Phono Cartridge Response Measurement Script

[Thomas_A]

many simple and easy tests exist on different test discs for antiskating or see vtf....
without measure..., are pretty self-explanatory...

it was approaches to the golden age of vinyl...

;-)

As mentioned elsewhere the optimal VTF/antiskate is when the armature is centered running in groove with nominal/average friction. I have not seen such a setup test though, other than using a microscope to watch/measure the stylus deflection during skate force.

 

[Newman]

That's just marketing material, and of note they don't call it a special line contact on that page, only line contact.

Incorrect.

I'll have to check Ray's thread on VE to see if anyone has photographed or measured the AT SLC.

That would be very interesting.

To date I'm not aware of any data that says the SLC is superior to the ML (MR). They price the Shibata higher than ML as well and we know that isn't superior.

In their top MM line they price the SLC $200 more than the Shibata,

And in the OC9 line compared to ML pricing, the Shibata is +$200 and the SLC an additional +$150

but for the ART9 the XA and XI are the same price...

...OTOH the motor systems are different, maybe the XA air-cored tech is more expensive to make, so not apples-to-apples.

I see the SLC profile is their only one that uses a rectangular shank instead of square or round. Whether this is more exotic or expensive to do, I don't know.

 

[morillon]

As mentioned elsewhere the optimal VTF/antiskate is when the armature is centered running in groove with nominal/average friction. I have not seen such a setup test though, other than using a microscope to watch/measure the stylus deflection during skate

then buy some tuning discs from the 70s early 80s
;-)
ps
thd
imd
etc
distortiometer
;-)
even the recent hfn offers an antiskating adjustment mode at 315hz ... (the same since... the 70s ...) face b
it is efficient and simple and..pedagogical
.... but very discussed by the use of test at high level..
or "the ultimate" with only a +12 not so hard..probably enought..
(Post in thread 'Help diagnosing distortion problem in LP playback, please.' https://audiosciencereview.com/foru...blem-in-lp-playback-please.18223/post-1534058)

 

[Thomas_A]

then buy some tuning discs from the 70s early 80s
;-)
ps
thd
imd
etc
distortiometer
;-)

I have a few discs, but none tells me if the cantilever/magnet is centered in the armature. Distortion is a function of VTF also due to other reasons such as tracking ability. IMD sidebands vs VTF.

 

[morillon]

I don't ignore the discussions etc on these all subjects...
(https://www.theabsolutesound.com/ar...proper-cartridge-set-up-fremer-at-fliax-2023/ )
;-)
but already reconciling the classic precautions and approaches makes it possible to broaden things considerably.
in the world of lp everything is "relative".. from protraction, if not tangential, to all the settings..
the lp support is relative in its design, its realization
...
manufacturing cartdrige "by hand" etc.
so
you use classical methode for help settings with simple distortiometres thd imd?
,-)

 

[Thomas_A]

I don't ignore the discussions etc on these subjects...
;-)
but already reconciling the classic precautions and approaches makes it possible to broaden things considerably.
in the world of lp everything is "relative".. from protraction, if not tangential, to all the settings..
the lp support is relative in its design, its realization
...
manufacturing cells "by hand" etc.
so
you use classical methode for help settings with simple distortiometres thd imd?
,-)

So which VTF is optimum? The highest possible?

Antiskate?

 

[morillon]

am not an expert.. do not wish to become one etc.
;-)
ps.. have you tried the ultimate d antiskating test? by ear (or thd "stereo")
or even type of approach on a 315hz more or less at the level?



modestly..
(and you don't want to have badly made cartriges)
;-)

(here. there is a subject that might interest you.. see the design and realization approach to lyra cartriges)

 

[Thomas_A]

am not an expert.. do not wish to become one etc.
;-)
ps.. have you tried the ultimate d antiskating test? by ear (or thd "stereo")
or even type of approach on a 315hz more or less at the level?



modestly..
(and you don't want to have badly made cartriges)
;-)

(here. there is a subject that might interest you.. see the design and realization approach to lyra cartriges)

You mean testing with level?
Here ya go:
OM40; AS 0.5, 1.0, 1.5 (blue filled triangle should be 3rd harmonic also in first panel):

OM10; AS 0.5, 1.0, 1.5:

So which antiskate is best? 0.5 for OM40 and 1.0 for OM10? ;-)

 

[morillon]

try a no-measurement approach, minimizing the force applied on a dedicated test such as ultimate..
it may avoid, as with the story of crosstalk, hours of measurements without much relevance on unsuitable tests
;-)
in the end it's not that complicated, (not enough?), see very simple and.. rather enough...
whether on modest om or ambitious cartdriges
(antiskating is both easier the old way...but also the most debatable...not to mention the potential "defects" in the realization of cartdriges...which now animates passions...discussions...and... a certain new business)
;-)

 

[JP]

Incorrect.

That would be very interesting.

