Fun with vinyl measurements

[Thomas_A]

Nice! Did you get them 3D printed? Sorry if I missed any earlier discussion of this.

Yes. A friend just got one and he's doing a few for me, different angles. Will test them when he's finished, and it will help me doing some systematic tests. With respect to setup differences, azimuth changes between mounting/remounting may cause some larger shifts as the azimuth gets closer to the ideal. Also as you used, three records as a "mat", may have done something? The curve is basically a power function.

 

[Balle Clorin]

Prototype shim 6 degree. Will get 5,6,7,8,9 degrees.
View attachment 514821

Where did you get that choice?

 

[Thomas_A]

Where did you get that choice?

A friend is making them for me. I also mentioned that there might be a niche market for such shims.

 

[USER]

I'm pretty niche.

 

[Thomas_A]

I'm pretty niche.

I will test them out, he's doing different materials at the moment. Not carbon fibre though.

(I guess he will charge a bit if he does more of them. The pats audio shim is sold out but they take $6 for the 6 degree shim, and the Wally ones are whooping $125 each)

 

[sqrl]

Hi - I've recorded and matched durations and levels for a couple of cymbal-heavy .wav clips with the MP-700 and some alternative sources. How would you like me to compare them or set them up for comparison?

 

[Thomas_A]

Hi - I've recorded and matched durations and levels for a couple of cymbal-heavy .wav clips with the MP-700 and some alternative sources. How would you like me to compare them or set them up for comparison?

You can take a part with cymbal crashes in audacity, split the tracks, invert one channel, make stereo tracks again, mix them to mono and finally do a spectrum analysis. Do the same from the same piece of track from another source or cartridge. Import the spectrum file into REW and plot them both. If the curves match there is no real problem but if the MP700 track is significantly higher the is an out of phase signal or vertical movement of the cantlever causing distortion. One can do the same as above without inverting the channels to look for the lateral cantliever movement.

Cymbal crashes or high-hats or sibilant sounds from voice are good choices.

 

[sqrl]

You can take a part with cymbal crashes in audacity, split the tracks, invert one channel, make stereo tracks again, mix them to mono and finally do a spectrum analysis. Do the same from the same piece of track from another source or cartridge. Import the spectrum file into REW and plot them both. If the curves match there is no real problem but if the MP700 track is significantly higher the is an out of phase signal or vertical movement of the cantlever causing distortion. One can do the same as above without inverting the channels to look for the lateral cantliever movement.

Cymbal crashes or high-hats or sibilant sounds from voice are good choices.

Is it better to start with a mono (without summing to mono) or stereo track?

 

[Thomas_A]

Is it better to start with a mono (without summing to mono) or stereo track?

if you have a mono track from start perhaps. Purpose is to look at either the in or out of phase signal in the recording. It should be appox the same using two different cartridges or a cartridge against a CD track. Best is to use two different cartridges since the source (LP) is the same.

 

[sqrl]

if you have a mono track from start perhaps. Purpose is to look at either the in or out of phase signal in the recording. It should be appox the same using two different cartridges or a cartridge against a CD track. Best is to use two different cartridges since the source (LP) is the same.

Ok - I have a first set of results for the Nagaoka MP-700, to check whether it shows elevated vertical or lateral distortion.

Methods and materials

UK 1979 7" The Police- Walking on the Moon (WOTM)
Technics SL-1210 + MP-700 or Clearaudio Maestro V1 > RME ADI 2/4 > Audacity > 24-96 Hz wav. 45 kHz R load and 210 pF C load (3ft BJC LC2).

WOTM, The Very Best of Sting and the Police, Tidal "max" > Audacity > 24-96 Hz wav
WOTM, The Police - Every Step You Take - The Singles, SACD > Foobar 2000 > 24-96 Hz wav

WOTM is famous for Stuart Copeland's cymbal work.


Loaded into Audacity, clips matched for duration and volume (RMS ± 0.05). I clipped the beginning (low volume) and end (digital versions had a different run out).

Here are the spectrograms :-

High and low passed at 400 Hz (rumble etc) and 19 kHz (12 dB drop below levels at 17 kHz), 48 dB per octave. Then divided the clip into 4 selections :-

You can see selection 1 < 2 ≈ 3 < 4 for high frequency content. If the MP-700 has a problem with HF content, it should be most obvious in Selection 4.

Mixed L and R to mono for lateral content.

Mixed L and inverted R to mono for vertical content

Selections exported as 24-96 Hz wav files and shown in REW.

Results


First, the CD lateral content shows the expected Selection 1 (crimson) < 2 (orange) ≈ 3 (green) < 4 (blue) for high frequency content :-

Next are the lateral selections 1-4. CD (orange), MP-700 (green), Maestro (blue), Tidal (brown)

The MP-700 tracks </≈ to the CD compilation single version and rolls off from around 5 kHz. The Maestro has a steeper roll-off from around 3 kHz.

Next the verticals - we might anticipate Selection 4 to be the worst ...

