Like what you see?
Subscribe to the Korf Blog!
- Be notified when new articles are posted
- Access members-only secret preview sales
- Get deals and discounts on Korf Audio products
Expect 1-4 emails a month. We will never ever spam you. You can always unsubscribe—no hard feelings!
The Korf Blog
The inside story: our research,
development and opinions
development and opinions
6 September 2018
Azimuth Adjustment Myths and Reality, Part III
There are many "azimuth setup tracks" on test LPs available today. Most provide a fixed frequency signal in one channel and then the other, allowing the user to measure crosstalk. Even with a very generous assumption that the LPs themselves are recorded to an acceptable standard, the best result one can get from it is case (b) as we described it in the first post of the series: generator coils would be in alignment, but not the stylus!
This isn't good. We need to align the stylus, not the coils. Fortunately, there is an easy way. You can use other, even poorly recorded, signals on commonly available test LPs to verify your azimuth setting. We at Korf Audio use the following procedure:
This isn't good. We need to align the stylus, not the coils. Fortunately, there is an easy way. You can use other, even poorly recorded, signals on commonly available test LPs to verify your azimuth setting. We at Korf Audio use the following procedure:
1
Rough visual setup. USB microscopes are cheap ($30 buys a good one) and easy to get. Using a mirrored surface and a microscope, try to position the stylus as vertical as possible.
2
Equal distortion test. Play back a stereo fixed frequency test track and aim for equal harmonic distortion in both channels.
3
High frequency output test. We use Denon XG-7002 or similar test LP, and aim for equal HF roll-off on both channels when playing back a series of 1-50kHz sweeps.
You will need some sort of instrument or software that can measure total harmonic distortion (THD) and plot accumulated spectral traces.
Please note that (2) and (3) above are only meaningful if the downforce and antiskating force are already adjusted and the cartridge isn't mistracking.
Please note that (2) and (3) above are only meaningful if the downforce and antiskating force are already adjusted and the cartridge isn't mistracking.
Visual Setup
We coulnd't find a single badly assembled cartridge in our collection! To illustrate the procedure, we had to use an adjustable headshell to twist the perfectly normal Sumiko Pearl cartridge a few degrees as in case (a). To be precise, α is 2.75°.
As we are not using a crosstalk method at all, the fact that its generator is properly aligned with the stylus shouldn't matter.
As we are not using a crosstalk method at all, the fact that its generator is properly aligned with the stylus shouldn't matter.
For the reference case, I have returned the headshell to its factory position. It resulted in an α of 0.45°.
Because our usual test rig is still unserviceable, I couldn't get satisfactory images from a microscope. To show the alignment angles, I have used macro photos of a stylus body. It works with our well-assembled Sumiko, but in general, it's a really bad idea. Focusing on the stylus and cantilever with a microscope will result in much higher precision.
Because our usual test rig is still unserviceable, I couldn't get satisfactory images from a microscope. To show the alignment angles, I have used macro photos of a stylus body. It works with our well-assembled Sumiko, but in general, it's a really bad idea. Focusing on the stylus and cantilever with a microscope will result in much higher precision.
Distortion Test
For distortion test, we have used a normal 3150 Hz track that is present on most test LPs for wow&flutter measurement. Any stereo fixed frequency track will do, the higher the better. The goal is to achieve equal harmonic distortion in both channels.
An azimuth that is off by 2.75° results in about 2.5-3 dB difference in THD between channels at 3150 Hz. If you use higher frequency test signals, difference will be proportionally larger.
Vertical axis of an FFT plot is in dB, horizontal is in kHz. Red/purple is for right channel, and green is for left.
Vertical axis of an FFT plot is in dB, horizontal is in kHz. Red/purple is for right channel, and green is for left.
With an azimuth within 0.5° of a vertical, the THD difference is close to the measurement error.
High Frequency Output Test
We have played back separate 1-50 kHz sweeps in left and right channels from a Denon XG-7002 test LP and plotted the results together. These sweeps are recorded without RIAA correction, so our result is nowhere close to linear — but it doesn't matter for the task at hand.
