Olympus Working On Technology That Would Allow Handheld Sensor Shift Images

Olympus Working On Technology That Would Allow Handheld Sensor Shift Images

Camera manufacturers are still cramming ever more pixels into their sensors. Canon has recently announced that it is releasing a 50MP full frame camera which has left many speculating about its technical limitations due to sheer physics. One solution of course is to move up to a larger medium format sensor but prices can be prohibitive for many. Olympus, in typical Olympus fashion, decides to tackle the want for more pixels in an innovative way.

With the release of the OM-D E-M5 ll Olympus has introduced sensor shifting technology in the camera which allows the onboard 16MP sensor to capture 8 frames in sequence and stitch them together for one much larger image; which it does by shifting the sensor ever so slightly. In the case of Olympus it would be 8 images that amount to 40MP or 64MP in raw+jpeg mode, and Olympus claims the resulting image will rival the quality of single capture sensors of equivalent pixel count.

https://www.youtube.com/watch?v=toYpuK7WaTY

While this sounds great on paper there is one major caveat to consider. The 40MP image would be captured over a 1 second time span and the 64MP image over a 2 second time span. Product photographers rejoice, but for any real world applications, this is quite a handicap. The camera would need to sit perfectly still for the duration of the capture if the photographer wants a usable image.

Fear not because Olympus General Manager Setsuya Kataoka has said in a recent interview that they are working on technology which will alleviate that concern. They want to speed up the process of the sensor shifting so that the photographer can remain mobile and create these stitched images while the camera is handheld. Just how fast are they thinking? The plan is to make these captures happen in under 1/60th of a second!

This is pretty exciting news because if the application of the technology proves practical, it could alleviate the physical limits imposed by sensor size, and we can introduce higher pixel counts while retaining a decent signal to noise ratio. This is all very preliminary and as of yet there is no word on when the technology will be implemented, but there are innovative things brewing over at Olympus, that much is certain.

[via DPReview]

Peter House's picture

Peter House is a commercial fashion photographer from Toronto, Canada. He shoots over 10,000 pieces of clothing every year for a variety of lookbooks. Clients range from small local boutiques to international brands such as Target, Winners, and Sears. In addition to that Peter runs one of the most popular rental studio's in the Toronto area.

Log in or register to post comments
11 Comments

Interesting idea in theory! I don't fully see the need for 50 megapixel, etc. however. I use a Nikon D800 and I often think that the 36MP is unnecessary and would prefer if there was an option to shoot RAW at 24MP for instance (that would be a feature I would be more interested in!).

I feel the strive for more and more megapixels is going to lead to larger file sizes, longer manipulating photos, etc. without necessarily improving quality- I would imagine the majority of lens' resolving power isn't optimal for a sensor of 50MP. I presume the simulated 40/64MP sensor would possibly not suffer to the same extent from this problem though, which is why I said it was an interesting idea...I'll withhold judgement for now! Still seems a bit gimmicky to me though, would much prefer better dynamic range and high ISO performance.

Fantastic observation regarding the resolution of lenses. As for dynamic range and better high ISO performance, to an extent those would benefit from lower pixel density, thus further development in this area could in fact pave the way for that. :)

The process actually improves DR as well. This is essentially the superesolution process. You can do this with any camera using a program called PhotoAcute:
http://photoacute.com/studio/

You handhold and fire off an array. The software does the same thing the Olympus is doing in-camera. Pentax alludes to being able to do the same thing or something along these lines with it's forthcoming 36 meg FF body.

Obviously the process is limited to relatively static scenes. In studio product and portrait photography the Olympus in-camera process can be a useful option if needed.

yeah. Canon have had Raw small a long time now. guess that's one of the reasons press photographers like Canon. I heard that the D810 has a similar feature, but with some weird limitation or something.

I now have to get a Rolleiflex TLR. Thank you Olympus for the great advertisement!

Actually, this message was brought to you by a Nikon shooter.

Cheers!

"Product photographers rejoice"? They'd have to replace their flash lighting with continuous lighting, or does the camera wait until the flashes are loaded? Do you get a RAW file that can be processed in "mainstream" RAW converters such as Capture One or Lightroom?
In the end you need a lens that resolves that much MP on a relatively tiny sensor. Hasselblad has done the same thing before with their H4D-200 and H5D-200.

Actually, sensor shift (with all the limitations mentioned here) is the best way to squeeze out more from your lenses. Remember, the pixels don't shrink, they are shifted. Hence a lens that resolves 16MP is perfectly fine for a sensor-shift-stitch 50MP or whatever. So the diffraction limit remains the same.
I do use a H5D-200. It works (sloooowly ...) ;-)

yes there is an option to make the sensor wait for the flash to reload,betwen everyshot, and i think a camera raw pluguin will be coming out for the 64 mp raw

yeah. This sure got me excited.

Nice feature, great for product or landscape, couldn't see it handling portraits though