Previous news story Next news story

KaleidoCamera teaches your DSLR new light field tricks

Jul 22, 2013 at 19:48:54 GMT

Print view

A prototype for a new DSLR add-on is poised to bring plenoptic capabilities to consumer cameras. The KaleidoCamera is designed to sit between a standard DSLR's sensor and lens. A diffuser splits light passing through the lens into nine different beams, each passing through a filter before it reaches the camera's sensor.

Depending on the configuration, each beam of light can be filtered to isolate particular colors, or capture a scene with nine different tone curves to create an HDR image in post-processing. With a slight adjustment, the KaleidoCamera is capable of light-field imaging. Lytro cameras have already made this capability available to consumers, but the KaleidoCamera would be the first device to work with a consumer's existing camera.

This is a picture showing the prototype KaleidoCamera fitted to a Leica medium-format DSLR.

Plenoptic cameras split the incoming light based on the angle it arrived from, meaning depth information is captured about the scene. The more images the scene is split into, the greater the depth information captured, but the lower the resolution of the final image.

Two configurations of the KaleidoCamera - the uppermost uses filters for selective color or HDR photography, and the lower configuration uses a light field design to enable plenoptic photography.

Applications extend beyond interesting photo experiments and post-capture focusing - the KaleidoCamera could see use in "scientific imaging, industrial quality control, remote sensing, computer vision and computer graphics," according to a paper published by the prototypes creators.

Video: KaleidoCamera for HDR, Multi-spectral, Polarization & Light Field Imaging

Via: Gizmodo Australia, Source: Saarland University

OTHER NEWS LIGHT-FIELD LYTRO

Comments

Total comments: 46

I noted how the proposed resolution was stated as less than HD. I assume it isn't much better than the 1K resolution used by the Lytro Light Field camera. I consider this extremely low resolution to be a major handicap that limits the usefulness of this technology.

Comment edited 2 minutes after posting

0 upvotes

For years people were happy with postcard size prints from 35mm film cameras and 1mp makes a nice postcard size print. Going to a birthday party and just want to be able to make postcard size prints without worrying about getting everything in focus and this would be great. The problem with Lytro is they made it all closed source, with this you would still have access to the raw files and other vendors could supply software (adobe or even rawtherapee). You also have to use their camera. With this you can use any camera.

1 upvote

Will it work with mirrorless slr's?

1 upvote

Interesting; it's rather like the old stereo mirror rigs gone wild -- or perhaps I should say gone to the other end of the lens. The diffuser is problematic as is the critical alignment of the image multiplier discussed in the paper's section 5.1, and there is a substantial loss in resolution (inevitable with multiple copies of the image being separated out) which I think would discourage use by most DPReview readers, but overall a very nice piece of work which could lead to some interesting devices. Being able to use conventional lenses and optical components that are relatively large (compared to things like sensor-stack microlenses) is a big plus.

0 upvotes

That's brilliant! Such tools will help to proceed to mature digital photography. Today's digital photography is basically replacing film by a sensor, combining it with abrasive mechanical shutters and loudly slapping mirrors (if you want the best camera technology available). With those camera anachronisms we produce loads of 2D images we've seen since ages. Of course, we all still love to do this, me too, and there are still plenty of new, wonderful images produced that way. But I am pretty sure that lightfield/plenoptic photography is the first real digital photography, since its images contain one dimension of information more than any classic print.

Comment edited 50 seconds after posting

1 upvote

Your statement "Today's digital photography is basically replacing film", might have been true almost a decade ago. These days, the replacement has already happened. Kodak is bankrupt and even Fujifilm is discontinuing entire lines of film. It is fait accompli.

0 upvotes

Shows what can be done with enough time, money and processing power.

0 upvotes

Very interesting is quite the understatement!

One thing to note though is, as you'll see if you look at the comments on YouTube, the proposed design will actually be in the 20 to 30 centimetre range.

1 upvote

Very interesting.

0 upvotes

more toys for us

0 upvotes

HDR, refocussable video and slide shots without a slider.
cool stuff in the pipe for later

0 upvotes

Pretty damn clever.

0 upvotes

Lytro just got pawned.

