I just tested my lenses for accuracy of focus. I only tried one thing - a ruler on the floor 6 feet from the camera, at 45 degrees. Lens wide open, on a tripod.
35/1.7 Ultron 3.3 inches front
21/2.8 Avenon 8.4 inches front
28/1.9 Ultron 0.6 inches back
50/2.0 Hexanon 2.2 inches back
35/2.5 C-S PI 2.8 inches front
21/4.0 C-S 6.1 inches front
These are line-of-sight distances, as you would see on the scale on the lens barrel, not the distance along the floor.
interesting, no? I haven't really got a clue what to make of this. The focus seems repeatable, so it's not random; it's both front and back, so it's not a systematic R/F bias. And they all show coincident rangefinder images at infinity. All the adapters are CV, and seem to be identical.
Anyone got any thoughts? Any input welcome