Phono Cartridge Response Measurement Script

[morillon]

thank you...
I will read with interest your feedback on the evolution of your tests....
especially as in the state your efforts are already rather convincing
;-)

 

[cgallery]

I was experimenting as I had nothing to lose. But, yeah, I put a little Sonax gummipfleger on the suspension with a small q-tip as the issue was clearly that it hardened. I put on a little at a time and measured the results after a few hours to see if anything was happening. I also adjusted the set-screw. I don't know the long term effects so I do not vouch for this and can only say do the same at your own risk.

The main point is that these graphs can help you do all sorts of things. I cringe when I read people writing that they do similar things but only confirm by ear.

Edit: Looking at the graph again and now knowing about the ~0.5dB dip between 5-10k, I can see that it still isn't right. (Though it is much better.) I'll have to experiment with the screw some more.

I'd have figured hardened rubber on an Audio Technica would result in a rising response. I always thought the suspension was primarily the wire and the donuts providing damping. I've purchased dozens of vintage Audio Technica styli though and have never experienced a suspension issue like that, though, it is one of the reasons I appreciate the vintage A-T stuff.

 

[USER]

I'd have figured hardened rubber on an Audio Technica would result in a rising response. I always thought the suspension was primarily the wire and the donuts providing damping. I've purchased dozens of vintage Audio Technica styli though and have never experienced a suspension issue like that, though, it is one of the reasons I appreciate the vintage A-T stuff.

My language wasn't precise. I should have said something like suspension damper instead. I'll correct it.

 

[cgallery]

My language wasn't precise. I should have said something like suspension damper instead. I'll correct it.

Not at all, I understood exactly your point and appreciated the data point. Sorry for cluttering the thread.

 

[narud]

i wish i could post some plots for you guys, but im on a windows 7 machine and between trying to download old versions of python and me having absolutely zero clue as to what im doing, its just not looking like its going to happen.

i can provide some reference though with the repaired stylus compared to all the rest i own using plots from a music sample. jp plotted this for me which is my oldest stylus which is visibly worn under the microscope and has audibly more surface noise than the rest of my styli which have either pretty much unused diamonds or one that has seen very light use. my tascam interface has a roll off after 15k or so for some reason, so thats what youre seeing.

heres that stylus and 3 of my others that pretty much match it. two of them are new and one has seen light use (all of these have been looked at under the scope). im only showing the high end because all of them didnt show any significant difference below whats shown.

on the next graph i included another stylus with a fresh diamond, but this one is slightly brighter.

the last graph is all of the styli i have with fresh diamonds. the 3 labeled new appear unused under the scope, 1 has seen light use, and the last one is the repair which also has an unused diamond under the scope.

the repaired stylus is the brightest of the bunch, but who knows how much is because of the repair. if i didnt have that one new stylus (shure new 2) that was a little brighter than the pack, i would have been ready to blame all the brightness on the repair. maybe the repaired stylus was bright to begin with like new 2 and the repair exacerbated it?

 

[Digital Mastering System]

I think we can all agree - this whole stylus - soft vinyl groove thing is a disaster. I am so happy I have Roon, Chromecast, and a good DAC. (and I have a SP-10 with a V15-MR Jico, - which is never used.)

 

[USER]

i wish i could post some plots for you guys, but im on a windows 7 machine and between trying to download old versions of python and me having absolutely zero clue as to what im doing, its just not looking like its going to happen.

