Interesting article on the Korf site

[tabarddn51]

Hi, As posted on the L/H site, here is a very interesting article on arm alignment from the Korf (no association) site.

The Trap of Mental Experiments

korfaudio.com

Cheers

 

[Vlad Soare]

Very interesting. But what's the conclusion? I'm not sure how we can use this information.
Also, he doesn't say which alignment he used for the tests. It would be interesting to see how the results would look with all three alignments, comparatively. OK, they will be different from theory in all three cases, that's interesting to know, but what's even more important is how different will they be from each other. Because at the end of the day that's what we need to know - which one to use.

 

[wd40addict]

He implies it's Baerwald as he puts that as a reference on the graphs, but it isn't explicitly stated.

Another interesting experiment would have been to stick with one cartridge, say a VM95 and move up through the different stylii. He doesn't state what cartridges he's used although one is obviously Ortofon.

 

[jrtrent]

Very interesting. But what's the conclusion? I'm not sure how we can use this information.
Also, he doesn't say which alignment he used for the tests. It would be interesting to see how the results would look with all three alignments, comparatively.

Yes, that would have been interesting to see, as well as seeing what happens if the cartridge is just lined up straight with no offset angle. The end of the article says, "So, does this mean the calculations are completely useless and we should all just use straight tonearms from now on? I don't think so. The three authors got the tracking error calculation right; they just failed at understanding its implications. It might still make sense to optimize for less tracking error, just not because of harmonic distortion." It's a shame they stopped with "I don't think so" and "it might still make sense to optimize for less tracking error," when I'd think they could have just given it a try to find out. Maybe that will be in a future Korf blog.

One thing it reminded me of, though, was Shure's published data on 2nd harmonic distortion across tip types using the V15 Type IV cartridge. This was not across the whole playing surface, and a different frequency was used (8 kHz, @ 5 cm/sec peak velocity, CBS STR-100, bands 3A and 3B), but they got the same pattern respecting tip type as Korf did: .6 mil spherical 6.4%, .3 X .7 biradial 4%, and a hyperelliptical 2.5%. Of course, lower distortion does not always translate to more enjoyable sound, and Stereophile's review of that cartridge said, "all of our auditioners indicated a preference for the spherical, despite its theoretical inferiority."

 

[cgallery]

Would have been interesting to superimpose the distortion figures from the same stylus on a linear-tracking arm.

 

[EWarrior]

Very interesting. But what's the conclusion? I'm not sure how we can use this information.
Also, he doesn't say which alignment he used for the tests.

Their null points (59 and 112.5mm) do not match any commonly used geometry. And of course there's no way to tell if the cartridge has been accurately aligned. Even when doing a good job, the alignment tolerance may shift the null points quite a bit.

Their calculated distortion seems way too high to me. Judging by the total amount of distortion, I guess the peak recorded velocity may be quite a bit lower than the 100mm/s that everyone keeps using in their calculators in order to keep results comparable. Also, they didn't make up for the RIAA correction. With these factors considered, the calculated distortion is probably half of what they assume.

And why didn't they bother to include a graph of the calculated tracing distortion? It's well known that tracing distortion can be considerably higher than tracking distortion. It would have been nice to visualize that and it would put the results into perspective.

What's interesting is that the replicant performs much better than the elliptical in their test, even though they have the same minor radius...

 

[Vlad Soare]

Indeed.
Mind you, it's an interesting experiment that raises some valid and intriguing points, so I've no intention to bash them. Quite the contrary. Hats off to them for doing it. But it needs more work. It's not particularly useful at this stage.

 

[Thomas_AA]

Korf needs to compare using the same level. If theoretical distortion is calculated from 1 kHz at 10 cm/s he need to compare with a test record of 1 kHz 10 cm/s. All other comparisons are meaningless.

 

[theophile]

"So, does this mean the calculations are completely useless and we should all just use straight tonearms from now on? I don't think so"

Intelligent conclusion.

 

[JP]

Korf needs to compare using the same level. If theoretical distortion is calculated from 1 kHz at 10 cm/s he need to compare with a test record of 1 kHz 10 cm/s. All other comparisons are meaningless.

And not conflate 2H and THD, consider how HD sums, etc.

 

[tothmes]

I assume there will be several follow-ups on this from Korf; far too original and interesting results to just leave hanging. I doubt any variations of the levels would affect the overall shape of the curves, which are the main point of his measurements. After all, he isn't attempting to analyse the THD from different carts, but to test if the measured levels of distortion is in reasonable agreement with the theoretical expectations.

One aspect always missing in discussions of distortion from tracking errors is the audibility of the 1st and 2nd order distortion products that everybody since forever agree are the only product of those errors. As far as I know those parts of the distortion spectra would be inaudible, or at worst, possibly benign, even at the levels shown for the worst cart measured. Those levels are below the distortions produced by the mechanics in the inner ear, which the neural portions of hearing does already have built-in filtering for. (The same mechanism that explains much of the nastiness of high overtone distortion at much lower levels - the ear is not evolved to manage distortion which doesn't occur in natural / mechanical sources).

What would be interesting to see - apart from some of the methodological questions above - would be a spectral analysis from the same setups, both for harmonic and intermodulation distortion. Of course, an IM test would require a test record which nobody probably has asked for before.

 

[JP]

I doubt any variations of the levels would affect the overall shape of the curves,

Frequency, amplitude, linear velocity, and groove radii define the shape of the groove. Amplitude is undoubtedly critical.

 

[tothmes]

Frequency, amplitude, linear velocity, and groove radii define the shape of the groove. Amplitude is undoubtedly critical.

