For reasons already set out, your 0dB has no meaning without a reference level which we do not know, Hans. In one plot, IIRC, you manually inserted a reference level overlay, deleted other aspects which would be expected as signatures of vinyl playback such as harmonic distortion spectra. Whilst I'm sure of your good intention, there is no way to verify the absolute levels of your plots. In fact, most of them lack characteristic features one might expect from normal vinyl playback, and are, as such, abnormal.
The absolute level of the noise floor is irrelevent, rather the contrast between ultra-low friction and normal playback. But, in fact, you did not measure friction coefficient, Hans. And your plots, from what we can tell, lack the characteristic attenuation of cart-arm resonant peak associated with reduction in friction coefficient. So there is no evidence your tests involved reduced friction at all, we simply do not know....... if you PM a link or post the raw recorded files, I'm more than happy to analyse them for the lf signature of reduced friction, I predict that would reveal the truth of the matter?
In fact you told us you had not repeated your test, you just repeated analysis - whereas I have repeated my tests with similar results across test discs many times over the years, to the point where there can be no reasonable doubt, including on 3rd party recorded test files which support my results, as already presented on this thread.
Furthermore, each repeat test consumes a test disc. It is typically ruined afterwards by permanent deterioration of surface noise for dry playback, a matter which you have either not tested or are silent on, Hans.
Great, that should yield characteristic looking results, similar to mine, so why not ? Again, if you share the raw recordings that should settle the matter, Hans. The plots just don't look like normal vinyl analysis, IME, wheras they should with that method - I suspect some mistake !
There is no dispute here. For example in the case where dry playback already had friction low enough not to contribute to surface noise, as already suggested may be the case in your test, Hans. However, you did not measure friction coefficients wet/dry, Hans, which is a pity, a missed opportunity, and an oversight in the circumstances.
OK, so did you record the sample in that state, or measure friction coefficient ? Then again after cleaning ? The act of cleaning per se can alter friction coefficient, depending on method and agents Pardee JAES Dec 1981. Sometimes better, sometimes worse, sometimes no change.
Yup. And be methodical before even thinking your work might be valid 😉
LD
Last edited:
It typically is though, but not always IME. What's more, 'normal' playback can be restored by pretty much any agent that reduces friction, not just water, IME. No reason that shouldn't include cleaning I suppose, with the right agent and regime, as per Pardee's '81 JAES article.
LD
In thinking what I need to consider with this, just revisiting noise sources (for a fairly wide definition of noise) and so far have:
1. Noise cut into the groove from master tape, cutting amplifiers etc. This is correlated, so each play gives the same noise.
2. Noise due the vinyl. Also correlated, with good vinyl should be much lower than 1.
3. Noise due to static charge. Uncorrelated
4. Noise due to dirt. Uncorrelated
5. Noise due to mistracking. Uncorrelated
6. Noise due to friction. Uncorrelated.
7. Noise due to the cartridge/preamp. Uncorrelated.
Have I missed any?
Still trying to work out the easiest way to make a friction measurement setup. First thoughts are that a garrard 301 style idler setup would be the easiest so that you could get the platter to speed then remove drive. I'm sure I have all the bits needed somewhere...
1. Noise cut into the groove from master tape, cutting amplifiers etc. This is correlated, so each play gives the same noise.
2. Noise due the vinyl. Also correlated, with good vinyl should be much lower than 1.
3. Noise due to static charge. Uncorrelated
4. Noise due to dirt. Uncorrelated
5. Noise due to mistracking. Uncorrelated
6. Noise due to friction. Uncorrelated.
7. Noise due to the cartridge/preamp. Uncorrelated.
Have I missed any?
Still trying to work out the easiest way to make a friction measurement setup. First thoughts are that a garrard 301 style idler setup would be the easiest so that you could get the platter to speed then remove drive. I'm sure I have all the bits needed somewhere...
It certainly fits the category of "not such a clear problem when you dig into it". Great discussion, enjoy it thoroughly.
The part that has most fascinated me is how much the stylus can change the noise spectrum. If you have a large vinyl collection then one stylus may not be enough. Which just makes it more confusing why the phools still stick by MCs*. It also might explain some of the arguments people have over diamond profiles.
* I was that phool, but I am getting better.
* I was that phool, but I am getting better.
On the manufacturing side why would there not be a standard profile for the cutter head of the lathe?
🙂 same boat. But I have the 'it just works' turntable. This is for the fun part (when I get odd minutes for that).
Has anybody ever inspected the grooves before/after wet playing by a microscope? I found here some interesting pictures:
Incredible photos of record grooves under an electron microscope
Google gives many others, Micrographia, Vinyl Asylum, etc.
Last edited:
Blimey, what a rogues gallery ! Amazing it works so well in practice, which it does of course. There's also spindle bearing rumble, motor whining/cogging, platter microphony, ground hum, off the top of my head.......
There's a nuisance ranking to all of that lot. For me anything that has an audible single tone like sound, such as hum, rumble, motor noise etc is top of the eliminate list. After that, it's crackle-pop-tick, so static/dirt/friction. I don't mind white/pink noise if its quiet enough. Part of analog heritage !
Here's a repost from some idiot a few years back on another forum 😉
This is similar to Robert Pardee's 1981 method, I subsequently discovered. Using this method, I found that friction coefficient varies greatly between different styli, especially those of different profile and 'quality', not always as you might think. Also varies critically with alignment, especially for line contact types, I found.
