Phonograph Stylus Wear Experiment

[USER]

The aforementioned JICO 3% THD claim opened the door to the possibility for verifiable measurements. In many ways I now better appreciate, the benchmark is a rabbit hole, sorry @morillon, fraught with issues. In January 2016, I met Hans Weedon, who holds crucial patents for logarithmic analog-to-digital conversion, from when he was at Analogic Corporation. We spent 10 days together and have communicated since. He regaled me with many stories about these AD devices, the pros and cons, the flaws, and about audio in general. He knew of my passion for vinyl records and my hobby of converting the music from them to digital formats. Because he tested such things, he shared with me that he took new and pristine vinyl records and measured their frequency response on the first play and subsequent plays. He noted that after one single play, a vinyl record loses forever its highest frequencies – likely inaudible to humans. Subsequent plays further degrade records, but to an increasingly lower degree. AC had the necessary equipment to make these measurements. I though we could measure those trends with some work, especially after seeing the skills developed herein. That, perhaps naively, is why we captured 96/24 dubs of three sides of vinyl.

My read from @JP, @USER, and @morillon is that task is much, much easier said than done.

You are likely not going to get conclusive results, but that's OK. Why all the pressure??? This is an imperfect and dying medium anyways. Have some fun. Even if you get conclusive results the average (forum) vinyl user is screwing something else up anyways. The public has very little knowledge of the medium and all I read on forums are old wives tales from people that I would never ever take seriously. Even the info quoted I would never take seriously. That's why I'm here. (And here specifically, I don't even bother with non- measurement or technically based turntable discussion on ASR! Have you seen the silliness? Eww.) I neither expect nor want people to take my opinions seriously either but if I can offer simple, well-done measurements then I feel as though I am actually contributing something. I am the least technically-minded person here and I am merely taking advantage of the incredible work being done by folks like JP in order to do some very minor grunt work that is within extremely limited parameters. Nonetheless I am seeing surprising results (that often run counter to accepted knowledge) and this is why I encourage others to run the measurements I am conducting myself.

Get whatever evidence you can and proceed from there. I think that Ray's images are going to be helpful so that's a big win in my book. More, and I don't mean this lightly, JP's script is absolutely revolutionary. After all this time the general public finally has an opportunity to measure the most important variables that influence what we we hear and how we like it. It's a shame it is only possible now when the medium is on its last legs (a good test record that can meet our current measuring technology will never be made), nonetheless digitizing and measuring to some degree is now pretty much democratized. If you do this right, you will still be contributing something even if its imperfect (just do it right, lol).

 

[JP]

Don't forget Scott - if not for him it wouldn't exist.

 

[USER]

Don't forget Scott - if not for him it wouldn't exist.

Amen.

 

[BendBound]

You are likely not going to get conclusive results, but that's OK. Why all the pressure??? This is an imperfect and dying medium anyways. Have some fun. Even if you get conclusive results the average (forum) vinyl user is screwing something else up anyways. The public has very little knowledge of the medium and all I read on forums are old wives tales from people that I would never ever take seriously. Even the info quoted I would never take seriously. That's why I'm here. I don't expect people to take my opinions seriously either but if I can offer simple, well-done measurements then I feel as though I am actually contributing something. I am the least technically-minded person here and I am merely taking advantage of the incredible work being done by folks like JP in order to do some very minor grunt work that is within extremely limited parameters. Nonetheless I am seeing surprising results (that often run counter accepted knowledge) and this is why I encourage others to run the measurements I am conducting myself.

Get whatever evidence you can and proceed from there. I think that Ray's images are going to be extremely helpful so that's a big win in my book. More, and I don't mean this lightly, JP's script is absolutely revolutionary. After all this time the general public finally has an opportunity to measure the most important variables that influence what we we hear and how we like it. It's a shame it is only possible now when the medium is on its last legs (a good test record that can meet our current measuring technology will never be made), nonetheless digitizing and measuring to some degree is now pretty much universally possible and democratized. If you do this right, you will still be contributing something even if its imperfect (just do it right, lol).

Brilliant, I like your attitude and approach. Thank you for your encouragement. As we've said, input from ASR has changed how we will conduct the experiment in our attempt to get it right. If not in this run, then in the next. I am still working to digest if not simply to understand what @JP has recommended. We are not yet turning to the right, as it were, so to the extent we can, we'll continue to tweek our process. But the primary goal is to record progressive stylus wear on cleaned records. You have offered assistance, expect us to call upon that offer. Again, thank you and @JP for the programming, and Scott who was in there from the beginning.

PS: Hans Weedon did not understand why vinyl did not die a quick death. He saw it as an inherently antiquated technology. Given the issues of just proper cartridge set up, its easy to see how a mid-level CDP is a huge step up in sonics and convenience.