i can provide some reference though with the repaired stylus compared to all the rest i own using plots from a music sample. jp plotted this for me which is my oldest stylus which is visibly worn under the microscope and has audibly more surface noise than the rest of my styli which have either pretty much unused diamonds or one that has seen very light use. my tascam interface has a roll off after 15k or so for some reason, so thats what youre seeing.
View attachment 261889
heres that stylus and 3 of my others that pretty much match it. two of them are new and one has seen light use (all of these have been looked at under the scope). im only showing the high end because all of them didnt show any significant difference below whats shown.
View attachment 261890

on the next graph i included another stylus with a fresh diamond, but this one is slightly brighter.
View attachment 261891

the last graph is all of the styli i have with fresh diamonds. the 3 labeled new appear unused under the scope, 1 has seen light use, and the last one is the repair which also has an unused diamond under the scope.
View attachment 261892

the repaired stylus is the brightest of the bunch, but who knows how much is because of the repair. if i didnt have that one new stylus (shure new 2) that was a little brighter than the pack, i would have been ready to blame all the brightness on the repair. maybe the repaired stylus was bright to begin with like new 2 and the repair exacerbated it?

It would be helpful to see a wider range, say 1-20k. Assuming that's the CBS test record, they all have issues in that region you highlight so it is difficult to draw any conclusions from what is seen. It's normal to see minute differences between the styli so I wouldn't say the differences in the first are related to wear. The second however seems consistent with what I have seen.

Here is a comparison of different VN5MR styli: https://www.audiosciencereview.com/forum/index.php?threads/mm-vs-mi-vs-mc.18636/page-63#post-1383126

The differences there are all in the range you show. Perhaps we can continue this discussion there.

And this is what a super worn stylus look like on the graph. It's not so much FR that changes as it is distortion. Of course, above 10k could be affected but it is too difficult to say unless there are before and after measurements of the same cartridge.

To help move things back to the topic at hand, I do think the script can help us "see" wear in this way. I have measured all sorts of cheap cartridges and I rarely see distortion above -18dB if ever.

 

[narud]

It would be helpful to see a wider range, say 1-20k. Assuming that's the CBS test record, they all have issues in that region you highlight so it is difficult to draw any conclusions from what is seen. It's normal to see minute differences between the styli so I wouldn't say the differences in the first are related to wear. The second however seems consistent with what I have seen.

Here is a comparison of different VN5MR styli: https://www.audiosciencereview.com/forum/index.php?threads/mm-vs-mi-vs-mc.18636/page-63#post-1383126

The differences there are all in the range you show. Perhaps we can continue this discussion there.

And this is what a super worn stylus look like on the graph. It's not so much FR that changes as it is distortion. Of course, above 10k could be affected but it is too difficult to say unless there are before and after measurements of the same cartridge.

View attachment 261897 View attachment 261898


To help move things back to the topic at hand, I do think the script can help us "see" wear in this way. I have measured all sorts of cheap cartridges and I rarely see distortion above -18dB if ever.

the plot jp ran is the str100. the eq comparisons showing 10-20k were all the megadeth track holy wars. i always go back and forth between the original cd and uk pressing to quick and dirty check what a cart is sounding like. if a cart is neutral, the cd and vinyl match up pretty evenly when listening and flipping back and forth. i already had recordings of it with half the stylii, so i recorded the track again with the rest.

 

[morillon]

I think we can all agree - this whole stylus - soft vinyl groove thing is a disaster. I am so happy I have Roon, Chromecast, and a good DAC. (and I have a SP-10 with a V15-MR Jico, - which is never used.)

amazing...
but it is respectable of course....


the main thing is that you are happy to listen to music
and this....measures or no measures...the finality..remains in our head...
;-)

 

[morillon]

I'd have figured hardened rubber on an Audio Technica would result in a rising response. I always thought the suspension was primarily the wire and the donuts providing damping. I've purchased dozens of vintage Audio Technica styli though and have never experienced a suspension issue like that, though, it is one of the reasons I appreciate the vintage A-T stuff.

were you satisfied subjectively... or via objective measures...?
;-)

 

[JP]

i wish i could post some plots for you guys, but im on a windows 7 machine and between trying to download old versions of python and me having absolutely zero clue as to what im doing, its just not looking like its going to happen.

3.8 works on W7 and is what I was using until I upgraded a few weeks ago just because.

