How long until AF with 4/3 lenses is fixed?

~tc~

Mu-43 Hall of Famer
Joined
Oct 22, 2010
Messages
2,494
Location
Houston, TX
No way - there's no money in it.

Evidence? If bright, super high quality zooms sold to the public well, everyone would have bought into 43s (the SHG zooms are widely acclaimed as the best available), and Olysonic would have come out with the native versions instead of umpteen iterations of slow zooms.
 

dhazeghi

Mu-43 Hall of Famer
Joined
Aug 6, 2010
Messages
4,381
Location
San Jose, CA
Real Name
Dara
No way - there's no money in it.

Evidence? If bright, super high quality zooms sold to the public well, everyone would have bought into 43s (the SHG zooms are widely acclaimed as the best available), and Olysonic would have come out with the native versions instead of umpteen iterations of slow zooms.
It's not immediately obvious to me that the failure of 4/3 is a reflection of the popularity or lack thereof of the lenses.

More to the point, the R&D and production costs of all 4/3 lenses have already been paid. If implementing PDAF allows them to unload their existing stockpile at more than firesale prices, as seems reasonable, they could stand to make a good amount of money. Redesigning those lenses for m4/3 from scratch would be far more expensive, and likely far more risky.
 

~tc~

Mu-43 Hall of Famer
Joined
Oct 22, 2010
Messages
2,494
Location
Houston, TX
But also, probably much smaller, lighter, and cheaper to produce - and therefore more sellable to the "general public".

Remember here, that the enthusiast market is a drop in the bucket. The real money is with the soccer moms, not us, and, heck - hardly any ENTHUSIASTS are buying the expensive, large, heavy, 43s lenses - despite how fantastic they are!
 

amalric

 
Joined
Jul 24, 2012
Messages
183
Location
Rome. Italy
You can't directly compare lenstip's 4/3 and m4/3 tests directly because they use different sensors - the E-3's sensor can't record much more than 50 lp/mm, whereas the E-PL1's sensor can record nearly 80 lp/mm.


What indications are there that the newer lenses are matching the higher resolution sensors? Even most good m4/3 lenses are unable to resolve 80% of the E-PL1's possible resolution across the frame, to say nothing of the newer 16MP sensor.
Do you have any proof of this? Lenstip showed precisely that the resolution of 4/3 lenses didn't improve when changing from the 10 to the 12 Mpx 4/3 sensor.

Olympus m4/3 lenses instead have increased resolution since then, according to Lenstip.

Since their tests measures the combined effect of both lens and sensor, 4/3 and m4/3 can be compared. I take that the bottleneck is in the lens.

I have seen no proof that 4/3 lenses are better than the expensive m4/3 ones. So i take it as an assumption, with no evidence provided. Miniaturisation doesn't entail necessarily less resolution.

I seriously doubt that Oly's engineers are restricting themselves in order to leave the advantage to 4/3 lenses. It makes good sense instead that they are increasing resolution of native m4/3 by optimising the sensor and lens combination.
 

dhazeghi

Mu-43 Hall of Famer
Joined
Aug 6, 2010
Messages
4,381
Location
San Jose, CA
Real Name
Dara
But also, probably much smaller, lighter, and cheaper to produce - and therefore more sellable to the "general public".

Remember here, that the enthusiast market is a drop in the bucket. The real money is with the soccer moms, not us, and, heck - hardly any ENTHUSIASTS are buying the expensive, large, heavy, 43s lenses - despite how fantastic they are!
Remember that Olympus sells roughly 1.5 lenses per camera body, meaning the vast majority of the unit sales are kit lenses, which are both low cost and low revenue. Each 12/2 they sell makes them more lens revenue than 8 camera kits. Profits are less clear, but they are certainly higher as well. So while I agree that the volume of high end lenses is small, it's quite possible for them to produce significant money.

Of course, none of that's to say that 4/3 sales are at such a level to matter and indeed at this point I doubt they are. Whether that can be changed is another matter.

