Stitching confirmed to be the only way forward...

[BernardLanguillier]

http://m.dpreview.com/news/2012/06/22/Gigapixel-camera-suggests-ways-to-offer-high-pixel-counts



Cheers,
Bernard

[Pingang]

technology can be useful and boring at the same time.....

Pingang

[shadowblade]

I want one! Preferably one with a 150-degree image circle, for ultra-wide panoramas.

It avoids the usual problem of subject movement with stitching and scanning back techniques, since the entire frame is exposed at the same time.

[BernardLanguillier]

technology can be useful and boring at the same time.....

Technology is never more boring than what you decide to do with it.


Cheers,
Bernard

[BJL]

Part of the strategy here is getting very high optical performance by using a fixed focal length lens of very high angular resolution with very large aperture to control diffraction, and a very high angular field of view. Then narrower FOV options could be achieved by cropping while still getting resolution mostly limited by diffraction (as its military sponsors will surely do with satellite mounted versions). As far as the lens, this is a very large-scale version of the strategy of the Nokia PureView! But the clever idea of being able to adjust focus and exposure time for each piece of the mosaic separately could overcome the fundamental limit on DOF at very high resolutions when recording the image the normal way, at the focal plane of the lens.

I wonder if more modestly scaled versions could become a large format "landscape" camera option in the future.

[BartvanderWolf]

But the clever idea of being able to adjust focus and exposure time for each piece of the mosaic separately could overcome the fundamental limit on DOF at very high resolutions when recording the image the normal way, at the focal plane of the lens.

But then it's nothing new. Calling it clever would risk overstating its level of innovation, since (and he was not alone) Roger also  knew the benefits of such an approach.

The camera was used in autofocus mode on the center sensor so the focal point of each image was different and optimum for that frame.

And I use the same approach for very wide panoramas or 360 degree VR shots when exposure and whitebalance is quite different with some shots being back-lit and others front-lit, in the same scene. It just requires a good blending engine to even out the luminosity/WB transitions.

Cheers,
Bart

[theguywitha645d]

I wonder if more modestly scaled versions could become a large format "landscape" camera option in the future.

I doubt it. The requirements for a spy camera and one for pictorial photography are very different. Resolving power is not the end-all-and-be-all of photography. What is the point of having information that cannot be either displayed nor perceived. And I would imagine this kind of system far more expensive and cumbersome.

It is an interesting camera and an interesting solution. If it ever proves to be a useful system--and from these results, it might be less than ideal--I would imagine it simply be employed in very technical imaging like remote sensing.

[John Nollendorfs]

The D600E with 60m MicroNikkor, used vertically, panned and stitched delivers pretty incredible 3-400 megabyte files. Do double deck shooting and stitch--you probably have gigabyte picture files! ;-)

Pretty sweet, for a $3300, that's not a special purpose camera.

[BJL]

But then it's nothing new. Calling it clever would risk overstating its level of innovation, since (and he was not alone) Roger also  knew the benefits of such an approach.
And I use the same approach for very wide panoramas or 360 degree VR shots ...

I should have said that the clever idea is allowing mosaicing with these local adjustments in combination with recording the whole image at the same time, as opposed to stitching of images taken sequentially. That seems to require the ability to reconstruct the image from data recorded well behind the focal surface, and perhaps also a lens whose focus is on a highly curved surface, not a plane.

[BJL]

Resolving power is not the end-all-and-be-all of photography. What is the point of having information that cannot be either displayed nor perceived. And I would imagine this kind of system far more expensive and cumbersome.

I am also not betting on it ever coming in a "civilian" version, and clearly this extreme resolution is only relevant to a tiny fraction of all photography, but there are photographers who use stitching to get far higher resolution than any current camera can get in a single shot (including at least two contributers to this thread), so I would not totally rule out a market for it.

Of course it would be expensive and cumbersome, but maybe not enough to deter the most zealous. Consider for comparison the cost of top of the line DMF system, or the bulk of the 10"x8" and larger view cameras that some have hauled up and down mountains for the sake of landscape photography. Note also that this prototype is for a project aimed at even more extreme resolutions like 50GP, so considerable downsizing might be possible for a camera of a mere few hundred MP, using a smaller number of "sub-cameras" and a smaller lens.

[John Camp]

At what point does air limit resolution?

[BartvanderWolf]

At what point does air limit resolution?

Hi John,

It varies, with distance/angle/and aerosol/dust content.
Ground surface temperature and air temperature/turbulence also make a difference depending on exposure time.

Cheers,
Bart

[Ray]

At what point does air limit resolution?

At billions of points I imagine. Not only the air, but the fine particles of dust and carbon.

[BJL]

At what point does air limit resolution?

There is some discussion of this at the gigipixel camera project site:
http://www.gigapxl.org/technology-atmosphere.htm
and it makes me skeptical about getting 50GP resolution in terrestrial photography. But it probably helps for the camera to be deployed above the atmosphere, which is probably what DARPA intends for the 50GP version.

[OldRoy]

The D600E with 60m MicroNikkor, used vertically, panned and stitched delivers pretty incredible 3-400 megabyte files. Do double deck shooting and stitch--you probably have gigabyte picture files! ;-)

Pretty sweet, for a $3300, that's not a special purpose camera.

No, indeed it isn't, the D600E is a Samsung mobile phone. And you can definitely buy one at a better price than $3300. Who makes that F mount adaptor for it?
Or do you have some advance info about the rumoured "affordable" FF Nikon model?
Roy

[John Nollendorfs]

No, indeed it isn't, the D600E is a Samsung mobile phone. And you can definitely buy one at a better price than $3300. Who makes that F mount adaptor for it?
Or do you have some advance info about the rumoured "affordable" FF Nikon model?
Roy

Pardon my bad typing Old Roy, of course I meant D800E!

[John Camp]

Hi John,

It varies, with distance/angle/and aerosol/dust content.
Ground surface temperature and air temperature/turbulence also make a difference depending on exposure time.

Cheers,
Bart

The reason I ask is, I have homes in Los Angeles and Santa Fe. I live on a hill only a few miles from the San Gabriel mountains in LA, and a lot of days, I have trouble picking out the major features because of smog/dust. The same can be true in Santa Fe, which is situated in one of the harder deserts in the US. The days can be sunny, but the atmosphere is often thick...Given the slight differences between a D800 and a D800e, was I dumb to hold out for the e? Would I ever be able to detect the differences in resolution, given where I shoot? Would I ever see the effects of diffraction in landscape photos?