a few thoughts on Lytro

(Posted originally on the AHCI forum)

http://pictures.lytro.com/ei8htohms/pictures/214663

http://pictures.lytro.com/ei8htohms/pictures/214664

Hello All,

The Lytro technology is inarguably immature, and as it stands currently is little more than a toy or a gimmick, but the promise of light field images is profound in my opinion.

With the hardware as it sits, there are at least 4 or 5 different outputs that will likely (certainly for some) emerge, only one of which is available now: the living picture (as you see in these examples). Living pictures are fun primarily for their novelty at this juncture, as there are not a great many applications where adjusting the focus point after the fact actually improves the experience. I would argue that macro photography of timepieces is such an application, but I'm pretty sure it would be more useful to have the entire image in focus rather than needing to interact with it (more on that in s moment).

One interesting way to limit the need for special viewing (interactive) software is to generate an animation of the living picture with a specified focal point shift. This can be done now with a Lytro image, but the process is laborious. It's easy to imagine this being automated in the near future once Lytro's APIs become available.

All-in-focus is a software advancement that Lytro has promised this year, and it will be possible to apply that software to images already captured. If anyone knows of another camera capable of all-in-focus macro pictures, I'd be interested to hear about it.

The last major near future output promise of the software is true 3D. Any image captured today will be able to be converted to 3D at some point in the near future but not yet specified by the company as far as I know.

I'm hopeful also that creative software approaches to the image processing will result in effectively higher resolution images. Something akin to interpolation applied to focal layers close to each other should be able to tease out a great deal more information that what is currently visible from any single layer. If this can be done to the all-in focus images, that alone would be worth the price of admission.

Another likely output that will interest some is tilt shift. I haven't heard anything official from Lytro, but once they make their APIs available this will certainly come as well.

For now, I think it's a lot of fun with a steep learning curve and I look forward to PC software availability and to my wife letting me use the camera a little more often.

_john