[Bf-committers] A new photo format

[Lars Krueger]

-------- Original-Nachricht --------
> Datum: Thu, 23 Jun 2011 22:18:54 +0200
> Von: Knapp <magick.crow at gmail.com>
> An: bf-blender developers <bf-committers at blender.org>
> Betreff: Re: [Bf-committers] A new photo format

> > To quote the reply from kalast: What uses do you have in mind?
> >
> > --
> > Dr. Lars Krueger
> 
> 
> First of all thanks all for the great explanation. I just thought it
> was really cool to be able to change focus like that and found the
> idea that you could even change the point of view very interesting. I
> guess I would have to say that my use was purely artistically
> motivated; something I thought would be fun to play around with. My
> question was also just based on curiosity. I was thinking of making
> pictures or even films where the user could change the focus and
> decide themselves what was important in a picture. I could build a

You could fake that by recording the depth map and the image and compute the defocus online. See http://developer.amd.com/documentation/presentations/legacy/Chapter05-Filion-StarCraftII.pdf for an example.

> picture with many interesting elements at each depth. The idea of
> having built in 3d was also interesting.
> 
> I see 3d as the wave of the future. There is even a hand held 3d game
> now that does not need glasses as well as many glasses based TVs and a

That's exactly the problem of all 3d content: The display technology. Until Star-Wars-like holographic displays are here, you have to decide which body part should hurt after the movie: The bridge of your nose from the heavy shutter glasses, your neck from sitting very still (handheld), your eyes from the red/green images or your wallet from the home cinema with lightweight circular polarisation glasses and projectors.

> big 3d film push. I suspect that a technology like this could really
> change the future of pictures and wondered if Blender will be ready

You can render stereoskopic images, even with proper convergence angles and correct defocus. The setup is pretty simple: An empty for the focal point and a tracking constraint per camera. The empty can also serve for the focus distance object.

> for it or not or if I could maybe even play with it now. I think the
> key to plenoptics' future will be how open the patents are for the
> normal person to use the stuff. As for cameras,  I have to say that

I don't think there are many patents around. Plenoptic cameras have been researched for 20 years, I think. The trick of this specific camera is the placement of the microlenses, which is published in the PhD thesis. Unless he patented it before publishing it, there can't be no patent on it.

> 80% of my failed pictures are failures because of bad focus; the rest
> bad light or a missed shot.

This is a problem that can be solved by money: Get an SLR, a decent set of lenses and a professional flash. It won't reduce the failed pictures to zero, but quite a lot. I saw this when I switched from compact to SLR.

Speaking of money, I don't think the lightfield camera will be cheap. The hardware in his PhD was a 16 Megapixel middle format backplate. Those things are expensive. I'm not sure if you can get a decent quality out of a APS imager, where the pixels are much smaller.

-- 
Dr. Lars Krueger


Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir
belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de