And in the OC9 line compared to ML pricing, the Shibata is +$200 and the SLC an additional +$150

...OTOH the motor systems are different, maybe the XA air-cored tech is more expensive to make, so not apples-to-apples.

I see the SLC profile is their only one that uses a rectangular shank instead of square or round. Whether this is more exotic or expensive to do, I don't know.

The first sentence in the Line Contact description here says the same thing as your link, but there they rank line contact below ML(MR) which has generally always been the case.

I've not read everything here in detail, but it's one of the most comprehensive listings of the various profiles from different cartridge/stylus manufactures I've seen.

Generally the better styli have square shanks, the smaller the better so lower moving mass can be achieved. I'm looking in to a re-tip for a cartridge that purportedly has a 0.06mm square shank where the smallest MR I chan get is .08mm. I think it'll physically fit, but at 0.077mg ETM I'm very concerned about the affect on the system.

 

[Thomas_A]

try a no-measurement approach, minimizing the force applied on a dedicated test such as ultimate..
it may avoid, as with the story of crosstalk, hours of measurements without much relevance on unsuitable tests
;-)
in the end it's not that complicated, (not enough?), see very simple and.. rather enough...
whether on modest om or ambitious cartdriges
(antiskating is both easier the old way...but also the most debatable...not to mention the potential "defects" in the realization of cartdriges...which now animates passions...discussions...and... a certain new business)
;-)

I've done many tests in the past and I usually use the blank surface method + microscope deflection test to set AS.

To question the high-level antiskate setting test. What is the friction force at +18 dB vs -5 dB 315 Hz track? VTF is another question though.

 

[morillon]

so you haven't tried a dedicated test...

(forget the blank and 15db is already rare conditions.. 18db torture with wear etc)

 

[Thomas_A]

so you haven't tried a dedicated test...

View attachment 275485

What's a dedicated test?

 

[cport101]

@JP -- some "use-case" suggestions you might find interesting (or not):

- Add the ability to flag which trace "sounds better (less distortion)" [ a subjective note [8 ascii characters]: SUBJ_NOTE = "GOOD|BETTER|BEST|BAD|STINKS"
- Add the ability to flag/comment on Azimuth settings [ degree of rotation (from last position or relative to 90 degrees vertical)]
- Add ability to see (plot) a "diff" -- plot the frequency/crosstalk/distortion diff between one run and another -- make clear you are seeing the diff

 

[JP]

@JP -- some "use-case" suggestions you might find interesting (or not):

- Add the ability to flag which trace "sounds better" [ a subjective note [8 ascii characters]: SUBJ_NOTE = "GOOD|BETTER|BEST|BAD|STINKS"
- Add the ability to flag/comment on Azimuth settings [ degree of rotation (from last position or relative to 90 degrees vertical)]
- Add ability to see (plot) a "diff" -- plot the frequency/crosstalk/distortion diff between one run and another -- make clear you are seeing the diff

• This is a measurement tool.
• Would only be relevant to that copy (those copies, really) on that setup. Even if the desire is to detect patterns amongst copies, there's too much reliance on the arm/headshell AZ being spot-on.
• The ability to plot multiple combinations of files is on the list - just working through larger structural issues in my head first. It is really easy to do by hacking some stuff around a little; this was one of the primary motivations to refactoring the plot data creation in to functions. The stories for it can be quite varied which makes a sound stab at the structures a bit more important.

 

[Balle Clorin]

The 1kHz reference tracks on the current Ortofon record are pretty good. The don't all agreed so it's pick your poison, but go with something that can show ~-35dB. I know someone who uses UTR and say they get good crosstalk valued with that one as well.

Tangentially I got a shipment notification for my PTG/II from CD Japan today. Funny as the second cart I bought circa 2010 was a PTG/II.

Looking forward to see how it compares with mine…

 

[Balle Clorin]

Direct comparison of CA-TRS1007 and Denon XG7001, 1 kHz-20 kHz.
Left and right:
View attachment 275290View attachment 275291
Crosstalk in left is quite different, similar in right. Distortion generally similar but consistently poorer in right channel. Has anyone tested antiskate force vs distortion and channel?

Yes, this how I did it..

Spoiler: anti-skate

 

[JP]

Looking forward to see how it compares with mine…

It's in the country so should be here tomorrow or Thursday. When I'll have time to mount it is an entirely different matter.

 

[JP]

• This is a measurement tool.
• Would only be relevant to that copy (those copies, really) on that setup. Even if the desire is to detect patterns amongst copies, there's too much reliance on the arm/headshell AZ being spot-on.
• The ability to plot multiple combinations of files is on the list - just working through larger structural issues in my head first. It is really easy to do by hacking some stuff around a little; this was one of the primary motivations to refactoring the plot data creation in to functions. The stories for it can be quite varied which makes a sound stab at the structures a bit more important.