 

[levimax]

See below original vs Beta 13 with the same file. They look pretty consistent but what am I doing wrong with the over 20KHz plot for the Beta 13? How do I get greater resolution on the FR access like I see here. Thank you for the great work on this!

 

[Thomas_A]

Anyone with macOS instructions to install and run the script?

 

[USER]

See below original vs Beta 13 with the same file. They look pretty consistent but what am I doing wrong with the over 20KHz plot for the Beta 13? How do I get greater resolution on the FR access like I see here. Thank you for the great work on this!

View attachment 261997
View attachment 261998

STR-100 wasn't meant to go higher than 20k. What you are seeing is what is there. Perhaps it is salvageable like how below 500 was, but for now that's what we have.

 

[levimax]

STR-100 doesn't go higher than 20k.

What line(s) do I change to get rid of over 20 Khz? The lines I see are commented "not implemented". See below

normalize = 1000
startf = 20 # not implemented
endf = 20000 # not implemented

 

[USER]

What line(s) do I change to get rid of over 20 Khz? The lines I see are commented "not implemented". See below

normalize = 1000
startf = 20 # not implemented
endf = 20000 # not implemented

Try cutting the wav file off at the natural end point like before.

 

[levimax]

Try cutting the wav file off at the natural end point like before.

The .wav file is the same for both the old and new script so it is "like before". I don't recall doing anything at the "end" of the file.... only eliminate silence at the beginning. I will take a look but this is different behavior and I know one of the goals was to make the script less dependent of a perfect .wav file.

 

[levimax]

Try cutting the wav file off at the natural end point like before.

See below in Audacity.... it looks like I did cut off the silence at the end of the file not the beginning. Is there any instructions on preparing the .wav files? Last one I did was 6 months ago and I forget all the details but want to do some more. I know you are not "tech support" and don't want to bother you too much. Thank you.

 

[USER]

The .wav file is the same for both the old and new script so it is "like before". I don't recall doing anything at the "end" of the file.... only eliminate silence at the beginning. I will take a look but this is different behavior and I know one of the goals was to make the script less dependent of a perfect .wav file.

Spoiler

Now this is weird. V13 was working perfectly for me.

I then put in a wav file that was not cut at around 20k.

And now every wav file I throw at it spits out the same graph.

Even when I restart spyder

However I can still change the header info. I'll play around with it some more later. Try different recordings and see if you are stuck as well. If so, the issue is not you. It's a working draft so things like this will happen.

Also: make sure you plug in a stereo wave file onto V13+. You can even fake it by copying the left channel if that is all you have.


Edit: I deleted the wav files in my working folder and started fresh and everything was resolved. Regardless, make sure to cut off at the natural end-point for now. What is odd is that all the similar graphs above had the same time stamp so something did get stuck.

Edit 2: I now believe that the issue has to do with Spyder. It is resolvable, however, so not a big deal.

In order to clean up the thread, I tucked away my response.

 

[levimax]

Now this is weird. V13 was working perfectly for me.
View attachment 262010

I then put in a wav file that was not cut at around 20k.
View attachment 262011

And now every wav file I throw at it spits out the same graph.
View attachment 262012

View attachment 262013

Even when I restart spyder
View attachment 262015

However I can still change the header info. I'll play around with it some more later. Try different recordings and see if you are stuck as well. If so, the issue is not you. It's a working draft so things like this will happen.

Also: make sure you plug in a stereo wave file onto V13+. You can even fake it by copying the left channel if that is all you have.


Edit: the rest of my scripts appear to be broken now as well. Perhaps a library was modified.

TO ANYONE READING THIS, FOR NOW, MAKE SURE TO CUT YOUR WAV FILE AT 20K/ THE NATURAL END POINT.

I tried another file I had which seems to work fine in the new script. I will try to figure out the difference. I also had no problem "switching" to the new file and the 13 Beta seems to be working for me.