Originally posted by redcat wow, just wow, you really take your time to fine tune these. Thanks a lot for sharing the results with us
You're welcome
---------- Post added 07-06-16 at 12:20 PM ----------
Originally posted by c.a.m tomO2013:
Would you mind checking the pdf files for the FA 43? The data appear to be identical in both files, so I assume they are the same Focus Tune run (at either 25X or 50X). I'd be interested in comparing the results at the two different focus distances.
Thanks
- Craig
Great spot - my bad sorry!
I'll setup again later today and see how I get on.
One thing that I am interested in is simulating backlighting situation with the same focus tune and seeing the behaviour of the AF. Right now, I am wondering if the spherical aberrations at wide open to close to wide open is screwing with the PDAF points.
Also more precise ring motor lens would probably give more accurate.
---------- Post added 07-06-16 at 12:25 PM ----------
Originally posted by johnkrumm I spent a couple hours trying to fine tune my FA 50 1.4 for single point auto-focus and could never get it to work properly at all distances. Then I tried the 9 point cluster and most problems went away. It focuses on what I aim at. However, because the camera has some choice, it won't always choose the exact spot I want, so I have to pay attention to what focus points engage. Overall I'm happy, but if they can get that single point working better in firmware, awesome. Later today I'm renting the 28-105 and I'll see how it does with a modern lens.
That's a good suggestion.
The thing with the 9 point is that it does generally (for fast moving portraiture with the FA Limiteds) give focus on the face, but doesn't single out the correct part of the face i.e. the eyes and so can focus on the tip of the nose or forehead etc.. I prefer to use a single point and sometime focus recompose so that I can ensure eyes are in focus.
If you have ever used the Sony A7ii, A7rii, Em1 - they can focus in on the specific eyeball! Similarly with the D750 and D810 in OVF they use the metering sensor to track and identify the eyes. I would love to know if Ricoh are doing this, as the physical hardware looks to be present (PDAF and advanced metering sensor). This could likely be something easily improved via a judicious firmware update.