qualities of a simple rotating surface.
There is only one way to definitively measure rotational accuracy and consistency and that is by using a qualified angular diffraction grating. These are used to certify the accuracy of angular and rotational encoders used in practically every application from high precision machine tools and robots, to satellite dish’s. Indeed diffraction gratings themselves are used in high precision units.
For any meaningful measurement based on a test standard, then the known and measured accuracy of that standard must be at least an order of accuracy higher than the predicted accuracy of the measurements we are trying to make, otherwise the result are meaningless.
If test records do have anomalies, then what are we looking at when we a
review the impressive plots and data produced my the resulting mathematical analysis??
I doubt if any test disk, or the lathes which cut the masters, are calibrated in this way.
indeed I doubt if they need to be!!
When we get the results, what happens next week when the mains voltage drops by 10V
or we switch on the aircon.??
The most reliable and meaningful long term approach to asessing the accuracy of your TT, would be to use one of the laser strobe devices I mentioned earlier.
Maybe Marco or Martin could buy one of these devices and rent it out to those who would like to use it.
Lodgesound
05-11-2012, 22:49
A cutting lathe has a servo system fed with a once-round pulse generator along with a diffraction-grating based frequency generator feedback loop.
The master servo reference would be from an outboard sync. pulse generator with an accuracy of around 6 decimal places to it's master reference frequency.
These facilities sound almost insane but are completely standard in a professional installation and ensure an extremely high standard of operational stability that massively exceeds systems in most domestic situations.
Masters made on such lathes are obviously to an exceptionally high standard - most home turntables have rotational errors which you will as a listener never hear or even be aware of in a million years unless a fault develops with your turntables. Technics SL1200 series are extremely stable as are many belt drives.
The biggest problem by far and one that causes severe rotational stability issues is the hole in your record NOT being precisely central relative to the grooves. Even a very slight error here will utterly trounce any performance of the platter itself.
Some professional designs I came across had the ability to offset the centre spindle and had a runout meaurement on the end of the tomearm to minimise this each time a disc was loaded.
It never ceased to amaze me that such a facility was never made available even on turntables costong £10 000 or more.
Eccenticity is the biggest bugbear of playing analogue discs as regards speed stability.
Mike, perhaps you are unaware as to how wow and flutter figures for turntables are normally measured? If you look on YouTube you can see the strobe you describe being used - they show some interesting results - but not the kind of results that Paul's technique gives. I'm not sure why you want to decry this measurement procedure - it is, essentially, a higher resolution version of a recognised process.
________________
Stewart, Nakamichi made a turntable that measured the eccentricity of a record and automatically adjusted the position of the record to correct it. It is also possible to do a similar thing (although manually) on Roksan's decks (as the record spindle is removable).
The Grand Wazoo
05-11-2012, 22:56
It never ceased to amaze me that such a facility was never made available even on turntables costong £10 000 or more.
Lots of people will now be shouting the word 'Nakamichi' at you!
Lodgesound
05-11-2012, 23:25
Yes I know - remember seeing one in tottenham court road at the age of 15 and thinking what a good idea....
Then I got in broadcasting and aquired my first pro tape machine which kind of changed things for me...
Mike_New
05-11-2012, 23:38
Hi Lodgesound,
I must agree with you the hole in the middle and it's diameter, is the single biggest problem to obtaining concentricity of the recording, why don't recording people pay more attention to this? On my Bearings I have specified a spindle diameter which it just larger than the nominal diameter which helps but does not totally illiminate the problem.
Only recently I got some new recordings from the US, where the holes were on the large size.
Some professional designs I came across had the ability to offset the centre spindle and had a runout meaurement on the end of the tomearm to minimise this each time a disc was loaded.
It never ceased to amaze me that such a facility was never made available even on turntables costong £10 000 or more.
Eccenticity is the biggest bugbear of playing analogue discs as regards speed stability.
I think a Nakamichi TT also had such a facility.
Agree re the eccentricity, and of course record warps since no records are perfectly flat. Then, if they aren't flat the cartridge generator bounces at the natural resonant frequency set by the moving mass and hinge compliance causing the stylus to scrub back and forth in the groove as it plays. Anyone for jitter?.......bucketloads of the stuff compared to digital ;)
Good job they still manage to sound good on the whole!
Hi Lodgesound,
I must agree with you the hole in the middle and it's diameter, is the single biggest problem to obtaining concentricity of the recording, why don't recording people pay more attention to this? On my Bearings I have specified a spindle diameter which it just larger than the nominal diameter which helps but does not totally illiminate the problem.
They aren't all even in thickness across the disc either. I've seen some suspended TTs rock the sub-chassis as the record rotates.
Rob.
sq225917
06-11-2012, 09:00
Mike, maybe you should read the thread on PFM, you certainly don't seem to understand the process or the veracity of the results. Sure it can only be as accurate as the test record cut, but then any record can only be as accurate as the master cut so that is the best we can hope for.
By comparing all the decks via the same record, we eliminate any discrepancies due to differing pressings and achieve a flat base line. Paul lines the samples up, so for example the 4-O'clock tick is at the same point on each plot and can be compared directly.
Interrogation of the plot can tell you all of the frequencies other than the fundamental that the deck is experiencing, and in doing show indicates drive anomalies, suspension anomalies, arm and tracking anomalies. The wavelet analysis of my deck indicated that i had a higher level of energy at 33 1/3 than Ynwan did, this indicated I had a noisy bearing- which i then fixed. No spinning laser can tell you that it just tells you if it is above or below a nominal speed, making it a very blunt tool at best.
Not sure I entirely agree with eccentricity causing speed problems. This is a plot Paul took of my KD550, you can see the eccentricity and yet the speed stability is very good. This is using the same test record I used throughout my PL-71 and SP10 testing.
http://i94.photobucket.com/albums/l82/_NRG_/other/KD-550_polar.png
sq225917
06-11-2012, 10:12
kenwood KD 550
Mike_New
06-11-2012, 10:24
Neil,
What are the base units for the KD-55 polar plot?
Neither the X, nor Y, are relevant in this context, except in a comparative sense. I really do urge you to read at least some of the pfm thread (particularly if you wish to criticise the methodology) - it's all explained in detail.
kenwood KD 550
Ah, makes sense.
Mike, I dont think the units represent anything. The plot is take over 5 revolutions of the platter, so its the relative deviation between each rotation thats shown. On some plots you can also see HF noise imposed on the trace of each rotation.
sq225917
06-11-2012, 16:08
Paul varies the axial scale to best suit each polar plot, he can put markings on the axis to give an idea of what the irregularities are. In fact somewhere on that thread he did, 0.1% and 0.01% plots for a basic 50 sinewave deviation.
Powered by vBulletin® Version 4.2.3 Copyright © 2026 vBulletin Solutions, Inc. All rights reserved.