When you take apart a lot of lenses, test them, re-arrange the optics in them: this makes perfect sense to me.
My advice to you: if the lens is giving good performance up close and wide-open, where you want it- leave it as is. If you are going to use the lens at infinity and wide-open, use the distance scale and push the RF past infinity.
Spherical aberration is responsible for focus shift. The focal length of the center of a lens that is over-corrected for spherical aberration, like the Canon 50/0.95, is longer than the focal length of the edges. This means the best focus point of light entering the lens from the center is behind that of light entering from the edge. Used wide-open, the image is dominated by light coming in at the shorter focal length of the edges. SO- on your lens, TEST THIS: focus close-up and wide-open; Focus at F2.8 using the RF to determine focus; AND infinity focus at F0.95 using the distance scale. If all work correctly, tell the Technician that he optimized the lens beautifully. When using deep yellow, orange, and red filters: the focus shift is towards infinity. Make some test shots. I end up optimizing lenses for my M Monochrom for Orange filters, close-up, wide-open. 0.02mm difference in the Shim.
This is exactly how I optimize a 52.4mm Sonnar formula Zeiss and Jupiter lens to focus on a Leica. I allow the hard stop for infinity to go past the infinity as shown by the RF. That allows me to use the lens wide-open at infinity. I have also rearranged the optics lenses to change the focal length to agree with the RF across range. Increase the distance for a 5cm F1.5 Summarit to optimize for F1.5 as the true focal length at manufacture is 51.1mm, and decrease the distance for a Zeiss Sonnar and Jupiter which is nominally 52.4mm.
To me, this would make sense, in theory. Skyllaney got back to me with this (very fast!):
"If your lens is focusing correctly at the minimum and mid focus distances, it could be that the lenses effective focal length is below 51.6mm, which is the calibration standard for Leica. As an example, a lens that had an effective focal length of say 50.5mm , and calibrated to be accurate at the minimum focus distance, wide open... would cause the lens to arrive at true infinity before the Leica rangefinder split patch arrives at infinity.
This is because the conversion of the Canon 50mm f/0.95 lens, uses a fixed RF cam that is mounted off the rear of the optical block. In theory, only a lens who has a true effective focal length of 51.6mm, would this work perfectly.
Lenses below 51.6mm would arrive at infinity before the rangefinder does, if optimised for MFD.
Lenses above 51.6mm would not arrive at infinity when the rangefinder does, but fall short, if they were optimised for the same MFD as the above scenario.
[...]
Its possible we might need to make a few custom shims, and find a way to adjust the lens from say a 50.5mm lens to become a 51.6mm lens (an example). Once at 51.6mm effective focal length, and calibrated so the lenses infinity coincides with the rangefinder patch infinity, all other distances by nature of the lens being 51.6mm now, will be in correct focus."
I'll have a go at the tests you've suggested and, failing that, send it over to Skyllaney for an evaluation. If it's OK as is for close up and medium focus, then that's all I need and I'll use it as is -- that is.. until Skyllaney tempt me with a good price.
