Computational Photography

[jdlaing]

Advertisement (gone after registration)

10 hours ago, scott kirkpatrick said:

Has anyone spotted any results of Leica's statement around midyear 2019 that a new "competence  center" in computational photography  with about 40 people would soon be staffed and beavering away somewhere in Silicon Valley.  Leica has made an investment in Light, Inc, and has a partnership with Huawei, leading to the P20 and P30 products.  Huawei, which has some problems operating in the US, gets good reviews for their 3-lens interpolating cameras.  Light, which made an impression with their Light 16 16-lens wonder in 2017 now seems to be emphasizing their newest vision system, intended for smart cars.  (And they raised over $100M from SoftBank.)  Qualcomm and Nokia are active in multi-lens cameras as well.

So where does Leica fit in and when will there be something to talk about?  Does a "competence center do research, develop products, or something in between?

Leica has invested, as some other companies have, in Light. 

[adan]

If one's goal is simply to put a ball into a hole, there are lots of better ways of doing it than hitting the ball with a crooked stick.

But what happens to, say, "The Masters" - when every player reaches the clubhouse with a score of 18?

I'm not sure who would be the bigger "control freak." Me with my manual RF and occasionally difficult lenses - or someone assembling their pictures computationally?

Maybe H-CB was right - throw away the camera and go back to the pencil.

Although someone at the gallery was showing me his AI-computed "paintings" just a couple of days ago. Rather boring printed flat on metal - but with 3D printing to reproduce the "brushstrokes"....?

I'm glad I experienced photography when the technology was "diverse" - tens of film types, a dozen different formats that could not easily be substituted for one another, alternative processes (silver, gum, carbon, C-41, Cibachrome, dye-transfers, cyanotype, ambrotype/tintype), lenses from Leica to Lomo, each with their own characteristic limitations - and thus their own unavoidably unique character.

Of course that can all be simulated in a virtual world - we can create a "75 Summilux" or a "Hasselblad SWC" algorithm.

But I like the authenticity of a real SWC image, where there is a black border with notches in it, and a specific DoF for a specific FoV - because that particular device was built that way in the first place. And is in B&W not because it was computed from a color image, but because I made a command decision - before even leaving the house - that my cameras would be loaded with B&W film that day. Nor colorized because I changed my mind later.

Not because someone selected a menu item labelled "Hassy SWC_border_f16_B&W." (BTW, such a computed image would be rejected by the press contests I contribute to.)

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

I'm not an old fogey - change and the future are inescapable, and humans will develop any possible technology eventually. Change, in fact, makes for more interesting reportage photography - humans are always doing the darndest new things, and that makes for "new" pictures even with "old" gear.

But I have a subtle worry that a technological sameness - everything reduced to 1s and 0s - may produce a gray social conformity and leveling that the authoritarian regimes of the 20th century could have only dreamt of.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama