Sometimes patents can be very…funny. The last one found by Egami (translation here) shows how Olympus wants to improve the AF performance by having a different height of the micro lenses. It’s a bot hard for me to get into the details of the patent without having an appropriate translation. But it sounds like this technique can give you a sort of phase detection alike performance.
14 Responses
Again, will it help focusing legacy 43 lenses or just help finding the right direction for µ43 ones?
i have good faith 🙂
with oly faith is all you need.
Considering that Olympus has unceremoniously discontinued a camera line before… no, not really.
“Maybe, maybe not.”
yes
It’s an ingenious idea, but the implementation is not possible with current sensor technology. This design should work well with near telecentric lenses such as the ones in the 4/3 lineup. It won’t work as well for some m4/3 lenses (notably pancakes) which project rather oblique light rays to the edges of the sensor.
maybe if they only had these special micro lenses in the middle of the sensor? where light usually comes straight (sans tilt shift lenses I guess)
I dont think telecentrics should make all that difference, ideally they would design one microlens layer for both mFT and 43 lenses. Microlenses exist to correct off axis light, better fill sensels, and create a gapless cover for max light gathering
This would indeed be able to synthesize pdAF from cdAF wired sensors, although the wiring would end up a little different and of course processing more difficult.
Actually the concept of micro lens has existed long time. The problem is that it’s very difficult to get to the production status. None of current technologies is plausible (both financial and technical)to make it become true. Leica seemed to invent something similar to this patent.
Microlenses exist right now, and are used in nearly every imaging sensor as used in digital cameras.
I posted this on DPR 3 days ago
and it actually follows from some thoughts posted months ago, although no doubt these guys thought of it first. Not sure the micolenses can be achieved as illustrated, but all that is required is to alter focal lengths over some cdAF points, could possibly be done in other ways (I thought of a brilliant one the other day)
While one of the limitations is the scope of FL for various microlenses, which is why it appears staged as 3x FL, not sure thats completely necessary
I really like this patent as it demonstrates that Oly still are working hard on solving the C-AF problems (with the FT lenses?).
Don it would establish an improvement in both camps, but to be honest, remaining competitive in mirrorless makes this a vital step with or without 4/3rds
Opera gates