The “Korf method” for azimuth alignment

[tony22]

The thread title refers more generally to azimuth alignment that targets stylus perpendicularity to the groove as the goal for optimal azimuth, using either distortion or HF measurement as the method. I imagine most vinylistas already know about this technique, but for those who do not, here’s the link -

Azimuth Adjustment Myths and Reality, Part I

korfaudio.com

The stylus perpendicularity method essentially ignores shooting for equal channel separation. I hadn’t read this in some time, and after seeing it again I wondered what parameter is being used by the old Adjust Plus+ or AnalogMagik v1 or v2 - or even the Fozgometer - for setting azimuth? Generator alignment I guess? I’ve spoken with Peter Lederman a few times about azimuth, and he points out that at least his own carts when properly aligned, given their very high channel separation to begin with, can have different numbers from L>R and R>L. That would suggest (what I’ll keep calling) the Korf method is a more useful technique, keeping the stylus from excessive vertical skew because of maybe shooting for the wrong goal. Wondering what other folks think.

 

[mtemur]

I wondered what parameter is being used by the old Adjust Plus+ or AnalogMagik v1 or v2 - or even the Fozgometer - for setting azimuth?

They all use the correct method, maximum channel separation and minimum phase error.

The thread title refers more generally to azimuth alignment that targets stylus perpendicularity to the groove as the goal for optimal azimuth, using either distortion or HF measurement as the method. I imagine most vinylistas already know about this technique, but for those who do not, here’s the link -

Korf blog contains some useful insights for vinyl playback but their azimuth method is not one of them. When you read the blog you get a feeling like every manufacturer before Korf is stupid and cheated us with poor products. I don’t know maybe it’s just me or their writing technique. Of course there is some reality in their assumptions but not as dramatic as they portrayed. IMHO take everything in Korf blog with a grain of salt. I still believe that their tonearm can be incredibly good though.

I used to think like them about Azimuth in the past but soon I realized perpendicular stylus is not the correct way. Scanning sides (edges) of stylus in contact with groove walls are not straight lines. They’re rather like an arc or hyperbolic.

What is the benefit of arc shaped edges on a stylus? Well, when the stylus is tilted to one side 1-2 degrees it will still be in perfect contact with groove walls. Modern profiles are cut like an arc on the edges and tip for this purpose rather than straight lines. Don’t let illustrations on the Korf blog fool you. They’re drawn as a triangle.
I don’t want to use their illustration without permission. It’s similar to the one below

It is illustrated as if stylus shape is a triangle with straight lines on the edges and whole triangle fits inside the groove. But in reality stylus is not triangle and only a small part of the curvature area fits inside the groove as you can see below. And that area is close to the tip which has a curved shape. So, modern profiles are not exactly circular shaped like a spherical but still keeps contact with groove walls up to a couple of degrees tilting.


As a result it’s better and correct way to adjust azimuth according to maximum channel separation and minimum phase error up to 1-2 degrees. If You need to tilt more than that than change the cartridge or use it as it is. Reducing phase error is more important cause human ear is very sensitive to phase error. IMHO using AP test disc is very good for setting azimuth.

 

[tony22]

Korf blog contains some useful insights for vinyl playback but their azimuth method is not one of them. When you read the blog you get a feeling like every manufacturer before Korf is stupid and cheated us with poor products.

Right! I'd been using the accepted methods for years and I go back again to this blog and see this again.

As a result it’s better and correct way to adjust azimuth according to maximum channel separation and minimum phase error up to 1-2 degrees. If You need to tilt more than that than change the cartridge or use it as it is. Reducing phase error is more important cause human ear is very sensitive to phase error. IMHO using AP test disc is very good for this purpose.

If the goal is to get maximum channel separation and minimum phase difference, it seems a tradeoff will have to be made if phase difference is the priority. That would suggest Adjust Plus+ as being perhaps a slightly easier tool to use for this purpose (at least IMO). AM does provide phase numbers but I think the graphing technique used by AP makes it easier to "see" where things are going as the cart is rotated. Having said all that, we've all had the discussion in another thread about the numbers between AM and AP not being the same for azimuth. Beyond that I've had numbers be virtually dead identical from L>R and R>L with AM and be unequal in AP, and vice versa, ignoring the values not being close.