PS: Moment of inertia of the platter calculation above is good for solid uniform platters only. Otherwise, one has to calculate I depending on the distribution of mass within the platter. I'm a physicist, all my stuff is only good for spherical cows in a vacuum ;-0 But calculating/estimating I is not too hard with a bit of reading up for the rusty/uninitiated, and is good for the soul......
LD
Last edited:
There we go again.
Posting #50, here displayed up to 2Khz, was the Original Recording for 1) the level of a 0dB test tone, 2) the noise level of a silent groove and 3) the noise of the playback chain. All 3 recordings played at the same gain level of the recording chain.
So here is the reference of 0dB against noise.
Posting #33 shows exactly this noise level, now transferred to excel together with the level of the 0dB 1Khz, the whole set now normalized to 0dB for the 1Khz test tone on the vertical scale.
From there I calculated the other lines in the figure and the dynamic range in dBA ref 0dB 1Khz.
Apart from the above, I measured the amplitude of the mono 1Khz 0dB test tone at ca. 36 um which turned out to be exactly the value it should have for a 0dB mono test tone.
So dr. Feickert did his homework correctly.
Here cannot be any possible dispute about references and absolute levels, period.
That is exactly what you see in again the Original Recordings in posting #122, only the relative differences are important, the absolute level here is not normalised and has no meaning.
The original recordings are in posting #122, I even showed the plot of the resonance of my arm / cart combination.
I think you will have to be very clever to extract a "signature of reduced friction" from this.
Just compare posting #50 from 2 years ago with the recording in posting #122, from last week and you will notice they are the same.
And to conclude, I'm not interested in friction coefficients, at this moment they just have no meaning to me.
I just wanted to check whether wet playing reduces the noise, which is clearly not always the case.
You should turn your mind 180 degrees and ask yourself "assume that all test are performed correctly, what might be the reason for the unexpected outcome" instead of trying to finding errors in the way I did the testing.
You know what test disk I used, but for whatever reason, you never told so far which test disk you used ??
Hans
Understand on the coast down. Problem is I don't have a turntable I can do that with. So need to make something out of the bits I have. When doing this some form of rim drive has it's charms.
But we do not know the absolute level of your 0dB test tone, Hans, because you either do not know it or haven't told us. Decibels are a ratio, and to be meaningful as an absolute measure, there must be a reference. There is no definitive meaning of '0dB', for the reasons already set out this changed over the years between various standards bodies. You have to know, and tell us, otherwise we have to guess.
But not at the same time, nor at the same radius on the record, and none of them with measurement of friction for correlation.
Strictly, you mean a tone of unknown level against noise measured at a different time on a different part of the record........
Trust you can see the scope for mis-assumptions here, as above.
Strictly, you measured 36 um at the second attempt, Hans, and stopped. Amazing. And we still do not know what the reference level is, so strictly whether 36um is 'correct', nor what you might have measured if you checked for a 3rd time.
If you post or PM a link to your raw files for wet/dry of the silent groove, in any lossless format, I can analyse the cart-arm resonant spectra and reveal enough of a signature to comment on the likelihood that the friction coefficient was altered significantly between your wet/dry tests.
This is because reduction in friction results in reduction in skate force, and dynamic variation in skate force is a significant stimulus for the cart-arm resonant system.
If you aren't going to provide them, just say and I'll stop asking.
You totally miss the point, Hans. The correlation between friction coefficient and surface noise is the point of interest being tested. Unless you are successful in altering friction coefficient, you will not alter surface noise via wet playback IME.
Over several years I probably went through dozens of test records of various types and from various eras. I obtained several batches of factory test records, some of which have a whole side recorded with tones for IMD tests. I believe my original post was based on one of those, for which the absolute IMD tone levels are irrelevant for the purpose at hand.
Having characterised wet playback and its correlation with friction coefficient, and the audible and measurable effects on playback, I reached the point where there was no point in ruining more test records.
As Billshurv posts, it's mostly for curiosity and interest, and, for me, understanding what other factors might influence friction, and so surface noise.
LD
Last edited:
Unfortunately SEM photos are destructive of the LP since only a piece is used and it has to be coated with a conductive layer. This is cool - https://www.youtube.com/watch?v=GuCdsyCWmt8
Some interesting comments I have read are that Dynagroove records were intended for a conical stylus, and, that changing from a stylus with a older profile to a modern type would show up in replay, the extent of wear in the grooves of a recording caused by previous incumbents.
For information on recent Ortofon and an old Nagaoka cartridges - frequency response graphs, tracking information etc and reviews - see
https://system.netsuite.com/core/me...ocodeaction=overwrite&addrcountry=US&gc=clear
MC cartridges are very expensive but Ortofon are still into that game.
Maybe the Nagaoka is acceptable for use with worn discs "as is".
With modern SEMs, you don't need to vapor coat or sputter with gold any more. Our company has tightened up the use of our SEM, but if I can figure out a way to sneak one past, I'll try to get some photos. This is almost certainly a mechanical effect, not chemical, but my knowledge of fluid mechanics is at the undergrad level so I wouldn't speculate as to what it is. Maybe some good SEM data will point us to the right answer.
Yes, it is cool. Stop frame animation though, wouldn't real time high frame rate be great ?! SEMs have fantastic depth of focus that optical can't touch !
LD
I'll have to think about it, but I'm wondering if I'd be able to see anything with some form of interference microscopy to elucidate anything about surface state (I'm not equipped in my lab with the right filters, but have a stage big enough that I could image select parts of a record without damaging the record), even if the spot size is large and it becomes more of an ensemble measurement.
I know people are using lasers as read-heads for vinyl, so the surface must be reflective enough to garner something.