After years of focus on capturing as much information in a scene as possible with computational photography off-camera (stitching, DoF stacking,...), I am excited to see the field expanding to in-camera applications.
I find it interesting that the most relevant application today is to get less DoF.
Interesting but not surprising since the control of DoF has always been one key aspect of photography. I would argue that it is probably the first moment of truth materializing the fact the photography is an interpretation of reality and not a mere attempt to capture a snapshot or reality.
The camera in phones was first about recording reality. But now, thanks in great part to DoF control, it is IMHO turning into photography as perceived intuitively by non expert users. It is one essential milestone in the perception of the positioning of smartphones amidst the real cameras.
Cheers,
Bernard
Bernard,
You dish out abstraction with the panache of a french chef serving crêpes.
It's interesting to see how long it took for the computer in the camera to catch up with the camera. But is was to be expected that the computer in the camera would swallow the camera.
If I understand the Apple keynote rightly, the variable bokeh is obtained by using a fairly coarse depth map to attempt an inverse render, determine the objects and then re-render. I guess that's more or less what I do when I draw. In which case I would say that a key feature of photography - the assumption of reality - has now been abstracted away, and the capture has become a chimera dreamed by a machine. But then we knew that already. Just look at your average auto-cleaned selfy.
So I agree with you.
What is interesting is that for you the smartphone is the smart object which can do that by itself, according to the user's desires. The big camera is not hooked up to the Google or Apple cloud, it can't add mascara and lipstick, reshape a chin, move away a frenemy or resize a waist. The big camera is as dumb as that big hairy lout who's using it, the man who doesn't understand how teens see the world - Papa.
So the smartphone is a tollgate to mediated perception. It can view what is received, but it can edit what will be sent. The big camera is just a capture device that cannot edit.
But we knew that too, from the beginning. The brains are in the net. As time moves on, anything that is not connected is insignificant, onboard software will be deprecated in favor of symbiotic local/remote coprocessing. The present state where online software is cached as "apps" is just a momentary phase. The smartphone "terminal" is mightier than the 150Mp camera "standalone device".
Which is why teens have stopped buying cameras, and they're not going to start again
Edmund
PS. I prefer crêpes to abstraction any day of the week; actually the one nice thing about winter is one can eat crêpes outside.