 

[mtemur]

If the goal is to get maximum channel separation and minimum phase difference, it seems a tradeoff will have to be made if phase difference is the priority. That would suggest Adjust Plus+ as being perhaps a slightly easier tool to use for this purpose (at least IMO).

It looks like Adjust+ is easier to use but I’m not sure if it’s better than AM. Normally -but not always- phase and best channel separation coincide.

Beyond that I've had numbers be virtually dead identical from L>R and R>L with AM and be unequal in AP, and vice versa, ignoring the values not being close.

I get equal numbers for L>R and R>L on both AM and AP discs only on vacuum hold down platters. Numbers are not the same but equal on both channels, something like 28dB for AM and 35dB for AP. Other than vacuum hold down platters I still reach to the same adjustment using either test disc but presented numbers are only equal with AP disc. What I’m trying to say is both test discs lead you to the same correct adjustment when you try to achieve highest possible numbers for both channels. If they’re also equal it’s even better.

 

[tony22]

but presented numbers are only equal with AP disc

I do get the best coincidence of results with the Analog Productions Ultimate Analog (I’m hoping that’s the one you’re referring to) test record. The AM and the Adjust Plus records when used with their respective programs yield the differing results I see. Makes sense I guess. Nobody really knows how well all these test records are cut and pressed. And for some reason Mikey doesn’t seem to like the Adjust Plus disc for azimuth.

 

[mtemur]

I do get the best coincidence of results with the Analog Productions Ultimate Analog (I’m hoping that’s the one you’re referring to) test record.

Yes, that’s the one.

 

[tony22]

You're right of course, @mtemur. I got out the Ultimate Analog test record, and after a tiny (and I mean tiny) bit of tweaking I got Adjust Plus+ and AMv1 to show almost identical (off by a couple of tenths of a dB) azimuth readings, on both programs L>R and R>L off being identical by .4 dB. Good enough for me. APlus shows the phase delta to be 4 degrees. Hoping that's low enough. Trying to get that any smaller was an exercise in frustration. And as others have noted, when all that was done I looked at azimuth on AMv2, with the v2 test record (since on v2 azimuth can't be done with separate L and R tracks) and the difference was on the order of 12dB. Really not sure what to think about that, as two other programs (oh, and the Visual Analyzer SW oscilloscope) all show very close results to each other.

 

[mtemur]

You're right of course, @mtemur. I got out the Ultimate Analog test record, and after a tiny (and I mean tiny) bit of tweaking I got Adjust Plus+ and AMv1 to show almost identical (off by a couple of tenths of a dB) azimuth readings, on both programs L>R and R>L off being identical by .4 dB. Good enough for me. APlus shows the phase delta to be 4 degrees. Hoping that's low enough. Trying to get that any smaller was an exercise in frustration. And as others have noted, when all that was done I looked at azimuth on AMv2, with the v2 test record (since on v2 azimuth can't be done with separate L and R tracks) and the difference was on the order of 12dB. Really not sure what to think about that, as two other programs (oh, and the Visual Analyzer SW oscilloscope) all show very close results to each other.

I believe 0.4dB channel separation and 4 degrees phase differences are extremely good. I have similar experience with AM test record.

 

[tony22]

I know (think) we kind of settled on agreement with this, but I still scratch my head about Korf’s method. My thinking is along these lines - let’s say one uses his 50KHz swept tone method. Of course we have to assume the tracks are cut identically for Left and Right - let’s assume they are. If this is true, then wouldn’t lining up the swept curves as best as possible be “as good as you can get” in aligning the azimuth? Or is this method primarily really getting us a theoretically optimized channel balance, and is really missing the point of cross talk cancellation?

 

[Balle Clorin]

@tony22
Two different systems MM and MC both Microline and Boron cantilever , used Ortofon test record to adjust azimuth for best crosstalk gives this on 50kz track. I think using Ortofon record for best crosstalk is good enough. Sweep here is Denon Xg-7002 test record.
I use 400hz high pass filter when setting azimuth, that removed low frequency rubbish from the crosstalk level

 

[tony22]

@Balle Clorin, this is rather interesting. I have both the Ortofon and the Denon records, and in both cases I get traces with amplitude descending as the frequency increases, much like that shown on Korf’s web page. How are you getting a (relatively) flat amplitude? I’m running the signal through my phono pre before sending it to my A/D USB converter.

Also, have you compared this method to standard crosstalk techniques (using either AM or A+, or even a Foz)?

 

[Balle Clorin]

The XG-7002 sweep record is without RIAA, I record without. If you use RIAA the level will drop as frequency increases.
I use the standard crosstalk method with the Ortofon L/R 1khz record as described in post above.
When I optimize Azimuth on the headshell on Ortofon record, I get this crosstalk on a Hungarotone record recorded with varying azimuth, Good match there also, this is for another cartridge. So setting Azimuth optimal on Ortofon also gives very good performance on both XG-7002 and Merolemez/Hungaroton . I do not have Analog Magic, Alalog Productions Ultimate Analog test record is quite poor and do not agree with many of my other 25 test records.EDIT but the crosstalk track is not too bad..

 

[tony22]

I use the standard crosstalk method with the Ortofon L/R 1khz record as described in post above.
When I optimize Azimuth on the headshell on Ortofon record

So you’re first using the Ortofon 1 KHz tone to set crosstalk, and only after that you’re using the 50KHz sweep simply as an azimuth check? What software are you using for these two traces shown here?

 

[Balle Clorin]

Yes.
For plots I use the script at ASR

Phono Cartridge Response Measurement Script

I thought neodymium had been in use for MM magnets for ages? I think it is specifically the “55” version which is the extra strong neodymium.

www.audiosciencereview.com

I record using Soundforge/CooleditPro via Parks Audio Puffin. But any method can be used. Puffin gives the option to record without RIAA, Test frequency sweeps rarely are RIAA, often only RIAA below 1000hz. The script can fix this

 

[tony22]

Thank you Balle. I often use an oscilloscope or spectrum alalyzer (depending on what I’m looking at) and just do the math myself. The ASR scripts look interesting, though, so maybe I’ll give them a try.

so you like the Ortofon tracks the best for crosstalk? What do you like about the Hungaraton disk?

 

[Balle Clorin]

See the plot it gives . It is recorded with 40-50 degree groove emulating 85 to 95 Azimuth. But more complex( analysis of digital recorded file) than using the Ortofon which is only recording L an R signal level while adjusting cart. My Puffin RIAA gives the crosstalk directly on the display…

 

[tony22]

Balle, do you have more than one copy of any given test record? In a different thread I was remarking that I’ve discovered among the test records I have where I do have multiple copies, that the crosstalk results from one copy to the next show more than a minor difference. It was suggested that this is because of imperfect record flatness from one to another. I use a record clamp and periphery ring, but this may still be true. But I wonder how anyone can ever know, then, that an ideal crosstalk alignment has been achieved?

 

[Balle Clorin]

Yes one I have 3 of, two others I have 2 of andthen 25 single ones.
The pressings and results are similar for the same record name, what different is the flatness, which has a effect in on WF.. Different records ( not copies) show very variable crosstalk.See table above. Azimuth test should be done with high pass filter, and is best don’t when recording flat without RIAA, if not low frequency grunge may obscure the actual crosstalk level

You may find som interesting stuff here

Fun with vinyl measurements

I think Vinyl people in other forums do not want to know about turntable and cartridge measurements as they may get disappointed. I like accuracy and good design but that does not prevent me from listening to vinyl or measuring it. Measuring is a way to learn,understand and improve. Let us...

www.audiosciencereview.com

 

[tony22]

The pressings and results are similar for the same record name, what different is the flatness, which has a effect in on WF

Thanks.

 

[tony22]

@Balle Clorin, I still feel that we’re essentially “chasing our tails”, with regard to crosstalk as there is too much variation between test records - and even between multiple tracks on the same test record. I give you the following example. The first two samples are the first pair of L,R tracks on the Ortofon test record, followed by the second pair on the very same record! The third sample is one of my Analog Productions records, the last sample is from another copy of the same AP record.

I am definitely coming to the opinion that this is all just a game meaning next to nothing. If copies of the same tracks in the same side of the same record can’t yield similar results, the the whole thing is a joke.