 

[ray_parkhurst]

To be clear my earlier input was to correct the misconception that harmonics couldn't be measured with sweeps and that my opinion is that THD as a percentage is a very broad brush that's not that insightful beyond that. I doubt the increased distortion with wear will be readily audible short of the stylus reaching a pathological state, and maybe not even then. FR is where we're sensitive, which is why I'd advocate for that but also why I'm advocating that FR for both outer and inner grooves is important. Most FR tracks are on the outer grooves, but it's the inner grooves where the wear will present first.

...

I'm a strong advocate of swept harmonics as long as the software is capable to resolve them. Being able to separate individual harmonics rather than THD surely gives a more informative result.

Your hypothesis regarding distortion from the worn stylus is noted, though even the mfrs (eg JICO) quote THD numbers as criteria for limits to wear, so you're going up against an industry. I've been there on occasion as well. We'll just have to see from the data we're able to collect and analyze.

 

[JP]

Your hypothesis regarding distortion from the worn stylus is noted, though even the mfrs (eg JICO) quote THD numbers as criteria for limits to wear, so you're going up against an industry.

By saying I don’t think the distortion will be readily audible I’m going against an industry?

 

[ray_parkhurst]

By saying I don’t think the distortion will be readily audible I’m going against an industry?

Yes, the stylus manufacturing industry, JICO in particular. They say that once the distortion rises above 3% the stylus should be replaced. By inference they are setting a lower limit of audibility of 3%.

It's debatable of course, and it seems you believe 3% is not audible, which I don't dispute, especially at 15kHz. They also don't provide measurement data or experimental results to back up their claim. But audible distortion is indeed a claim by others in the industry as to why you might want to replace your stylus. If you say it is not audible until a pathological condition, you may be right. Certainly folks who claim they hear no degradation after 2000 or 10,000 hours are in agreement. I'd personally like to see those styli. Only comparison data like we're about to make will tell us how much distortion is present vs wear, or at least how much of an increase in distortion. Audibility is another matter.

There are folks on ASR who believe diamonds can't be worn by vinyl grooves, and probably many other "interesting" hypotheses floating around even on this seemingly enlightened forum.

 

[JP]

Inference isn’t universal.

Who here has said diamonds don’t wear?

 

[morillon]

I don't understand anything you say.
of course the distortion is noticeable.. this is what tells us that a cartdrige is worn out and puts it in the drawer

the thd is extremely simple to observe and explicit.. at least in relative value... which interests us..
the use of vinyl distortimeters has been common since the end of the 60s.. and the same for the observation of imd...
that s the principe of shure or audiotechnica measuring benches from the 70s and their test discs..


when has the audible aspect of this 15k alone...
of course that's not the idea
it is just that it has been observed that if we estimate a used cartdrige. it is found that it is correlated to more or less 3% at 15k..
invites you to observe all the interest of the famous 16k-16k300 that I pointed above ... the interest of imd
..and really special ..
a ccif principle

but yes.. the main thing is the choice of test discs for given tests, and their condition, state of wear etc.

 

[ray_parkhurst]

Inference isn’t universal.

Who here has said diamonds don’t wear?

Check the 12th post (if I counted correctly) in this thread: https://www.audiosciencereview.com/forum/index.php?threads/macro-photos-of-phono-styli.44822/

He made similar statements in other threads.

 

[morillon]

can be the occasion to point out discs, types of music, very demanding demanding which point to possible worries of wear...
it could explain that the words of each other diverge... according to their practices and requirements
;-)

 

[USER]

For what ever it is worth and so you can be on the lookout for similar phenomena:

Here is a comparison the CBS record before and after about 100 measurements or so. Note that the difference is due to set-up as is discussed on the thread. Obviously, this is just the outer groove area of the record.

The only egregious frequency response alteration (record damage) I have seen is this:

Note that this is not due to a "worn" stylus damaging a record. This is the result of a stylus I damaged when I tried to photograph it. (An ADC K8 conical using a 3.5g tracking force in case that matters.) I had it on a small z-axis pedestal that I accidentally raised too much causing the stylus to smash into the microscope glass and crack it. Luckily it was the glass of a lens filter. Unfortunately I ruined a valuable test record. (Not the first time and, I'd bet, not the last.) I didn't notice as the first measurement came out seemingly normal. Consequent measurements did it in. I will photograph that stylus when I can through my limited means. Perhaps the CA test record is easily susceptible to wear and damage. I'll be sacrificing a CBS test record in order to compare once I am done with all my other measurements. I have loads of weird cartridge situations to investigate. Fun.

