advertisement


¿End-of-side distortion — inherent to vinyl?

Robert there's not 'only one applicable' to the Rega arm, you can choose between, Lofgren, Baewald and Stevenson as you see fit to bias towards lower overall distortion or minimum EOS distortion. They are different solutions that provide different results.
 
Nobody spotted my mistake: inner radius is typically 0.06m, not 0.1m, so phase distortion is about 1/30th of a wavelength at 20kHz, not 1/50th. Still a fairly minor effect. Perhaps a digital guru could tell us whether it's even resoluble by the CD specification.

CD has a time resolution of approx. 600 picoseconds, which is 0.0000000006 seconds. Even if we allow your initial estimate, 20,000 Hz and 1/50 of a wavelength covers a time period of 0.000001 seconds.

So, 0.000001 / 0.0000000006 shows that CD is 1666.67 times more accurate than is required to show up the ultimate limitations of vinyl, in this regard.
 
CD has a time resolution of approx. 600 picoseconds, which is 0.0000000006 seconds. Even if we allow your initial estimate, 20,000 Hz and 1/50 of a wavelength covers a time period of 0.000001 seconds.

So, 0.000001 / 0.0000000006 shows that CD is 1666.67 times more accurate than is required to show up the ultimate limitations of vinyl, in this regard.

I don't understand this. How can CD resolve picosecond scale data when the sample period is 23 microseconds? Equally, with a frequency brick wall at 22kHz, if it can resolve a slight phase shift at 20kHz then it can only do so by the skin of its teeth. My experience tells me that cymbals invariably sound very slightly tinny and suffocated on CD or streamed 44.1, and I can't discern a 16kHz sine wave, so I suspect that redbook artefacts extend well below 20kHz. Can anyone produce more convincing numbers for redbook 20kHz phase resolution?

Just to clarify my skepticism, it seems to me that you are claiming that CD can resolve tones in the GigaHertz range. Impressive technology.
 
It's in the detail of the Nyquist-Shannon sampling theorem.

Basically CD has a theoretical temporal resolution of 1/ (sampling frequency x no of samples)
Or 1/(44100 x 2^16) = 1/3.46x10^-10 or 289pS.
 
It's in the detail of the Nyquist-Shannon sampling theorem.

Basically CD has a theoretical temporal resolution of 1/ (sampling frequency x no of samples)
Or 1/(44100 x 2^16) = 1/3.46x10^-10 or 289pS.

OK - thanks - understand now. Large trends very accurately resolved.

Could the tinny cymbals be moiré?
 
Thanks for this, but I'm not yet convinced by either of these suggestions. I confirm that the variation for a 3 degree delta is about 13% (e.g. sin(20)/sin(23) ~= 0.87) but don't forget that this delta IS BETWEEN TRACK 1 AND TRACK 3, NOT TRACK 1 AND TRACK 5. You can verify this here:
http://www.sme.ltd.uk/content/Series-M212R-1536.shtml

If this is the cause of a need for variable anti-skating force, then the anti-skating force should increase as far as the middle of track three, peak there, and thence decline towards the last groove. I may be wrong, but that does not appear to be how the Caliburn Cobra is set up.
I'm sorry but I obviously did not make myself clear. I agree entirely with what you say. I gave those figures only so that we can dismiss this factor (offset variation) as a reason for increased bias at the inner grooves.

This is very counter-intuitive as well. The frequency does not change due to a slowing speed, so why should tighter wiggles matter. The stylus has correspondingly more time to navigate them, so the stylus doesn't really experience tighter wiggles. Think of fluid flowing through a more sharply bending pipe, but more slowly. The stylus sees the same snare drum frequencies it saw in the first groove and it navigates the same number of peaks of the same amplitude in the same time period. Assuming the energy transfer to the stylus is the same in each situation, why should drag increase? (BTW, fluid flow is a bad analogy because viscosity and turbulence mean that the pipe example is more acutely affected by speed, and there certainly would be less drag with a slower fluid in sharper bends. Turbulence within the fluid would wreak havoc with bend navigation.)
Again, I have to agree, but how else to explain the increase in bias (if it's real)?

FWIW I've always understood that bias compensation should be pretty much constant or increase slightly towards the centre. The example of an increase from 12% to 18% of VTF (a 50% increase!) seems far too great a change, no matter why we think it's needed in the first place.

This is turning out to be really interesting, isn't it? What do we know and why do we think it's correct? Unfortunately there's been so much BS from arm makers (including SME) over the years that this has always been a murky area.
 
Nobody spotted my mistake: inner radius is typically 0.06m, not 0.1m, so phase distortion is about 1/30th of a wavelength at 20kHz, not 1/50th. Still a fairly minor effect. Perhaps a digital guru could tell us whether it's even resoluble by the CD specification.
I did but ignored it because you were giving ball-park figures.
 
The only one applicable to a Rega arm, but most importantly a cartridge with a Micro Ridge profile.

Think of it this way (sorry if I'm being too simplistic).

The effective diameter of the record is reducing as the stylus moves across the disc surface. Assuming the same material is being recorded there is a 'bunching' and squashing of the undulations that form the analogue signal into a tighter space, making it harder for the stylus to trace them. In the orange/corner example above, imagine the corner as a more acute angle - you get an even worse fit.
You could get around this problem by driving the disc past the stylus at constant velocity (think how the speed of a CD varies as the laser moves across the disc).

Tips with a small minor radius cope far better with this effect and it dominates HF distortion, more so than arm geometry so long as this isn't badly off.

Best test (other than measuring which shows this all too clearly) is to play a vocal with strong sibilance at EOS. Horrid with many cartridges compared to the outer grooves, and of course digital. Doesn't bother some people but it drives me nuts.
I always found sibilance a real annoyance until I started spending a lot more on TT-arm-cartridge. Maybe it was just that the cartridge had a better stylus and the setup was more precise?

BTW I think you meant 90° earlier, not 45°. ;)
 
I don't understand this. How can CD resolve picosecond scale data when the sample period is 23 microseconds? Equally, with a frequency brick wall at 22kHz, if it can resolve a slight phase shift at 20kHz then it can only do so by the skin of its teeth. My experience tells me that cymbals invariably sound very slightly tinny and suffocated on CD or streamed 44.1, and I can't discern a 16kHz sine wave, so I suspect that redbook artefacts extend well below 20kHz. Can anyone produce more convincing numbers for redbook 20kHz phase resolution?

Just to clarify my skepticism, it seems to me that you are claiming that CD can resolve tones in the GigaHertz range. Impressive technology.
I agree. Where does the 600 picosecond number come from?
 
It's in the detail of the Nyquist-Shannon sampling theorem.

Basically CD has a theoretical temporal resolution of 1/ (sampling frequency x no of samples)
Or 1/(44100 x 2^16) = 1/3.46x10^-10 or 289pS.
Is this right? 2^16 samples, not 2^16 levels for each sample?
 
I guess we'll just have to agree to differ, but I'll leave you with a few recommendations that I hope you might consider.


http://www.amazon.co.uk/Uchida-plays-Schubert-Mitsuko/dp/B000654OUG/ref=sr_1_7?s=music&ie=UTF8&qid=1299672614&sr=1-7

Basil, I'm going to veer off topic to ask you about this Uchida Schubert boxed set. All I have of Schubert's piano music is the Jeno Jando cd of the Impromptus on Naxos and I've never really liked it; I find it a bit heavy handed and stolid really, but perhaps that's too harsh. I love Uchida's Mozart - do you think her refinement and subtlelty will convert me to Schubert's piano?

Sorry to interrupt the row; please continue.

Do they play records in hell and cd's in heaven, or the other way round? If they exist you can all carry on the debate FOR ETERNITY!!
 
This site contains affiliate links for which pink fish media may be compensated.
Basil, I'm going to veer off topic to ask you about this Uchida Schubert boxed set. All I have of Schubert's piano music is the Jeno Jando cd of the Impromptus on Naxos and I've never really liked it; I find it a bit heavy handed and stolid really, but perhaps that's too harsh. I love Uchida's Mozart - do you think her refinement and subtlelty will convert me to Schubert's piano?

