Raist3d wrote:
hindesite wrote:
Raist3d wrote:
The DFD will work better with Panasonic lenses, and yes, a full profile will realize the full benefit. However, as the Japanese GX850 brochure states, there's a partial DFD benefit even using non-Panasonic lenses (which does include Olympus lenses).
[citation needed]
I already did that before. Go ahead to Panasonic's Japanese website, go to their section on GF9, and translate the Japanese text. It's their Japanese website, not the PDF brochure (my small mistake):
http://panasonic.jp/dc/g_series/gf9/high_image_quality.html
Translate the section on AF- look for "空間認識AF"
"Improve autofocus speed and tracking performance
Adopt space recognition technology (DFD * technology) to instantaneously calculate the distance to all subjects appearing on the screen. Contrast AF which high precision was popular in the past, has also realized faster AF (autofocus) and improved tracking performance. In GF 9, even when using micro Four Thirds lenses other than ours, we can improve AF performance by utilizing a part of DFD technology.
* Abbreviation for Depth From Defocus."
Emphasis mine. You are welcome
I have also mentioned a few times that I thought Panasonic used a form of "DFD Lite" to get its quicker than usual CDAF performance and this manual snippet more or less confirms what I had thought common sense.
Anyone who has used manual focus lenses for more than a moment can agree that a slight wiggle turn in one direction will soon enough visually disclose whether the lens is moving towards focus or away from it. The degree of out of focus also give some indication of how far out of focus the lens is and some indication of how far the lens must be adjusted to get within precise focus distances.
I would imagine that if we were to make a science of it with one favourite lens our eye-brain-finger co-ordination might get quite good. Experienced users might not even need a preliminary test turn to determine which direction to twist.
Electronically a DFD lite system might be able to not only tell in which direction the lens is out of focus but also give a good estimation of how far it must travel quickly to get close to good focus.
Therefore "DFD Lite" must only be an electronic version of what any user of manual focus lenses can see and manipulate on any lens. Full DFD must be when individual lens focus characteristics are fully documented and the process refined.
Therefore if (guess say) DFD lite gets 80% of the speed of PDAF (maybe faster?) then perhaps lens statistic based full DFD might add another 10% to this. I simply do not know facts but use the figures only to illustrate my thoughts. As PDAF is supposed to be very quick but lacking in finite precision even good pdaf systems can be hybrid and switch to cdaf for absolute precision.
Whatever method is used there is no doubt that Pansonic bodies focus Olympus lenses acceptably quickly.
I doubt if there is any secret on what needs to be done, just how it is done electronically is much harder.