A more thorough comparison:

It turned the prestigious Shure V-15 V-MR into an Audio-Technica cartridge, lol! As long as tracking is still decent I think most would not know the record is f*cked. (I am using an extremely good cartridge so that certainly helps here.)

As far as I can tell, the well-worn styli I presented earlier on this thread did not cause any noticeable changes to that CBS test record I used to measure them. But I never really tested them thinking about the issue and may do so in the future.

 

[restorer-john]

Note that this is not due to a "worn" stylus damaging a record. This is the result of a stylus I damaged when I tried to photograph it. (An ADC K8 conical using a 3.5g tracking force in case that matters.) I had it on a small z-axis pedestal that I accidentally raised too much causing the stylus to smash into the microscope glass and crack it. Luckily it was the glass of a lens filter. Unfortunately I ruined a valuable test record. (Not the first time and, I'd bet, not the last.) I didn't notice as the first measurement came out seemingly normal. Consequent measurements did it in. I will photograph that stylus when I can through my limited means. Perhaps the CA test record is easily susceptible to wear and damage. I'll be sacrificing a CBS test record in order to compare once I am done with all my other measurements. I have loads of weird cartridge situations to investigate. Fun.

Confused here.

You raised the stage and crashed the stylus into the objective lens, which was somewhat protected by a lens filter? How do you know you didn't merely damage the cantilever or its overall geometry? What makes you think the diamond itself was in any way damaged?

 

[USER]

Confused here.

You raised the stage and crashed the stylus into the objective lens, which was somewhat protected by a lens filter? How do you know you didn't merely damage the cantilever or its overall geometry? What makes you think the diamond itself was in any way damaged?

My first thoughts were similar. I wasn't worried about the diamond as it went up against glass and so I went ahead and measured it. This is what I got.

The thing that struck me was that rise at 20dB so I focused on that. (See the script thread for discussion about it possibly showing cartridge-tonearm incompatibility as this is a low mass tonearm. I was very excited about this.) Everything else looked "OK" so I thought nothing of it. I figured that the super short cantilever of the K8, with its low dynamic compliance (it can take 4g+) was able to take it. As the record got damaged, I think it's fair to blame the stylus. This was a super cheap cartridge after all. Further, I am also basing this on before and after comparisons of different cartridges. I haven't touched this one since I realized what happened. I started getting strange results from cartridges I had previously measured and the differences are consistent with what is shown in the post above.

In addition, before the incident I had already established a reasonable consistency with respect to a comparison of that record with a CBS record:

https://www.audiosciencereview.com/...measurement-script.41148/page-11#post-1501312

That was thrown off.

Edit: I am also now thinking that this consequent test of how a servo tonearm would affect the resonance in the first measurement is actually showing the damage--at the very least in the higher frequencies. It lines up similarly and it would make sense. I'll have to redo the test with a NOS copy, which is fine as all of these were exploratory measurements as I made clear.

But now that I'm looking at the first measurement of the K8 again, perhaps there is some minor suspension damage there after all, which would make sense as you say. Thanks for getting me to take another look at it. I have another NOS copy and will measure it on a working test record to make sure.

This is a measurement of a cartridge with known suspension issues.

The dip is at around the same area. Half a dB dip is normal for these if loaded correctly (more, part of it is the CBS test record after 5kHz), but this is excessive, especially with the capacitance boost. And, yes, these are merely measurements. I'm sure listening to it would illuminate the issue further.

Just in case, here is one with a hardened damper that I was able to improve somewhat with rubber restorer, just so you know what *I* look for.

Another hardened damper (and extreme stylus wear). Don't buy old A-T cartridges, folks!!

 

[Icewater_7]

I'm a strong advocate of swept harmonics as long as the software is capable to resolve them. Being able to separate individual harmonics rather than THD surely gives a more informative result.

Your hypothesis regarding distortion from the worn stylus is noted, though even the mfrs (eg JICO) quote THD numbers as criteria for limits to wear, so you're going up against an industry. I've been there on occasion as well. We'll just have to see from the data we're able to collect and analyze.