A 2.75° azimuth error gives us 5-7 dB difference between channels, culminating in almost 10 dB at 50 kHz.
Please note that the channel colours are reversed here. Green is right channel, and red is left.
A 2.75° azimuth error gives us 5-7 dB difference between channels, culminating in almost 10 dB at 50 kHz.
Please note that the channel colours are reversed here. Green is right channel, and red is left.
With a correct azimuth, the sweep curves are basically identical.
In our experience, THD and HF roll-off tests correlate well with each other. If you do not have an LP with a very high frequency sweep, a harmonic distortion test might be sufficient for most stylus shapes. The sweep test gives more precision with advanced linear contact styli though.
Like with most other static alignments, there's no point in chasing every 1/10th of a degree or millimeter. For azimuth in particular, styli are designed to perform optimally even with some alignment error. Getting it within 0.5° is fine.
I understand that the Denon XG series of test records is virtually unobtainable. Ortofon is selling an LP with a very high frequency sweep. While we haven't used it ourselves yet, it just might work.
Like with most other static alignments, there's no point in chasing every 1/10th of a degree or millimeter. For azimuth in particular, styli are designed to perform optimally even with some alignment error. Getting it within 0.5° is fine.
I understand that the Denon XG series of test records is virtually unobtainable. Ortofon is selling an LP with a very high frequency sweep. While we haven't used it ourselves yet, it just might work.
Note: if you are using a tonearm with detachable headshell, an SME/JIS bayonet mount can easily introduce significant azimuth errors. 1-1.5° slack is common if either a location pin on the headshell or a slot on the tonearm side is worn, and would prevent you from making repeatable measurements.
To end our series of posts, let's revisit the "common knowledge" about azimuth alignment that we've mentioned at the beginning, and see if there's any truth in it.
1
Azimuth is a normal and standard alignment procedure, just like overhang, SRA, tracking force, antiskating etc
In a correctly designed and adjusted playback system, the only source of static azimuth error is imprecision in cartridge assembly. Azimuth alignment is only needed to try and correct such assembly errors.
2
Correctly assembled quality cartridges require azimuth alignment to sound their best
3
Consequently, any tonearm that does not have azimuth adjustment is severely deficient
If we limit ourselves to correctly assembled cartridges, having an azimuth adjustment on a tonearm is a disadvantage. It can only introduce an error where there was none.
4
There is only one correct azimuth alignment for a given cartridge
For an imperfectly assembled cartridge, there are two "correct" azimuth settings: when the stylus is aligned and generator isn't, or when generator is aligned but stylus is askew
5
It is achieved by rotating the cartridge along its longitudinal axis until crosstalk between channels is at a minimum
Measuring and minimizing crosstalk will give us a cartridge position where the generator coils are aligned, but the stylus isn't. Compared to aligning the stylus itself, it would result in worse HF reproduction, more harmonic distortion and worse tracking.
The common knowledge about azimuth alignment is completely wrong
Basically, all the common knowledge about azimuth alignment is completely wrong. Like many things in analogue sound reproduction, it is a mix of misapplied 1950-70s research ("minimize crosstalk for perfect azimuth"), desire for market differentiation ("the more adjustments the tonearm has, the better"), questionable manufacturing practices, and lack of consumer education ("a cartridge assembly error of 5° is fine").
I hope that our series on azimuth alignment might help the industry professionals and the LP playback enthusiasts achieve quality playback, understand the problems involved, avoid frustration and save money. If you liked the series, we'll be very grateful for you sharing it — we have included social media buttons below and in the top right corner. Please give them some exercise!
I hope that our series on azimuth alignment might help the industry professionals and the LP playback enthusiasts achieve quality playback, understand the problems involved, avoid frustration and save money. If you liked the series, we'll be very grateful for you sharing it — we have included social media buttons below and in the top right corner. Please give them some exercise!
Don't forget to subscribe to get the latest blog updates!