EDIT: As example this is how I did the spot frequency rundown plot.

if __name__ == "__main__":

    infoline = 'V15VMR / Spot Rundown'

    riaamode = 0 
    riaainv =  0
    str100 = 1

    str100fmax = 22000
    _FILE = 'St-V15VMR-STR170-20k.wav'
    input_sig, Fs, minf, maxf = openaudio(_FILE)
    freqout1, ampout1, freqoutx1, ampoutx1, freqout2h1, ampout2h1, freqout3h1, ampout3h1 = createplotdata(input_sig, Fs)

    str100fmax = 11000
    _FILE = 'St-V15VMR-STR170-10k.wav'
    input_sig, Fs, minf, maxf = openaudio(_FILE)
    freqout2, ampout2, freqoutx2, ampoutx2, freqout2h2, ampout2h2, freqout3h2, ampout3h2 = createplotdata(input_sig, Fs)
 
    str100fmax = 5500
    _FILE = 'St-V15VMR-STR170-5k.wav'
    input_sig, Fs, minf, maxf = openaudio(_FILE)
    freqout3, ampout3, freqoutx3, ampoutx3, freqout2h3, ampout2h3, freqout3h3, ampout3h3 = createplotdata(input_sig, Fs)

    str100fmax = 3300
    _FILE = 'St-V15VMR-STR170-3k.wav'
    input_sig, Fs, minf, maxf = openaudio(_FILE)
    freqout4, ampout4, freqoutx4, ampoutx4, freqout2h4, ampout2h4, freqout3h4, ampout3h4 = createplotdata(input_sig, Fs)

    str100fmax = 2200
    _FILE = 'St-V15VMR-STR170-2k.wav'
    input_sig, Fs, minf, maxf = openaudio(_FILE)
    freqout5, ampout5, freqoutx5, ampoutx5, freqout2h5, ampout2h5, freqout3h5, ampout3h5 = createplotdata(input_sig, Fs)

    str100fmax = 1100
    _FILE = 'St-V15VMR-STR170-1k.wav'
    input_sig, Fs, minf, maxf = openaudio(_FILE)
    freqout6, ampout6, freqoutx6, ampoutx6, freqout2h6, ampout2h6, freqout3h6, ampout3h6 = createplotdata(input_sig, Fs)
  
    str100fmax = 600
    _FILE = 'St-V15VMR-STR170-500.wav'
    input_sig, Fs, minf, maxf = openaudio(_FILE)
    freqout7, ampout7, freqoutx7, ampoutx7, freqout2h7, ampout2h7, freqout3h7, ampout3h7 = createplotdata(input_sig, Fs)


    i = find_nearest(freqout7, 500)
    i2 = find_nearest(freqout6, 500)
    norm = ampout6[i2] - ampout7[i]
    ampout6 = ampout6-norm
 
    i = find_nearest(freqout6, 1000)
    i2 = find_nearest(freqout5, 1000)
    norm = ampout5[i2] - ampout6[i]
    ampout5 = ampout5-norm

    i = find_nearest(freqout5, 2000)
    i2 = find_nearest(freqout4, 2000)
    norm = ampout4[i2] - ampout5[i]
    ampout4 = ampout4-norm
 
    i = find_nearest(freqout4, 3000)
    i2 = find_nearest(freqout3, 3000)
    norm = ampout3[i2] - ampout4[i]
    ampout3 = ampout3-norm

    i = find_nearest(freqout3, 5000)
    i2 = find_nearest(freqout2, 5000)
    norm = ampout2[i2] - ampout3[i]
    ampout2 = ampout2-norm

    i = find_nearest(freqout2, 10000)
    i2 = find_nearest(freqout1, 10000)
    norm = ampout1[i2] - ampout2[i]
    ampout1 = ampout1-norm

...

    ax1.semilogx(freqout,ampout,color = 'b', label = '20kHz Spot')
    ax1.semilogx(freqout3,ampout3,color = 'b', label = '5kHz Spot')
    ax1.semilogx(freqout5,ampout5,color = 'b', label = '2kHz Spot')
    ax1.semilogx(freqout2,ampout2,color = 'r', label = '10kHz Spot', linestyle = 'dashed')
    ax1.semilogx(freqout4,ampout4,color = 'r', label = '3kHz Spot', linestyle = 'dashed')
    ax1.semilogx(freqout6,ampout6,color = 'r', label = '1kHz Spot', linestyle = 'dashed')
    ax1.semilogx(freqout7,ampout7,color = 'b', label = '500Hz Spot')

Parameters could be done in a list of tuples and looped through for the plot data, as could creating the plot axis, though the latter is far more nuanced.

I think the first order of business is to consolidate the data lists/arrays in to a 2-D array as it's far easier to hack stuff around when the entirety of the data for a signal is contained in one object. A lot easier to manipulate the data sets as well. Downside is that it's far less tangible for a novice to navigate.

 

[cport101]

I have a few discs, but none tells me if the cantilever/magnet is centered in the armature. Distortion is a function of VTF also due to other reasons such as tracking ability. IMD sidebands vs VTF.
View attachment 275423View attachment 275424

How are you producing these graphs? Script based as well ?