9 upvotes

Not quite, Lytro delivers living pictures and tools for manipulating and playing with after the fact and all that garbage about printing Lytro images at a max of postcard size is wrong, I have gone to 8x10 without issue.

That being said, this does look cool and I can't wait to get my hands on it.

0 upvotes

I think soon or late we will forget about clumsy lenses. The Photography will suffer the same fate as the biological vision. The eye, a very imperfect optical engine, coupled with the vast computational power of the brain, continuously scans the space and gives a sharp image with huge dynamic range.

2 upvotes

Computational Photography

I invite everybody to have a look at a lecture given by one of this project's authors last year:
http://www.mpi-inf.mpg.de/departments/d4/areas/giana/Teaching/ComputationalPhotographySS2012/
It has some interesting ideas about how the art of photography may change in the future, esp. that in the studio. E.g., scene illumination may be added after the shot etc. ...

0 upvotes

With todays huge megapixel sensors, dividing one into 1/9 sections still give you printable results unlike the Lytro. Cool beans if you have an SLR, but I wonder how much the lens/software package is going to cost you. :)

Comment edited 4 minutes after posting

0 upvotes

Never had a problem printing my Lytro pictures, already 8x10 without issue. The key for Lytro, though, will be to get ahead of this - their camera and its pictures are not infallible but certainly a good start. Now on the other side of the coin, I haven't seen any mention of the pictures coming out of this attachment being living pictures.

0 upvotes

Sweet stuff :)
I know the institute which did it and most of their research is top notch, actually.
I'd say the main application isn't what they describe (the conversion of a high end consumer camera into an industrial one). Even though it may be an important application for German machine engineering.

I'd say the main application is as a research and prototype device to path the way to lens-array based smartphones (which most likely will produce an array of 3x3 images, each of at least HD quality, too). Interesting for high end smart phone lens makers like Zeiss.

A 3x3 lens array-based smartphone reduces the crop factor by 3 (such as from 4.5 to APSC) and brings smartphones on par with dSLRs. Esp. as the array allows for parallax-accelerated autofocus (before shot) and plenoptics-like focus-tune after the shot, beating phase detect AF. This device from Max-Planck Institute will help explore stuff like this.

Comment edited 8 minutes after posting

7 upvotes

The German company Raytrix offers high-end plenoptic cameras for industry since 2010.

0 upvotes

Sweet! Something new for camera geeks to complain and argue about on the Internet!!!

6 upvotes

Fascinating

1 upvote

Yes but can it do 3d?

2 upvotes

Yes, but will it blend?

10 upvotes

Yes

0 upvotes

Before the technical discussion, il get my coat.

1 upvote

Fascinating.

BTW - inventions that nobody have asked for? I would guess that a vast majority of the inventions are not "asked for". How many asked for digital cameras in the film days?

All this said - maybe it will be nothing. Just another fun thing.

But .... I assume that something will happen with digital photography ... something you don't expect. And not is asking for. I am quite sure of it.

4 upvotes

Well back to the turn of the XX Century, a lab technician was testing the consistency of a resin to be used on car tires. In the end he discovered some moisten residual, not so hard to be used on tire's manufacturing, but the chewing gum was born.

1 upvote

the best inventions can be failures to achieve something completely unrelated.

The real problem is finding practical applications for stuff like this.

0 upvotes

Based on what I'm reading, it appears to turn your camera into 1/9th the pixel output too. Sooo...a 36MP D800 = 4MP camera?? Am I reading their site correctly??

0 upvotes

I have not read it. But ... that is a necessity. You can't just invent data. So ... if you want data from several angles ... it has to be taken from resolution.

Much better than Lytro though. There you lost a factor of much more.

Comment edited 39 seconds after posting

1 upvote

Which is probably why they are showing the prototype on a medium-format camera i.e. Leica S2 is 37.5 MP

So while it will turn your 36mp camera into a 4mp camera, it will be a 4mp image with 9-image HDR capability - with no tripod required for multiple images, and no software-based correction for varying camera position between the nine 4mp photos.

