Analog Magik
- Thread starter Tirebiter
- Start date
I think you’re mixing things up here. It seems like you’re trying to measure the time difference between channels, but you’re focusing on overhang. The time difference between channels is more closely related to azimuth than overhang. If you perfectly set the overhang on an SME V, the maximum error would be around 1.8 degrees. However, the measurements you’re discussing are more related to phase and azimuth. Of course, all adjustments are interrelated, and remember that high distortion can be caused by many factors beyond alignment.
I am not changing overhang or thinking about it, not intentionally. If the tracking angle error is X degrees should that not show up as time difference left-right channel since the cartridge sides are not 90degrees to the groove direction, and touching “different” spots , one channel before the other?
I have difficulty visualizing how different azimuth makes one channel movement occur before the other ,,,
PS The SME V mounting holes allows some rotation of the cartridge to change offset angle slightly.
I have difficulty visualizing how different azimuth makes one channel movement occur before the other ,,,
PS The SME V mounting holes allows some rotation of the cartridge to change offset angle slightly.
Last edited:
Regarding the IEC and Baerwald alignment, on which the SME V tonearm is based, tracking error reaches its maximum of approximately 1.8 degrees at the outer and inner grooves, as well as between the null points. Elsewhere, the tracking error is below 1.8 degrees, and it reaches zero at the null points. You aim zeroing tracking error by aligning the zenith at the outer null point, approximately at a 120mm radius. However, even with precise overhang and zenith alignment, azimuth remains a critical factor. This is because stereo records employ the 45/45 Westrex cutting system, where both channels are encoded together and read in relation to each other. Any tilt in the coils alters this relationship, introducing a phase shift. A phase shift results in a timing difference between channels, affecting playback and, if present in the cutting stage, potentially compounding phase error.
Even with a mono record, using a stereo cartridge like the AT OC9 inherently introduces phase error if azimuth is not correctly set. Azimuth alignment impacts not only channel separation but also phase accuracy. This insight also addresses your previous questions on azimuth topics: channel separation is just one aspect of azimuth alignment, with phase alignment being equally essential. BTW apart from calculations you made, this response is about measurements and matching calculations with measurements which I assume you're after.
Last edited:
Many thanks, then I wonder what method should be used to correctly set zenith error to 0 at the null point?
I can’t say with certainty which method is best, but in my opinion, AM is a solid choice for aligning zenith since it uses a dynamic approach with test records. Alternatively, sending the cartridge to Wally Tools for inspection and alignment with their proprietary zenith protractor offers a static alternative.
And Analog Magic uses a 2 tone lateral modulated track and calculates Intermodulation distortion?
It’s not disclosed how AM does the calculation. At least I’m not aware of any.
Last edited:
I have some trouble following your calculation steps, but 6degrees off zenith is a much bigger phase difference for a 10kHz tone.
You are on the right track, but something is wrong.
I suggest to use the wavelength of 10kHz as a measure to estimate the zenith angle.
Last edited:
I had a typing error in my calc. The circumference at 12cm radius is 2x3.14x0.12=0.75m and 1.8 ( not 118) seconds per revolution gives 0.419 m/s track velocity. . 10khz has then a wavelength of 41.9 um. Not sure how to proceed from here. The phase difference is then ..
Position difference 1.9um gives 1.9/41.9x360= 16 degrees phase difference at 10 kHz
Position difference 1.9um gives 1.9/41.9x360= 16 degrees phase difference at 10 kHz
The groove velocity at track A5 is correct v = 2πR = 2π x 120 x 33.33/60 = 419mm/s
The wavelength of 10kHz at that velocity is also correct λ = v/f = 419/10000 = 42µm
Measure the phase difference Δt [µs] on your OS.
The required zenith movement then is:
a = v x Δt x b / 2w,
where a is how much the cartridge should be moved at the screws, b is the spacing between the screws (1/2") and w is the width of the contact area (say 20µm). See below illustration.
If L channel is ahead of R channel on the OS, a should be in the clockwise direction.
The above is just an illustration. Here I assume rigid body dynamics, but in reality it is certainly not. Actual w depends on the contact area of the actual stylus shape, vinyl elasticity, vtf, cantilever stiffness etc.
But it can give an idea of how much the cartridge needs to be rotated.
