Rational Acoustics



PhillipIvanPietruschka
August 11th, 2015, 07:29 PM
Having just refurbished my turntable and being faced with the job of aligning the cartridge and tonearm I used Smaart on a whim to help set cartridge Azimuth. Aiming for maximally similar VHF response between L & R channels on a mono cut of pink noise. This is relatively easy to do, and I was able to achieve matching not worse than 1.5dB in the top octave, an improvement on the 3.5dB to 4dB difference I started with (and perhaps with patience and a steady hand I could do better). Phase difference improved from 26deg to 19deg (not much). Subjectively, clarity of stereo imaging of also improved. Not going to make any claims that this is better than the popular method of setting for best and equal cross talk, however I suspect the end result would be very similar.

There are some more appreciable differences in the low end < 100Hz that I am not sure how to account for, and given it looks more like a drunk line, rather than a clear bias to one side or the other, I am inclined to guess this is characteristic of the cartridge (or cart-tonearm system) and not a something I can readily improve upon.

836

Purple was the starting configuration. Teal was as good as I got it. Red was what I ended up with after trying to improve further (ie slightly worse). With patience and delicacy I am sure I could match or improve upon the Teal trace.

Post edit: On further reflection a contributor to the poor data in the LF is probably a combination of surface noise, motor rumble, and residual buzz. It is however a good quality clean record, an a fairly quiet system for rumble and buzz.

PhillipIvanPietruschka
August 12th, 2015, 09:18 PM
837
Here is a better image showing 0.5deg variations, and then a 0.25 deg variation between the best two candidates. Range is from +1deg to -1.5 Deg. The best result, -1.25 deg is within 0.5dB except at two points in the HF.

Chris Tsanjoures
August 18th, 2015, 02:06 PM
This is a really neat post, thanks for sharing!

PhillipIvanPietruschka
July 2nd, 2016, 01:42 AM
This is a really neat post, thanks for sharing!

Thanks Chris. A very late follow up - I was persuaded, perhaps, that method is perhaps not the best because it conflates errors in channel balance inherent to the cartridge/stylus with errors in the azimuth. Essentially I am fixing, or so it has been suggested, the channel balance error with a deliberate mis-setting of the azimuth.

The best way seems to be to set for equal cross talk between the channels. I did this with a LP with a 1kHz tone tracks for centre left and right. I measured the result in the RTA and notated levels into a spreadsheet. I used two different ways to measure the cross talk, a 'normal method' of difference between tone in left, residual in right, and a proposed alternative method to account for channel balance issues difference between tone left and residual in left (when the tone is in right). Hope that makes sense.

I wish I could give you some comparative listening notes between the two methods. The reality is that to me they both sounded pretty good to me, and the end results where quite close to each other. Furthermore the whole process takes so long, that it is not even vaguely practical to do any kind of AB listening for subtle differences. I would never cut it as the kind of journalist that does reviews of expensive hifi cables!

915

The actual azimuth measurements, done with a mini spirit level on my SME tonearm, should be considered relative, as the azimuth clearly changes when in arm rest (where the measurements were done) vs when on a record.

One day when I have time I will use my cuts of pink noise to measure response differences with changes in SRA/tonearm height. I tried doing that with a loaned USB microscope and it was a right pain. You can probably tell I am in a real rush to do this.