Lytro camera review - post shot refocusing

[Peakbagr]

From today's NYT: http://gadgetwise.blogs.nytimes.com/2012/02/29/a-review-of-the-lytro-camera/?ref=todayspaper

Just the first step for the future of digital cameras? How interesting if point and shoot and DSLRs adopt or license this technology.

 

[DougPaul]

This design, while interesting, appears to have some rather significant disadvantages...

It basically works by putting a mask--basically a grid of relatively widely spaced holes (there are micro-lenses in the holes to minimize the loss of light)--in front of the regular sensor. Each individual pixel on the sensor gets light from only one hole and from only one part of the lens. (Each hole illuminates a number of pixels each at a different angle.) Software then selects a subset of the pixels to focus after-the-fact.

The disadvantage is the effective number of pixels is greatly reduced to just the number of holes in the mask (note that the camera is rated in "megarays", not final image megapixels*). I'm sure the mask blocks some light--there is no info on the ISO on the website...

*The article mentions 11 megarays but later hints at 1 MP in the final image... The website doesn't give a spec for the final pixels...

The mechanical design and user interface hardly look like it is targeted at serious photographers.

A review: http://www.dpreview.com/articles/7237351494/lytro-light-field-camera-first-look-with-ren-ng

Doug

 

[Tom Rankin]

Some serious downsides:

• Mac only for now
• Takes 1 minute per picture to upload!
• Grainy Prints
• No Filters
• Does not (yet) have 'focus on everything' mode
• $$$

This sounds like the typical first effort. Interesting, expensive, and not quite ready for prime time.

 

[Peakbagr]

Agree, as I mentioned, probably the first step that could be much more refined in the future.
For P&S cameras, forum and Facebook posts, 1Mp images would be fine most of the time.

 

[bikehikeskifish]

It made NPR this morning as well. It would require a plug-in for viewing on the internet (here, FB, etc.) in order for the viewer to control the focus, which further limits its mass appeal.

Tim

 

[nartreb]

"Focus everywhere at once" should be pretty easy to do. If they wanted to make a mass-market product that produces images (e.g. JPG), that's where they'd concentrate.

Instead they're pinning their hopes on controlling the data and how you use it. ("We'll write all the code so that other people can connect and view the images"?? BZZZT! Wrong answer! Publish the file format and let ANYBODY write the code to view the images or do anything else with them, if making useful stuff is your goal.) They're showing contempt for their customers, and customers will respond accordingly, even if it takes a few years for customers to catch on. (e.g., Facebook.)

 

[DougPaul]

I recently read that this technology is already commercially established: http://www.raytrix.de/index.php/home.html. The Lytro camera is just a toy version... (The technique has a lot in common with the technique used to make color CRTs for the last 60+ years.)

A nice discussion of the trade-offs in the Lytro: http://terragalleria.com/blog/2011/06/28/light-field-camera-from-lytro/

(Ref http://groups.google.com/group/comp.dsp/msg/e586f9bac5f9ffe1)

BTW, the author of the article on trade-offs is an established photographer with lots of nice outdoor pics on his website: http://terragalleria.com. Worth a visit or three...

Doug