Light-field photography(bit-player.org)
bit-player.org
Light-field photography
http://bit-player.org/2012/light-field-photography
12 comments
I was one of the first adopters of the Lytro and the article hits on all the good and bad points I've found in my own usage. (I built a hackintosh VM just to play with my photos). What I really wanted was a camera that would work in any lighting or at any speed without me having to manually adjust the settings. I'd thought light-field photography was something that could accomplish this, and perhaps it still is, but for now it's still out of reach.
How would that work? Part of what makes photography a creative art is choosing a creative, personal exposure from a range of visually valid possibilities for a given scene. Once a shutter speed has been selected you're pretty much stuck with it and I'm not sure any camera design can really compensate for that after the fact.
As a hobbyist photographer who enjoys shooting in low-light and is frequently disappointed by focusing errors, I totally see the value in the concept of the Lytro (my most recent disappointment being unable to correctly focus the smoke in this shot and having to post-process to bring it out more, but still not well enough: http://www.flickr.com/photos/lusr/6830872670/).
Too bad they don't ship internationally yet, although I feel the current lack of manual controls for shutter speed and ISO (and the presumably lower-end sensor) for this first revision would probably be disappointing if I did order it.
As a hobbyist photographer who enjoys shooting in low-light and is frequently disappointed by focusing errors, I totally see the value in the concept of the Lytro (my most recent disappointment being unable to correctly focus the smoke in this shot and having to post-process to bring it out more, but still not well enough: http://www.flickr.com/photos/lusr/6830872670/).
Too bad they don't ship internationally yet, although I feel the current lack of manual controls for shutter speed and ISO (and the presumably lower-end sensor) for this first revision would probably be disappointing if I did order it.
Once a shutter speed has been selected you're pretty much stuck with it and I'm not sure any camera design can really compensate for that after the fact.
Imagine a camera that constantly records video at a very high frame rate. You can then "go back in time" to pick a moment, and the creative decision of choosing the shutter speed can be done at that point. Technically it's just a matter of blending a range of frames around your chosen moment (a fast shutter uses just a few frames, whereas a slow shutter would combine dozens of frames to create your image).
Imagine a camera that constantly records video at a very high frame rate. You can then "go back in time" to pick a moment, and the creative decision of choosing the shutter speed can be done at that point. Technically it's just a matter of blending a range of frames around your chosen moment (a fast shutter uses just a few frames, whereas a slow shutter would combine dozens of frames to create your image).
Fair point - I was a bit too rigid in my statement. High-end DSLRs already support shutter bracketing of this sort. From a technical perspective, my mid-range Nikon D7000 can take 16.2 megapixel images at a rate of 6 frames per second, which may or may not be adequate for what you're proposing.
Of course at that point the camera's image buffer fills up and the speed of your storage and the compression engine becomes the bigger technical challenge.
Whether you can take decent resolution photos on a device priced at consumer-level at a reasonable rate (without much of a delay between shots) becomes an interesting question and perhaps an interesting camera feature. If I had a Canon I would use CHDK to experiment with this.
Of course at that point the camera's image buffer fills up and the speed of your storage and the compression engine becomes the bigger technical challenge.
Whether you can take decent resolution photos on a device priced at consumer-level at a reasonable rate (without much of a delay between shots) becomes an interesting question and perhaps an interesting camera feature. If I had a Canon I would use CHDK to experiment with this.
Take a look at Lytro CTO Kurt Akeley's talk for the Stanford EE Computer Systems Colloquium (EE380) on the Lytro camera: http://www.stanford.edu/class/ee380/spring-schedule-20112012....
Has anyone tried eye-tracking to decide where to re-focus? I found it slightly tedious to look at part of a picture and to have to click there...
I was thinking that on top of video for a pretty awesome experience
That's a really great idea! You should write to the company. Maybe patent it?
Your comment made me think of "Everything is Amazing & Noone is Happy": http://www.youtube.com/watch?v=8r1CZTLk-Gk
Your comment made me think of "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." (George Bernard Shaw)
While I completely agree with Luis C.K, I also like when somebody actually propose something cool to create.
While I completely agree with Luis C.K, I also like when somebody actually propose something cool to create.
Not refocusing, but Canon made film cameras that selected the focus point based on eye-tracking. They were actually pretty damn cool. Actually, they still are pretty damn cool.
i imagine the best use is not interactive focus for the end user, but the creative possibilities of having all that source material. video with this could be amazing, one example being autofocus problems in poor lighting conditions.