Hi guys,
Debates on this level of refinement (which alignment pattern to use or which tonearm offers lowest tracing distortions) are really insignificant compared to zenith error tolerances in the cartridge industry. Your zenith error should be known and corrected before any of this inquiry will deliver results. The average zenith error in the last few hundred cartridges I’ve analyzed is a hair over 2.5 degrees. If we look at the dataset with Gyger replikants removed, that number goes up substantially. Industry assembly tolerances are so far beyond the differences between Loefgren, Baerwald, Stevenson, etc that the debate and inquiry offers limited “yield” until zenith correction is satisfied.
I don’t mean to rain on you fellows. There still may be value in this inquiry but it can’t benefit us until we’ve first managed the elephant in the room.
I am still in development with a tool to correct for zenith error in the field so it does not require sending to me in the lab. Technically, the WallyZenith allows for it already but the correction involves using your ears in a rather tedious process of listening for best results. The tool makes it very easy to know where you are and return to any previous position with a high level of accuracy.
You might ask why I bothered to put alignment variations in the WallyTractor in the first place then. Simply: I redesigned the WallyTractor before I came to realize how poorly stylus/cantilever assemblies are made and mounted.
In a product development process I had to do repeated static and dynamic repeatability tests for zenith error in which I repeatedly aligned the cantilever by 0.5° increments. We were able to confirm that these attempts were accurate though I do believe 0.5° is likely my limit of resolution when aligning a cantilever. One of the engineers on our research team who has a fantastic analog rig and is a VERY discerning headphone listener insists he can hear 0.5° zenith error differences on his pivoted tangential tracking arm. I have not blind tested him on this but I’d say that I’m not worried about resolving below 0.5° accuracy based upon what I’ve seen in Matlab and Abaqus with respect to fine changes in zenith error.
The tracing error difference between Loefgren and Baerwald is less than that. So is the difference in tracing error between a 9” arm and a 12” arm.