I still fail to see how adjustment of amplitude could modify three evenly(ishly) falling curves to the W expected from the maths? Please enlighten me. The only relevant results from this work are about those two basic shapes.

 

[Sleiven]

Here is one result I got with a 15 minute 3150hz W&F track,, hard to get repeatable and consistent results..
Here is another try

 

[Vlad Soare]

One aspect always missing in discussions of distortion from tracking errors is the audibility of the 1st and 2nd order distortion products that everybody since forever agree are the only product of those errors. As far as I know those parts of the distortion spectra would be inaudible, or at worst, possibly benign, even at the levels shown for the worst cart measured. Those levels are below the distortions produced by the mechanics in the inner ear, which the neural portions of hearing does already have built-in filtering for. (The same mechanism that explains much of the nastiness of high overtone distortion at much lower levels - the ear is not evolved to manage distortion which doesn't occur in natural / mechanical sources).

I did hear distortion very clearly when I had a slight alignment error (I had made a mistake with the protractor). Most records used to play just fine, which made me unaware of the error for quite some time, but two of them had an issue with a couple of louder, higher pitch passages. It went away completely after redoing the alignment.
So, I'm not sure that tracking errors are solely in the inaudible harmonics domain.

 

[EWarrior]

I still fail to see how adjustment of amplitude could modify three evenly(ishly) falling curves to the W expected from the maths? Please enlighten me. The only relevant results from this work are about those two basic shapes.

They were never supposed to get any W-shaped curve here as they can’t measure tracking distortion.
Tracing distortion plays a major factor here.

Getting the amplitudes right will just put things in perspective and will eliminate "unexpected" results like the THD being lower than the calculated tracking distortion.

 

[tothmes]

I did hear distortion very clearly when I had a slight alignment error (I had made a mistake with the protractor). Most records used to play just fine, which made me unaware of the error for quite some time, but two of them had an issue with a couple of louder, higher pitch passages. It went away completely after redoing the alignment.
So, I'm not sure that tracking errors are solely in the inaudible harmonics domain.

Nor am I; if the cart is badly enough aligned, of course even the lowest harmonics will become audible, especially on inner tracks. My point is rather that possibly even the worst of the levels shown in Korf's measurements may be below that level (at least before that violin solo in the final track).

Alex doesn't specify how his carts are aligned, but I would assume Stevenson, since that seems to be his alignment of choice. I would hope a maker of tonearms testing for distortion from tracking errors would know how to align his carts.

Mostly, what I'm trying to say is that much of vinylista arguments on tonearm geometry is based on a miscomprehension of the audibility of THD and distortion spectra, failing to weigh in the actual audibility of the different harmonic distortion products. A THD of even 4-5% - figures that once scared the heck out of me - is a totally different beast if it only consists of the first three harmonics, than it would be if it were the high harmonics distortion coming out of badly feedbacked transistors. This is mostly since our hearing is evolved to filter out a fair bit of those harmonics, since our inner ears do produce that same kind of distortion due to mechanical vibrations in their constituent parts.

Of course, if intermodulation enters the chat, all bets are off, likewise if the actual resulting distortion spectra aren't limited to the theoretical first three.

What we would need would be more of actual measurements like the ones Korf has shown, preferably with full spectra. After all he seems to be doing something that's so far has been missing in every previous publication on the matter, by adding a bit of measurements to the theory. The results he shows are intriguing; I would love to see others doing similar ones; if he's shown to be wrong, all the better since that would add to our understanding of the subject.

And in the next step, once we've established the actual measured THD coming out of our aligned carts, we should compare those results to figures from studies of the audibility of distortion spectra.

 

[mr_petit]

Getting the amplitudes right will just put things in perspective and will eliminate "unexpected" results like the THD being lower than the calculated tracking distortion.

I think Korf already makes a conceptual error in his explanation regarding what he says that his measurements are with RIAA when he says:
"Interestingly, 1 kHz is a zero decibel point on a RIAA curve—this frequency is neither boosted nor attenuated. In our case, this means that even the absolute THD level should be correct!"
Of course the harmonics are not at 1khz but a multiplier of them and I think they are very much attenuated by the RIAA filter. That could be one of the explanations of him producing low thd figures, especially with that replicant stylus. I think if he had focussed on or extended his article to focus on 1 specific harmonic, it would've been more interesting (and given a bodeplot of his riaa filter).
Still part of his findings, especially the differences in stylii is interesting, as one would expect any stylus shape should be able to track 1khz (provided that these are not of ridiculous amplitude, I don't know that record.

Would have been interesting to superimpose the distortion figures from the same stylus on a linear-tracking arm.

That was my first thought too. Why not do the same experiment on a linear tracker of decent quality. That would give a very interesting insight, regardless of the outcome (either the same or different)

 

[Thomas_AA]

Any test using a test record have many unknowns. The quality of the production/pressing of test record and the setup including stylus type. As already pointed out, stylus type affects tracing distortion which is lower with fine-line type of stylii.

Below is a 1 kHz mono track from the CA-TRS1007 record. It is recorded at 3.54 cm/s RMS and 5 cm/s peak. The track is placed ≈115 mm which is close to one null point. 2nd harmonic is here -62 dB or about 0.1%. I have the Moerch tonearm setup with nulls at 65.345 and 121.754 mm. Signal is also played through RIAA.

 

[JP]

Interestingly, 1 kHz is a zero decibel point on a RIAA curve—this frequency is neither boosted nor attenuated. In our case,

Also not exactly true in any real sense in terms of RIAA as bass EQ doesn’t flatten out until around 2.5kHz. Albeit barely a dB from 1kHz, but still a head-scratcher to state an arbitrary reference as an absolute.