Lytro ?

[FoodMan]

I wonder if this would be interesting for Photogrammetry 

https://www.lytro.com/camera/#

f/

[Wishgranter]

10 mil rays - 480 x 480 pix = 0,3 Mpix

40 mil rays =  approx 1 Mpix ( 40 rays per pixel )

So marketing is one thing,  real experience is other :-D, but im could be wrong........

[FoodMan]

Hey Thanks Milos...

I was not sure , but you make it clear it is very small... ah well, too bad.. 

f/

[alexb]

to low rez?

[FoodMan]

yes too low res... only 1 m pixels... 

[Wishgranter]

lores video..... but... DEPHT info in realtime ?? :-D  https://www.youtube.com/watch?v=ZE60HihBYWw#t=75

[FoodMan]

well that's the beauty of it, but as you said... toooo low res.. hehe

[scanlab.ca]

Was thinking about this myself for the last couple of days after the announcement of their second generation camera.

https://www.youtube.com/watch?v=CevVGBZgdts

Thing is that we do have a whole lot of resolution with traditional DSLRs, but majority of it goes to waste since things go in and out of focus.

If Lytro camera can be used with focus stacking that can be a whole other ballgame, because the usable resolution of that camera may actually be more useful. But at f/2, not sure.


From what I managed to find out about First Gen Lytro:
Sensor array is 330?390 micro-lenses at 10pixels each
Capturing 3280?3280 RAW
Lytro software exports 1080?1080 interpolated images


If current stats are true, then I think the Second Gen will be:
Sensor array of 1320?1560 micro-lenses at 10pixels each
Capturing 13120?13120 RAW
Lytro software exports 4320?4320 interpolated images

Or something similar.

But, its the future, I think.

[FoodMan]

sure it has a great potential... I was thinking especially for small objects, where dof can  be an issue..

f/

[scanlab.ca]

sure it has a great potential... I was thinking especially for small objects, where dof can  be an issue..

f/

Exactly.

Check out this video:
https://www.youtube.com/watch?v=1XAgkJ5ypdU#t=29

31 positions for Focus stacking.
Why not replace it with a single Lytro and "fogetaboutit".

[Wishgranter]

hmmmm Scanlab thax for detailed info on sensors....

[FoodMan]

Interesting Video ScanLab..

I tried focus stacking a few times, but it takes so long and I would have need a rail stacking system.. mine was using lens focus... not too great..

f/

a good kickstarter for a Special camera dedicated to photogrammetry would do wonder.. eheh

[scanlab.ca]

If current stats are true, then I think the Second Gen will be:
Sensor array of 1320?1560 micro-lenses at 10pixels each
Capturing 13120?13120 RAW
Lytro software exports 4320?4320 interpolated images

Or something similar.

I just realized that my Second Gen Lytro projection is too generous, quadrupled things "accidentally".

The proper math should go more like this:

1080 ^ 2 = 1166400 Rays OR 11.664 Megarays

In one of the articles they mentioned that they've added 4x resolution, which goes something like

( 1080 ^ 2 ) * 4 = 4665600 Rays OR 46.65 Megarays (actually less, or they would've stated 46 Megarays )

sqrt( ( 1080 ^ 2 ) * 4 ) = 2160 pixels wide by 2160 pixels high

So, there. This could mean that you may need to take 2x OR 3x more shots, depending on your current setup.


That's not all, pixel pitch of Lytro is 3x less than of regular Nikon/Canon making it less light sensitive and more noise prone. But, because of the micro lens design you can probably reduce noise computationally having all that light incident information at hand.

Anyway, just wanted to correct myself on that previous wild estimate.

[alexb]

looking at the exemple of the image quality in the axemples, considering the result is a bad quality 2500*2000 image. what would be the advantage of the lytro compared to a good reflex closed a 22 ??