Phonograph Stylus Wear Experiment

[krabapple]

Good suggestion! Need to know typical measurement variance to be sure of wear variance accuracy.

It's crucial otherwise you won't know if (or how much of) your effect is due to wear.

 

[USER]

What plugin do you use for fully accurate digital RIAA? I experimented beck in the 90’s with the Waves RIAA EQ plugin preset but was not happy with the results. Just went with a fully balanced RIAA solid state pre made by Audio Research (SP2, i think).

Yeah, most on the web aren't good. I use these developed by Scott Wurcer. They only work for 96k recordings in Audacity as nyquist commands. It is what is used in the script.

 

[Balle Clorin]

I an happy to inform that Parks Audio Puffin have digital RIAA and other settings that suit the test records . Using setting E500R-0 makes JVC and ClearAudio TRS- 1007 test records appear as linear plots when using the Python Frequency Script posted elsewhere on ASR.

 

[USER]

It's crucial otherwise you won't know if (or how much of) your effect is due to wear.

Perhaps we are discussing different things.

If you are discussing testing the records multiple times at each step because you think that there could be enough variance to muddy results then no, we have not found evidence of that beyond the reasonable margin of error built into test records (given proper set-up) and I recommend you read over 2 years of experiments with test records and the script with examples on the script threads. You can do it, but it would be a waste of time.

If you are discussing establishing a baselines for the test records and then measuring at each step, then that is exactly what is happening and I am the person running the measurements through the script. My hope is that they wait for me to give them the go ahead to start once they provide me the initial recordings. Ideally this will also happen at each step as there could be set-up issues when they load the cartridge again. The problem will not be the test record--beyond its regular limitations--nor the very small regular measurement variance.

Edit: they are also, smartly, measuring and comparing multiple CBS records (with the cartridge to be used and with a control cartridge) so that they can distinguish between wear on the stylus and wear on the test record towards the end of the experiment.

 

[ray_parkhurst]

The experiment uses a sort of natural process to ensure we are able to interpret the results properly. We'll measure the test record at T0, then again at T24 and T48. We don't expect to see much of a change in performance up to this point. Folks will have some time to compare the T0 and T24 data before we measure T48, then time to compare T0 vs T24 vs T48 before we measure T96, etc. If the consistency and trends look odd, we can adjust the method accordingly.

 

[Icewater_7]

No real need for several reasons. First, there are plenty of examples on the measurement script thread and its not so bad once you establish how a record performs. And most importantly, there is no real benefit because of how the experiment is set up. As they will be removing the cartridge to image it, the concern should be on making sure it is comparably set-up after each step. That's the concern that was mentioned here and why the experiment is only a first step. To properly measure it over the long run, you will need to minimize variables, which also means leaving the cartridge alone. Still, the first run should show interesting results.

I wasn’t clear. my bad. By “each” I meant at the beginning of one setup and at the end of wear plays for that setup. Not each separate cartridge.

 

[ray_parkhurst]

No real need for several reasons. First, there are plenty of examples on the measurement script thread and its not so bad once you establish how a record performs. And most importantly, there is no real benefit because of how the experiment is set up. As they will be removing the cartridge to image it, the concern should be on making sure it is comparably set-up after each step. That's the concern that was mentioned here and why the experiment is only a first step. To properly measure it over the long run, you will need to minimize variables, which also means leaving the cartridge alone. Still, the first run should show interesting results.

Note that only the stylus will be removed, not the whole cartridge. This should minimize any setup shifts.

 

[USER]

Realistic (Radio Shack/Shure) RXT6 (R16X stylus)

Worn stylus compared to NOS stylus

The test record has a bit of warping so I have to flatten it. Not 100% whether right or left channel high frequency resonance is correct, but the effects of wear on distortion are pretty clear nonetheless.

 

[GXAlan]

How many hours were put on that set of graphs? It seems like one reasonable alternative is to have a cartridge that starts off a bit hot and then ends up a bit rolled off if you’re not so focused on accuracy but want convenience. The 5 kHz region actually seems a bit flatter too.

 

[USER]

I'm fairly experienced at loading cartridges now so it wasn't so bad, and it certainly wasn't hours, maybe 40 minutes to get decent results and run the script. I was trying to test that particular record to see how off it was as well as the NOS stylus I just got in the mail and I was able to add a quick comparison to what I was already doing. I don't really understand what you mean at the end, however. And I don't understand why I would want inaccurate measurements.

 

[Balle Clorin]

@GXAlan , Hours? It takes me 10 minutes from I finish brushing my teeth, until I have the chart ready for publishing. That is getting the laptop, connect it to my Puffin, get the record out, dust off the record, record the thrack in CoolEdit pro, edit the file, Open the Python script an run the edited files. Bingo , finished. Simple when you are used to it. But optimizing a cartridge can take anything from 1 hour, to 4 hours or even 2 days. Fixed headshell on SME V is cumbersome…not really made for cartridge swapping..

