Lytro - A new form of digital photography

NathanH

NathanH

http://www.lytro.com/

Check out this fancy new camera, it records a whole tonne of image data within its photographs. You can adjust the focus in post, try out the sample images - left click to pull focus, double click to zoom.
 
Sort by date Sort by votes
Dreadylocks said:
Sounds fascinating. I'm surprised no one has actually thought of it before.
Click to expand...

The camera is based on the CEO's PhD dissertation. Very complicated software processing happening here, I'm quite impressed. I imagine that other's have thought of it, but this guy is the first one to actually get it into working code.

Looking through the sample images is definitely impressive for the consumer space, though I would want to have some higher res samples to play with. Some of the ones that I fiddled with didn't quite seem sharp at all planes when moving focus around. Also hard to judge the zoom function at that resolution, but still very cool.

Gave them my email address on their "reserve a camera" thing. I highly doubt I'll be able to afford one, but who knows, maybe they'll have a give away contest I can enter.

Now the real question: Who'll have a motion picture version of their sensor first? Lytro or Foveon?
 
Upvote 0 Downvote
it smells fishy to me..

would be easier to take pinhole photos (perfect focus though the entire range) and blur in software. In short I think thats sorta what going on. The lens is supposed to be made up of many individual elements or "micro lenses" that represent each PIXEL. so I'm presuming 18 million micro lenses for an 18megpixel resolution, proably wont be that high res to start with though..

Anyway, the underlying idea SEEMS to be that if a light ray (all the that photons pass through one micro lens) spreads out on sensor more then a perfectly focused light ray would, then you CAN run some math and derive how much out of focus it is and figure out what should be under the lens IF it were properly focus. OF course you loose data, which means that farther out of focus the less detail sorta explaining why it seems a bit soft in the examples.. ok, thats what I came up with anyway..
 
Upvote 0 Downvote
wheatgrinder said:
it smells fishy to me..

would be easier to take pinhole photos (perfect focus though the entire range) and blur in software. In short I think thats sorta what going on. The lens is supposed to be made up of many individual elements or "micro lenses" that represent each PIXEL. so I'm presuming 18 million micro lenses for an 18megpixel resolution, proably wont be that high res to start with though..

Anyway, the underlying idea SEEMS to be that if a light ray (all the that photons pass through one micro lens) spreads out on sensor more then a perfectly focused light ray would, then you CAN run some math and derive how much out of focus it is and figure out what should be under the lens IF it were properly focus. OF course you loose data, which means that farther out of focus the less detail sorta explaining why it seems a bit soft in the examples.. ok, thats what I came up with anyway..
Click to expand...

I'm with you, I think it calculates synthetic bokeh.
 
Upvote 0 Downvote
This seems like a very interesting idea - any word on pricing yet?

On a consumer level, I imagine the product's success will depend on the ease of use of the software as much as the quality of the hardware. For all the people that either can't be bothered or don't know how to rotate their photos on their computer, I can't imagine many clamouring for the opportunity to manually refocus everything they shoot.

Perhaps you'll be focus pulling in post-production one day, David ;)
 
Upvote 0 Downvote
I think this'll be very much a consumer product and would imagine that the difficulty of converting the process to motion pictures would fall somewhere between extremely difficult and impossible.

But for consumers I think it could be very popular. The average chump, even ones with a DSLR, just use it in point and shoot mode and don't know how to produce depth of field with the manual settings- so I think there's definitely a market for it.

When I mentioned my incredible, revolutionary, awe-inspiring idea of using 3D mapping of a film set to artificial set the lighting in post to a DoP I was berated for being 'idiotic' and I imagine that's how a good focus puller would feel about the prospect of using this technology for video.
 
Upvote 0 Downvote
I think chilepie hit it with the average consumer not bothering to rotate photos so they show up correctly in facebook (I mean heck, you can even do it FROM INSIDE facebook)

the people who CARE about DOF, will learn how to get it out of existing tech, the people who dont will be happy without DOF, and sharp photos. Easy enough with current cameras.

As the good low light sensors in the high end DSLR's start finding there way into the sub $100 consume cams its a done deal. If there is a market for this camera, its very small and wont be around for long. Thats my $0.02.
 
Upvote 0 Downvote
A little off topic, but I've asked RED to modify my epic to shoot infrared. This technology is amazing, but I think practically that we're all going to see HDR and Infrared in motion picture cameras before this is ready.

But, pulling focus in post, what I would pay for that. Wow. Do you know how much it costs to house, feed, and salary a focus puller?
 
Upvote 0 Downvote
Interesting idea there Wheat - calculating it back instead of forward.

chilipie said:
Perhaps you'll be focus pulling in post-production one day, David ;)
Click to expand...

Hopefully I'll hiring focus pullers before that day comes. ;)

Nate North said:
But, pulling focus in post, what I would pay for that. Wow. Do you know how much it costs to house, feed, and salary a focus puller?
Click to expand...

Yes, as a matter of fact I do. :P I wish it was a fixed number sometimes. <_<
 
Last edited:
Upvote 0 Downvote
You must log in or register to reply here.
Back
Top