Except at the 9 kHz anomaly, I can't see see the MP-700 performing badly on this metric against the digital files or the Maestro. It tracks a little better than the Maestro in the 4,500-9,000 Hz range.

Here's the Maestro. The contacts were evenly worn, but I didn't optimize the load and azimuth.

Limitations are that it's only one piece of music, a 7" single, which will be better than an LP. I should have matched the cartridge FR first. The digital files may have been remastered with a treble boost. Nevertheless, I can't see anything egregious compared with the digital releases. The 9 kHz anomaly is certainly there.

 

[Thomas_A]

Thank you,

when it comes to the CD and Tidal versions they are similar in lateral/mono but not the out of phase content, and seems to have higher HF content vs vinyl. Other than it I cannot see any alarming for the MP700 for these tracks. For a better visual however you could do the arithmetic |A/B| using the tracks and choosing MP700/CD, Maestro/CD and MP700/Maestro. Try align the curves at similar 0 level att ≈ 1 kHz with more smoothing. The point is to see deviations over a larger fr range.

 

[sqrl]

Thank you,

when it comes to the CD and Tidal versions they are similar in lateral/mono but not the out of phase content, and seems to have higher HF content vs vinyl. Other than it I cannot see any alarming for the MP700 for these tracks. For a better visual however you could do the arithmetic |A/B| using the tracks and choosing MP700/CD, Maestro/CD and MP700/Maestro. Try align the curves at similar 0 level att ≈ 1 kHz with more smoothing. The point is to see deviations over a larger fr range.

Not sure if this is as requested

 

[Thomas_A]

I don't have original LP or single of that particular one but on two other older used records (best of, and the singles). The section around time 3.15-3.30 min seems best to test. My result does not say much since I do not have a another cartridge or CD to compare but below is the mono lateral (brown) and out-of-phase section (green) 3.17-3.28. @sqrl It's here for download and use for comparison , it will be taken away in a couple of days.

Dropbox

www.dropbox.com

Below os section 3.17-3.28:

Adding graphs for the corresponding selection 4 (≈2.40-3.30 clip) as in post #2630 above.

 

[Thomas_A]

Prototype shim 6 degree. Will get 5,6,7,8,9 degrees.
View attachment 514821

 

[Jimi Floyd]

View attachment 515587

Suggestion: next batch tell him to print a giant "6" on them, or 5 or 9 or whatever "size" they are.

 

[Thomas_A]

Suggestion: next batch tell him to print a giant "6" on them, or 5 or 9 or whatever "size" they are.

I will. Now they are just marked with stickers. Other side. Prototypes to be tested. The four hole ones are for turning them for azimuth purpose. One could make combination shims also.

I don’t know though if he wants to print more of them. I have no idea of the market and what people are willing to buy them for. $8?

 

[sqrl]

I don't have original LP or single of that particular one but on two other older used records (best of, and the singles). The section around time 3.15-3.30 min seems best to test. My result does not say much since I do not have a another cartridge or CD to compare but below is the mono lateral (brown) and out-of-phase section (green) 3.17-3.28. @sqrl It's here for download and use for comparison , it will be taken away in a couple of days.

Dropbox

www.dropbox.com

Below os section 3.17-3.28:

View attachment 515531

Adding graphs for the corresponding selection 4 (≈2.40-3.30 clip) as in post #2630 above.

Hi - did you have a low pass filter at around 18kHz - there's very little output above that?

 

[sqrl]

I'd buy some for $8, if he can do lower values around 2, 2.5 and 3 degrees.

 

[Thomas_A]

Hi - did you have a low pass filter at around 18kHz - there's very little output above that?

No. This compilation LPs from Police are not very good though. Many of the sibilant songs are put as the first tracks on the originals; the compilations are not. I’ve ordered an original now.

 

[sqrl]

Ok. The spectra look a bit like they were eaten by moths, so I low passed filtered the data at 15.5 kHz @ 48 db per octave. And high passed at 400 Hz, as before.

This is the section around 3:15 - 3:32

Lateral

Vertical

The comparisons are normalized at 1 kHz. Here's the lateral

Just the roll-off and the step at 9 kHz for the MP-700 that I can see. These were at on the ADI 2/4, so 45 kΩ and 150 pF, plus measured 221 pF with tt and 3ft of Blue Jeans LC2.


Vertical

The MP-700 is above the CD from around 1-2.6 kHz, but so are the other cartridges - here it is zoomed :

The worn LP does worse than the 7" single above 4 Hz and nothing much for the MP-700 and Maestro, except loading and the 9 kHz blip.

More on that, here's a zoom of the 8-10.5 kHz region at 1/48 smoothing, vertical and lateral.

Based on the sine-wave frequency sweeps, you might expect to see a sudden jump in lateral amplitude, phase error and vertical mistracking at 9 kHz (the peak is 9,050 Hz). Instead, the MP-700 falls between the CD and the Maestro. Music at multiple frequencies doesn't seem to excite the sudden 1.5-2.0 dB peak (it's still clear on the sweeps at 1/48th smoothing) like a sine wave.

What is still evident is a larger fall after the peak, with a trough at ~9.5 kHz.