Hi Ray, I’m an acquaintance of BenBound that has been following the stylus wear project on this and other forums. It appears that you now have a final definition of the process that will be followed and will execute on it. I’m impressed with the analysis and feedback you have gotten here from the members. Is it true that you will only be using tracks A1 and A2 of the CBS STR-100 for the digital file creation? Is it also true that BendBound’s 96k/24-bit digital recording of those bands will be sent to someone else for running “the scripts” which will generate plots of frequency response, 2nd and 3rd harmonics, and crosstalk as a function of frequency? I know your original goal was to measure THD on the DUT (cartridge stylus) in its virgin state and then again after many hours of playing wear and see what effect the wear has on the plotted results of post-wear analysis processing. Those bands do have a 1kHz spot tone at the end which could possibly be used to measure THD, but you seem to have decided to forgo that and wisely keep things simpler. There was some good advice regarding uncontrolled variables and you have done a very thorough job to think of many possible ones and apply guard bands in the test process to deal with them. I’m certain you will learn a lot from your first attempt that will improve future re-tests or point to different and better ones that will give you the data you seek. BendBound and I are experienced “needle-droppers” digitizing our big vinyl collections using post A2D software to remove the vinyl noise so we can enjoy listening to the “cleaned up” digital tracks without distractions. So I’m curious to know if the scripts perform any noise reduction as a part of its signal processing. Not a question that you can answer but maybe someone else can. Just wondering if a second file could be sent with BendBound’s vinyl noise reduction settings to see if an analysis of a file with less vinyl “N” might show up the effects of stylus wear with more accuracy. Pure curiosity on my part. Best wishes for satisfying outcomes.

 

[ray_parkhurst]

Hi Ray, I’m an acquaintance of BenBound that has been following the stylus wear project on this and other forums. It appears that you now have a final definition of the process that will be followed and will execute on it. I’m impressed with the analysis and feedback you have gotten here from the members. Is it true that you will only be using tracks A1 and A2 of the CBS STR-100 for the digital file creation? Is it also true that BendBound’s 96k/24-bit digital recording of those bands will be sent to someone else for running “the scripts” which will generate plots of frequency response, 2nd and 3rd harmonics, and crosstalk as a function of frequency? I know your original goal was to measure THD on the DUT (cartridge stylus) in its virgin state and then again after many hours of playing wear and see what effect the wear has on the plotted results of post-wear analysis processing. Those bands do have a 1kHz spot tone at the end which could possibly be used to measure THD, but you seem to have decided to forgo that and wisely keep things simpler. There was some good advice regarding uncontrolled variables and you have done a very thorough job to think of many possible ones and apply guard bands in the test process to deal with them. I’m certain you will learn a lot from your first attempt that will improve future re-tests or point to different and better ones that will give you the data you seek. BendBound and I are experienced “needle-droppers” digitizing our big vinyl collections using post A2D software to remove the vinyl noise so we can enjoy listening to the “cleaned up” digital tracks without distractions. So I’m curious to know if the scripts perform any noise reduction as a part of its signal processing. Not a question that you can answer but maybe someone else can. Just wondering if a second file could be sent with BendBound’s vinyl noise reduction settings to see if an analysis of a file with less vinyl “N” might show up the effects of stylus wear with more accuracy. Pure curiosity on my part. Best wishes for satisfying outcomes.

I think there are a few misunderstandings that need to be cleared up, and one change to the experiment that needs to be announced. First, @BendBound will only be recording the few records that he recorded at T0, and will record them after they are played. One will be played at the end of first test period, and one at the end of last test period, and one is being held as a control (I believe). We're looking to see if the stylus causes record damage when new and when worn. @BMRR(VE) will be doing the recordings of the STR100. He will record the entire record, both sides, and make the data available to the forums. Which tracks are used is up to the folks doing the analysis, but I expect both swept and spot harmonics are the primary focus.

I'm interested as well in the impact of the noise reduction algos on the other analyses. I'll let the relevant folks comment.

The one change we're making to the process is to add a reference stylus measurement at each test point. @BMRR(VE) will record the STR100 with both styli to give another level of calibration/reference to the system and eliminate any issues with potential wear on the record causing uncontrolled response shifts.

 

[Icewater_7]

Hi Ray, I apologize for my misunderstandings. Thanks for the details, I’m anxious to read the progress reports by the participants as they get going.

 

[restorer-john]

Another hardened damper (and extreme stylus wear). Don't buy old A-T cartridges, folks!!

So is the general consensus vintage AT dampers all go hard? That's my limited experience too.

Many other brands do the complete opposite and collapse or go super soft.

 

[USER]

So is the general consensus vintage AT dampers all go hard? That's my limited experience too.

Many other brands do the complete opposite and collapse or go super soft.

Those darn little ring dampers.

I wouldn't say consensus, just what I have noticed after measuring about 20 of them. All my more "modern" ones including AT7Vs, AT150MLXs, and studio reference series (8008, etc.) are good but I've had bad luck importing a NOS AT15Ea and NOS AT120Ea as well as with the cheaper stuff from the 80s such as the AT110E and AT71E. The 3 Signets I have tried, also from the 80s, are good, however. But this point I'm going to stay away from any AT cartridge 30 years or older, which is a shame as I want to measure more classic ones.

 

[JP]

I may need to try the 180 then