For info:
The phase difference in degrees is θ = 2π x f x Δt x 180 / π = 3.6 x Δt with Δt in [µs]
The phase difference in length then is Δλ = θ x λ / 360 [deg]
Last edited:
For your interest: the 1981 pressing of the NAB test record has about 6 degrees cutting zenith error. I say "about" because the error variable since approximately 3.8 degrees of it is true cutting stylus zenith error and the remainder is a lathe offset of over 3mm that happens to err in the same direction as the cutting zenith error, so the errors get stacked. The lathe offset makes for variable zenith error in the record depending upon the playing radius. The 1989 NAB pressing is much better.
We see stuff like this all of the time on test records in our research. I know of no test record on the market now or historically that has demonstrably accurate geometric alignment and cutting stylus symmetry, much less one that offers their measurement error on those parameters that matter on playback accuracy.
Do you know the accuracy of the 2006 issue of Analogue Productions AAPT 1, i.e. the UATLP ?
Is it ultimate or not so much?
We haven’t measured that one yet. There are many we haven’t measured that we own. Our work isn’t focused on measuring and comparing the field of test records but instead to refine our ability to measure and understand where the opportunity is with respect to improving cutting and playback alignment as well as cutting and playback forces.
The sloppiness is shocking. Happily for us audiophiles it is equally distributed about the mean so we can still use statistics to our advantage to get the most out of all records without worrying about tail events. It’s aligning for the tail events we need to worry about!
J.R., maybe if you're worried about 1,000 cartridges, where things like 'mean' begin to work to one's advantage, but when worrying about one cart on one table, the variation amongst test records IMO is just appalling. Here's yet another sample of results across multiple records using the same cart on the same arm on the same table.
These measurements were all from Adjust Plus+, and spot checked using some of these same disks on AM v1, where the difference was relatively small for the crosstalk numbers (but the AM v1 phase numbers were all over the map compared to what is shown here). I then used the AM v2 record to see what the crosstalk numbers would be there, and they were way out in left field compared to even this relatively sloppy span. Like L-R -18dB, R-L -30dB.
"Sloppy" and "shocking" are used appropriately here, I think.
Yup, no surprises there.
There are many things that can contribute to this variation between records when all else is held constant on the playback side: Cutting stylus symmetry, lathe offset, cutting stylus zenith error, cutterhead design, cutterhead chuck perpendicularity error, playing radius, bandpass filter effectiveness, noise floor of the playback amplification circuitry relative to the modulation velocity of the test track, etc.
The question isn't "which is correct"? The question is, "Since we know under what mechanical conditions the optimal test track would be cut, why have there been no test records EVER to proclaim their engineer's ability to hit those targets and offer their measurement error in doing so?"
In other words, we've been asked to trust the test records, but not been given any reason whatsoever why we should believe they got it right. I know from working closely with a number of mastering engineers that they simply don't have the tools to KNOW how well they are set up on all the parameters that matter.
I think the real question for relevance to what a TT lover needs to do is - do we just take a scope, attempt to determine the stylus offset angle from perpendicularity, and then set azimuth to compensate for that? If not, then using a test record is just a walk through a circus fun house, with the crazy mirrors that make you look fat, then skinny, then tall, then short. The result just randomly depends on what mirror you're looking at.
If you can be patient a bit longer, I'll have the answer.I think the real question for relevance to what a TT lover needs to do is - do we just take a scope, attempt to determine the stylus offset angle from perpendicularity, and then set azimuth to compensate for that? If not, then using a test record is just a walk through a circus fun house, with the crazy mirrors that make you look fat, then skinny, then tall, then short. The result just randomly depends on what mirror you're looking at.
I’m responding to my own post above. If one looks at the records and variations, one might say “Hmm, the Ultimate Analog records look pretty consistent” - especially since these two records were purchased 8 years apart! But then looking at the rest (not counting the A+ record which over time has been criticized for their crosstalk tracks), they all show a larger negative L->R value than R->L. So… is UA right(er) and all these others wrong(er)? How likely is it that all these other records - made decades apart from each other - just happen to show values larger in L->R and every single one is wrong in the same way relative to the UA records???
If one looks at the records and variations, one might say “Hmm, the Ultimate Analog records look pretty consistent” - especially since these two records were purchased 8 years apart! But then looking at the rest (not counting the A+ record which over time has been criticized for their crosstalk tracks), they all show a larger negative L->R value than R->L. So… is UA right(er) and all these others wrong(er)? How likely is it that all these other records - made decades apart from each other - just happen to show values larger in L->R and every single one is wrong in the same way relative to the UA records???