Do you have any proof of this? Lenstip showed precisely that the resolution of 4/3 lenses didn't improve when changing from the 10 to the 12 Mpx 4/3 sensor.
Where is this lenstip test of which you speak? There is certainly ample evidence to the contrary, starting with the resolution charts shot by DPR.
 

amalric

 
Joined
Jul 24, 2012
Messages
183
Location
Rome. Italy
Where is this lenstip test of which you speak? There is certainly ample evidence to the contrary, starting with the resolution charts shot by DPR.
LOL. You provide no evidence and ask me for one?

Despite this, here we go. There is a nested graph comparing resolutions in the original

Olympus M.Zuiko Digital 17 mm f/2.8 review - Introduction - Lenstip.com

The comment:

"The heart of the camera is a well-known 12-megapixel sensor of the 4/3 format from extensively tested E-30 and E-3 models, allowing recording images in RAW format, which we think are not improved too much. The question arises, can the results of a 12-megapixel sensor be directly compared to ones of a 10-megapixel sensor? Well, they can. When we conduct MTF50 measurements by virtue of unsharpened RAW files, the increase of pixels from 10 to 12 million doesn’t really cause noticeable improvement of MTF50 values. It’s worth to remember the results obtained on E-520, E-30 and E-620, for which one and the same lens was used – Sigma 1.4/30 mm.

(graph)

As you can see, there are no measurable differences between a 10-megapixel E-520 sensor and 12-megapixel sensors of the two other cameras. And that’s not because of a weak lens, as some may suggest. It’s because Sigma 1.4/30 is not weak, quite the opposite, around f/5.6 it is very sharp. We’ve ultimately explained the case, mounting lenses like ZD 25 mm f/2.8 or ZD 150 mm f/2.0 on Olympus E-3, E-30 and E-620, lenses which are everything but not weak. In all cases, whichever camera was used - a 10- or 12-megapixel, the maximum MTFs, obtained at apertures free from optical aberrations, reached just over 50 lpmm."


By comparison the latest Olympus m4/3 lenses reach much higher MFTs - 60 lpmm and over, so I don't see what mirrorless users are losing. There might be little point in using 4/3 lenses after all.

EDIT: except for those that already have them.

Now Oly seems to have a surprise in store:

http://www.43rumors.com/olympus-surprise-we-will-do-the-hybrid-ft-and-mft-camera-by-end-2013-no-classic-ft-e-7-camera-anymore/comment-page-1/#comment-257728

:)
 

Kiwi Paul

Mu-43 Top Veteran
Joined
Aug 15, 2011
Messages
729
Location
Aberdeen Scotland
While MTF figures are all very useful the real world test is to me the only one that matters and as a previous owner of the 14-35 and 35-100 lenses I can vouch that I've not seen any m4/3 zoom lens match the IQ of these 2 lenses, even the 12-35 and 35-100 mft lenses.
Sharpness and resolution are not the only characteristics that make a good lens anyway but rather how all the qualities of a lens combine to make it special and the 4/3 SHG lenses combine all their qualities to make them outstanding.

paul
 

dhazeghi

Mu-43 Hall of Famer
Joined
Aug 6, 2010
Messages
4,381
Location
San Jose, CA
Real Name
Dara

RDM

Mu-43 All-Pro
Joined
Jan 6, 2011
Messages
1,408
Location
Orange County, New York
Well, that sure was timely now, wasn't it! Thanks for the link.

I guess the only downside I see is that it's going to be available in a large body. Hopeful the feature will migrate down the lineup eventually though.
Very timely .. And when you look at the poll results we can see how little faith in Olympus we all had :rofl:

I don't believe that's a down side. If a little larger was something consumers didn't want i doubt Panasonic would;d hive gone that way too.
But Remember, Olympus does not ever make only one model camera.. many like the larger and many like the smaller .. Some prefer the OMD some the E-PM1 .
I think the Idea of this camera is to replace their Professional camera (The E5) and most of those users want a good sized camera in their hand.
Also Olympus usually releases the the more expensive enthusiast models first then the more prosumer friendly ones. I am still waiting for a cheaper version of the OMD or something with a Built in VF and IBIS from Oly that i can afford..lol
So don't worry about the size of that camera people, i think there will be a camera from Olympus to best suit your needs if its not this one.
 