 

[GXAlan]

@GXAlan , Hours? It takes me 10 minutes from I finish brushing my teeth, until I have the chart ready for publishing. That is getting the laptop, connect it to my Puffin, get the record out, dust off the record, record the thrack in CoolEdit pro, edit the file, Open the Python script an run the edited files. Bingo , finished. Simple when you are used to it. But optimizing a cartridge can take anything from 1 hour, to 4 hours or even 2 days.

What I was asking is how many hours separates the new vs worn stylus

Is worn 50 hours? 100 hours?

 

[Balle Clorin]

Opppsss. My wife said I never listen properly, now I cannot read good either,….

 

[GXAlan]

Opppsss. My wife said I never listen properly, now I cannot read good either,….

No worries. I was using a colloquialism which may be not be universally interpreted.

 

[ray_parkhurst]

What I was asking is how many hours separates the new vs worn stylus

Is worn 50 hours? 100 hours?

Yes, would be nice to know.

 

[morillon]

This is where there is a profound error in reasoning....


it is subjectively at what time (and what circumstances) I consider my diamond worn and that it bothers me..
and observes in different cases if corresponds to observation constants of wear of the diamond visually, or
by observing objective measurements of distortions etc.


but the first reflection concerns... the subjective...

 

[USER]

Yes, would be nice to know.

No clue. The cartridge came with a second hand turntable and was used
.

Hopefully we will be better able to tell after your project.

 

[ray_parkhurst]

This is where there is a profound error in reasoning....


it is subjectively at what time (and what circumstances) I consider my diamond worn and that it bothers me..
and observes in different cases if corresponds to observation constants of wear of the diamond visually, or
by observing objective measurements of distortions etc.


but the first reflection concerns... the subjective...

Since we have no information on how these two styli "sound" subjectively, all we can do is analyze the objective performance vs wear hours.

For sure I agree that when a stylus starts to sound bad, it should be replaced. I've been encouraging folks on my Imaging thread to send me their styli when they replace them due to sonic degradation, and most of those have shown heavy wear.

 

[Digby]

Can anyone give a quick summary of results so far? Apologies if these questions are answered, but a few quick things I noticed:

How are the hours elapsed counted?
Are different condition records being used for testing? Doesn't anything over than new records (granted, few play only new or NM records) introduce the variable of 'less good' records creating more wear than less?
Have you considered doing a digital recording of a piece of music as stylus wear progresses (perhaps inner track on an LP, to emphasize IGD changes?), to show how a single piece of music varies in sound quality compared to wear quality and what kind of correlation there is between the two.

 

[BendBound]

@Digby, I'm glad you asked. We got off on a huge boiler full of steam, only to be derailed post haste. That is to say, we or more accurately @BMRR (Vinyl Engine, SHF) who is running the actual turntable, got it all set up, AT VMN95E cartridge aligned, climate controlled and ready to turn to the right. Then he discovered the automatic return system on the Sanyo Q50 began to not work properly. So we delayed the start until we could sort this, we even considered using an alternative turntable. But after some searching and work, our pilot secured a new belt for the turntable for proper automatic return function. It's installed as seen in this link on Vinyl Engine by BMRR » 14 Jul 2023 11:32 (https://www.vinylengine.com/turntable_forum/viewtopic.php?f=19&t=134294&start=48), the Sanyo is operating properly so now we are up and running.

Today, @BMRR recorded the entirety – Side A and Side B – of the CBS ST-100 NOS test lp.

The experiment procedure in general begins as follows:

Records NOS CBS STR100 (T0).

Plays 1 vinyl record, both sides for ~12 hours each (T24), test stylus sent for photomacrographic imaging.

Imaged stylus (T24) returned along with a separate new AT VMN95E styli.
Re-records CBS STR100 with both styli.

Play record both sides ~12 hours each, ship test stylus for photomacrographic imaging. Returned for
Re-records CBS STR100 with both styli (T48).

Play 2 records both sides ~12 hours each, ship test stylus for photomacrographic iimaging and and return,
Re-records CBS STR100 with both styli (T96).

And so on until we are ~500 hours, possibly longer. Also see the initial post in this thread for our overall procedure.

See this link to the T0 recordings for both sides of the NOS CBS ST-100 test record:

https://drive.google.com/drive/folders/14le9fH-yCUWyEL3C0xBFaEJLriiqXwkH?usp=drive_link

USER has generously agreed to run the analytical program to sort out the progressive performance of the test stylus throughout this experiment. The performance will be referenced to the new stylus as well.