Stitch a few of those together, and you will have a nice HDR panorama. :)

Comment edited 7 minutes after posting

3 upvotes

Roland: I agree, it makes sense, but I would have thought, at least for the HDR functionality, that a 4-to-1 ratio would be ideal and a 9-to-1 would be a tad bit of overkill. But then I'm very binary-minded, so maybe its just me.

And yeah, the Lytro was way too much compromise.

Comment edited 2 times, last edit 2 minutes after posting

2 upvotes

I agree on the 4 [or 5] exposures making more sense than 9, but i can only assume the other uses - like color filters - make better use of 9 images.

As my fisrt dSLR was a 6mp D50, getting 6-images into a 6mp HDR would also be interesting.

1 upvote

This is an argument for a 100 megapixel high density sensor. Now just add a speed booster! ;)

1 upvote

The Lytro starts out with lower MP and has higher decimation to provide more powerful focus capability. If they build a new one with Nokia style high MP capability, they will be getting somewhere.

0 upvotes

You don't really need it for anything, so don't pay anything for it. Another "must have" invention that nobody had asked for.

2 upvotes

We never needed to fly either.

If we only ever invented stuff we 'needed' we wouldn't have had half the stuff we currently do.
Heck, we probably would still be stuck in the sea as a single cell organism.

1 upvote

"Another "must have" invention that nobody had asked for."
You mean... like cameras?

0 upvotes

It is a sophisticated optical filter, but placed in the "wrong" end of the lens!

Comment edited 31 seconds after posting

1 upvote

wrong end? there are plenty of lenses that place the filter between the lens and the body..

0 upvotes

The problem with putting this particular filter in the wrong end is that you need to handle the camra-lens communication and also sometimes screw drive.

0 upvotes

Maybe they can make ones for prime lenses in front of the lens but that would require a lot of R&D money and bother. Start with a universal is a smart business move.

0 upvotes

It has to be behind the lens.

2 upvotes

Have I been putting my teleconverter at the wrong end of my lenses then?

0 upvotes

Wrong endsesss?
I say assses and pussiess to be very different animals, they are...

0 upvotes

Total comments: 46

Editorial content

Product database

Community


Miniature wide angle lens under development at UCSD Sep 27, 2013	Editorial: Why I can't stop taking iPhone Panoramas Sep 24, 2013	Smarter cameras may help bring autonomy to space robots Sep 12, 2013	Full 360-degree camera rig takes your 3D portrait with 60 DSLRs Aug 18, 2013

	Nikon 1 AW1 Sep 19, 2013
	Fujifilm X-A1 Sep 17, 2013
	Olympus OM-D E-M1 Sep 10, 2013
	Leica C (Typ112) Sep 8, 2013
	Nikon Coolpix P7800 Sep 5, 2013
	Nikon Coolpix S02 Sep 5, 2013
	Sony Cyber-shot DSC-QX10 Sep 4, 2013
	Sony Cyber-shot DSC-QX100 Sep 4, 2013
	Sony Alpha A3000 Aug 27, 2013
	Sony Alpha NEX-5T Aug 27, 2013

Okay Fuji, so now that the experiment is over....	107
D610 to be announced within 48 hours (as of 10/5)	87
How many shots do you take to get a shot?	83
While everyone ooohs and aaaahs over the ...	78
More leaks about K3 specifications	76
Anyone for Sunday Scapes (2013-10-06) ?	66
After 3 weeks & thousands of shots: NEX-6 + ...	57
E-M1 Trouble focusing on moving objects	55
The Sunday BIRD Volume 9, Issue 15 October 6, 2013	49
Counterfeit lens caps from Amazon	45
GX7: Extended ISO125 Image Degradation?	43
E-M1 as E-5 successor - first week experience	42
Help deciding between Sony NEX-6 or Olympus ...	36
X-pro1 vs x-a1 (or is it X-A1 vs X-pro1) using ...	33
I'm going broke buying E-M5 Eyecups!	32
Sensor Technology (The Software Engineer Way)	31
Shutter shock verdict is not present here. ...	30
Worth selling my Canon DSLR to move to Olympus ...	30
Five Worst Clichés About Photography	26
Why is new ff nex E mount instead of A mount?	26

Your email address:
Your name: