A question of Azimuth setting up turntable


In process of setting up my VPI Prime table and am questioning myself of whether or not I'm setting azimuth correctly.  All else is setup perfectly VTA, SRA, VTA etc. While setting up azimuth where should the measurement be taken? At the headshell or at the cartridge? While setting up my Ortophone 2m Black using a Azimuth block I checked Azimuth at the bottom of the cartridge while lowered on the platter which showed perfectly horizontal but at the headshell it was off by a fair margin. Nothing outrageous but it was off. Am I missing something or? Cartridge was installed at dealer but not fine tuned. So where does the audio community check and adjust their Azimuth? In my view azimuth at the cartridge should be preferred but I could be mistaken. Please correct me if I'm wrong. What is the correct or preferred way to set azimuth?

128x128gillatgh
I don't get how you could have a different reading at the top verses bottom of cart as long as both readings are taken while the cart is lowered in the groove.
Short answer: azimuth should be measured at the stylus to LP interface.

Ideally, you'd use something to measure channel crosstalk while playing a test record -- something like the Fozgometer, but even that method seems a bit controversial around here. 

Many will place a thin mirror on the platter and rotate the cartidge until the stylus is perfectly aligned (straight up and down) with its mirror image. It's important that the mirror height is at the same level as the LP surface.

One can also get an analog to digital converter with USB output to connect to a computer. There are free softwares (sometimes included with the device) that can measure channel output when playing a test record. 

I personally fine tune it by ear after using the mirror method. I simply adjust it until the sound no longer improves while playing mono records. 
Fwiw, leaving azimuth as is using the bottom of cartridge horizontal with azimuth block I get beautiful center imaging and great soundstage. Everything sounds good but could it be better? Please keep feedback coming. 
How could the bottom of cartridge and headshell be different? Good question, I have no clue. I'll try the mirror method, sounds most reasonable after all it is all about the stylus angle.
Thanks for the replies so far.

helomechm
Short answer: azimuth should be measured at the stylus to LP interface ... Many will place a thin mirror on the platter and rotate the cartidge until the stylus is perfectly aligned (straight up and down) with its mirror image.
Yup, I think that's the best approach.
Depends on your level your looking for. Using the mirror method gets you very close but with your shibata stylus you want better than close. You need a way to measure output of each channel using test LP to optimize the output and also minimize crosstalk. This also gets you the optimum soundstage from your 2MB.

I use a Foz meter, not sure what is controversial about it. It reads both channels and also you can read/adjust for minimal crosstalk.
The Foz is controversial as some maintain that crosstalk will vary at different frequencies hence setting only for 1khz is unlikely to be optimal. Hence methods like this which I think works well
http://www.durand-tonearms.com/Support/Support/azimuth.html

Not sure how you can get a different adjustment/reading from headshell and cart- makes no sense. 

Just to add to the responces since no one has mentioned it, with a unipivot, the AZ adjustment tends to change with VTA. (That's why it was mentioned it has to be taken at record height). 

And visually, you should be looking straight at the cart, not straight down the arm.
There has been so much said and written about the Foz meter's failings that I personally stay away.

IMO if you buy an expensive cartridge (say, more than $700 or so) I would not accept it unless the stylus by a very careful examination by eye and by loupe is straight with the body of the cartridge.  Peter maintains that should be true with his cartridges.  It is also true on my Shelter and Audio-Technicas.  That way you would be able to get a first, and very good, approximation of good azimuth by lining up the cartridge.  Many of them have a horizontal line at the front.  That is shown on the cartridge in Peter's video.  It is also on so the Shelter and Audio Technicas. 

What I do, while a record is playing, is to shine a small flashlight on the front of the cartridge and adjust the azimuth so that the horizontal line lines up parallel with its reflection on the vinyl.  This is usually good enough for me, certainly on the A-T.  But it is easy enough to make very small adjustments afterwards if you wish and listen, listen, listen.

The ear is the ultimate judge.