~tc~

Mu-43 Hall of Famer
Joined
Oct 22, 2010
Messages
2,494
Location
Houston, TX
Olympus has been asking for quite some time what enthusiasts would like in a DSLR body to use up the extra space available if it had mirrorless "guts". Will be interesting to see what they came up with.
 

dornblaser

Mu-43 Hall of Famer
Joined
Aug 13, 2012
Messages
3,388
Location
Chicago-area
Real Name
David Dornblaser
That should be interesting. I wonder if that body would make sense for someone who is strictly m4/3s?

Post-Script - the link to the hybrid camera article made my vote in the poll easy. :cool::eek:
 

dhazeghi

Mu-43 Hall of Famer
Joined
Aug 6, 2010
Messages
4,381
Location
San Jose, CA
Real Name
Dara
Olympus has been asking for quite some time what enthusiasts would like in a DSLR body to use up the extra space available if it had mirrorless "guts". Will be interesting to see what they came up with.
I handled an E-5 and an E-M5 side-by-side yesterday. Olympus definitely knows how to build big bulky cameras if they want. That E-5 is larger than a lot of full frame bodies and it wasn't all the mirror box. The E-410 was after all one of the smallest DSLRs.

camerasize.com gives a good sense for the differences in size among the E-5, E-M5 and GH3.

Offhand, I expect a huge built-in grip, a top LCD, an articulating rear LCD and a lot of big buttons should bulk things out nicely, though I rather wish they wouldn't.
 

dhazeghi

Mu-43 Hall of Famer
Joined
Aug 6, 2010
Messages
4,381
Location
San Jose, CA
Real Name
Dara

amalric

 
Joined
Jul 24, 2012
Messages
183
Location
Rome. Italy
Not the size of the camera is the matter, but if the adapter will be compatible with previous cameras, notably the E-M5.

It it were not, the whole concept of adaptation would be somehow jeopardized by the (presumably) high price of the combo.

I hope that we have better leaks, because it would be quite convenient to buy some 4/3 lenses now, while their price might rise again at the introduction of the new camera.
 

Mikefellh

Mu-43 Top Veteran
Joined
Jun 7, 2012
Messages
938
Location
Toronto, Canada
The way I had the issue of PDAF lenses on a CDAF bodies explained to me is the following:

PDAF lenses are told what position on the lens to turn to.

CDAF lenses are told to turn the focus motor on and off, speed, and direction.

Now some of the PDAF lenses have had modified firmware to make them semi-compatible with CDAF bodies, but with a loss of speed.

Would be great if they could come up with replacement firmware to make them fully CDAF, but don't know if that's possible with the type of motor and circuitry PDAF lenses have.
 

usayit

Mu-43 Hall of Famer
Joined
Jul 2, 2010
Messages
2,964
Location
Some call it the arm pit of NYC.
PDAF = there is actually a measurement / calculation to determine the distance of the subject. Throw in the differential between two points in time and the camera has enough information for more complex predictive calculations for constant/tracking AF.

CDAF = is simplistic. It makes the assumption that the highest contrast is also the point of "in focus". In its most basic implementation, it simply has to hunt for that point of highest contrast.

My assumption is that the motor inside these lenses work very similarly to servos... pulse width modularization. If a lens was designed to take in a fixed position and map that to particular pulse length, I would surmise that it would be difficult to correct this completely in firmware. I've also been told that the demand on motor speed in CDAF is much higher since its incremental changes in position relative to current driven by a sample speed of the camera.... again something that can't be fixed in firmware.

IMO (based on assumptions), the only real solution is to make the body (via adapter if necessary) speak PDAF rather than make the lens "more compatible" with CDAF. Judging from the input I've read here, only 100% functionality will be acceptable from those with specific tastes in FT lenses... the rest of us would have already moved on the micro 4/3 native lenses.
 

dhazeghi

Mu-43 Hall of Famer
Joined
Aug 6, 2010
Messages
4,381
Location
San Jose, CA
Real Name
Dara
Sorry for the thread-resurrection, but it seemed now was the appropriate moment :smile:

Looks like the folks who voted on 'less than a year' were correct. I am extremely happy to have been proven wrong on this!
 
Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Mu-43 is a fan site and not associated with Olympus, Panasonic, or other manufacturers mentioned on this site.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Copyright © 2009-2019 Amin Forums, LLC
Top Bottom