Sorry to interrupt the row; please continue.

Do they play records in hell and cd's in heaven, or the other way round? If they exist you can all carry on the debate FOR ETERNITY!!

It's available as a download, so you could try a few movements for a couple of quid?

Some find her interpretations a little too 'over-emotional' if you're bothered by such things then it may not appeal?

I love them!

Brendel is well worth a try as well.

ps,

Superb recording, better than Brendel IMO.
 
Actually Basil I've already started listening to them on Spotify and got so quickly involved that I did a rash one-click order of the set. The emotional arcs she creates are extraordinary - wonderful stuff. And I can't wait to hear them in full resolution. I'm really loving the way the Harbeths reproduce good piano recordings.

Many thanks for the tip!
 
Actually Basil I've already started listening to them on Spotify and got so quickly involved that I did a rash one-click order of the set. The emotional arcs she creates are extraordinary - wonderful stuff. And I can't wait to hear them in full resolution. I'm really loving the way the Harbeths reproduce good piano recordings.

Many thanks for the tip!

For the health of my credit card, I have 'one-click order' disabled!

For a laugh, see if you can find Kissin playing 960 on spotify.
 
Gosh ! I think I'm mentally distorting after reading that. I don't dispute the physics, but bite-sized sentences may have eased my aged brain in grappling with it.

WHY is tracking force higher on the inner groove wall when one might think that outward pressure would prevail? I know you're right, but why?

Does the fact that the arm is pivoted outside the circumference of the record have any bearing on this?

If bias should increase with playing weight, why do Koetsu cart's much prefer little or no anti-skating when, say, a Lyra playing at .25 of a gramme lower performs better with a much higher bias?
I assure you you are not mentally distorting. What you stated, can be explained.

Back in the 60's when anti -skating bias on tone arms became popular, published illustrated tests were done in the U.S with sets of cartridges in various arms -with and without bias applied.

In order to reach the same clean test sine waves at various frequencies on display oscilloscopes, at a cartridge's given downward pressure WITH bias,....on a particular arm - it was repeatedly found.... WITH OUT the anti skating bias, in use - slightly extra tracking weight had to be applied to reach the same 'clean' standard,

In short: more extra downward pressure to deflect the inward skating tendency, and to achieve the same performance of equal pressure and contact - tracking, on both walls of a groove.

PS The mention here, previously by others: of 12" tone arms to 'overcome or lessen' the error problem... were initially designed for playing 16" radio broadcast transcription discs. The 12" tone arm was then adopted by some in the Hi Fi trade to keep tracking error at hitting no more than 1 degree rather than near 2 degrees.(that of 9" arms)

But in doing so, that automatically increases the bearing friction ...if considered against any available 9" arm 'identical equivalent' from the same maker and design. They continued on....due to being used by fans with rather heavy tracking moving coils at 2 grams and more - where 'bearing friction' is a complimentary part and parcel of using many 'low compliance' cartridges that certainly do not play around the 1 gram or slightly higher mark. .

If you increase bearing ( A.K.A sideway ) friction by certain amount ( rated in milligrams) - the possible lowest limits that a given cartridge will track in a given arm goes up.

One example of how it was tried, to be solved.: That old Garrard straight arm TT system -( was it called the Lab100?) the arm was a series of pulleys, pivots and strings down to the cartridge head. Bearing friction was 200 milligrams -which made it only possible to correctly track with a cartridge: by at least 2 grams downward pressure. Since the down pressure must be at least 10 times the arm's side- moving friction to not affect performance.
 
ter,

Yeah, 8-track was a phenomenally crappy format. Not sure what it was supposed to offer that cassette tape didn't already do much better.

Joe
 
The temporal resolution of cd is the number of samples per second, it has nothing to do with the number of levels per sample, it cannot define a time period smaller than that of the smallest sample duration. By it's very nature it's frequency response is limited to 1/2 the sample rate, as is its temporal accuracy.

It simply cannot play any sound for less than 1/44.100 of a second.
 


advertisement


Back
Top