Previous news story    Next news story

One-shot gigapixel camera offers a future beyond flat sensors

By Richard Butler on Jun 22, 2012 at 17:24 GMT

A gigapixel camera developed for the US Department of Defense's research agency (DARPA) provides an insight into the challenges that will need to be overcome to offer super-high-resultion cameras. A team from Duke University, Durham, North Carolina, has described its 960 megapixel (0.96 gigapixel) 'AWARE-2' camera in a letter to scientific journal Nature. The team says small, efficient electronics are the key to being able to miniaturize the camera, which currently sits in a 0.75 x 0.75 x 0.5m frame.

The team developing the AWARE-2 have considered scalability be one of the key considerations while designing it. Conventional camera designs will remain limited to megapixel resolutions because the small apertures they use are limited by diffraction softening. Scaling the design up and using larger apertures ends up limiting the system because lens aberrations increase, so you remain restricted to the megapixel scale, says team leader Professor David Brady.

This led to a design that arranges a series of cameras in a hemispherical arrangement, pointing at a single, spherical lens that the team have dubbed the 'gigagon.' The use of a single lens avoids the cost and complexity of having specialist optics on each sub-camera, while the curved design scales more easily than a flat array of cameras. The team believes this approach would continue to work for up to 50GP cameras.

A diagram showing the hemispherical arrangement of the sub-cameras, and their relationship to the 'Gigagon' main lens (top right).

The current design is made up of 98 individual 14MP sub-cameras, with focus and exposure set individually for each sub-camera. All the sub-cameras are exposed at the same time, meaning you don't have the problems of movement that occur in conventional, scanning gigapixel images. HDR techniques are then used to combine all the different 8-bit exposures into a single 32-bit file, which is then tone-mapped back to an 8-bit image that can be displayed.

 An image highlighting the contribution made by each sub-camera.

The AWARE-2 weighs 93kg and captures a 120x42 degree field of view but  the current design allows for a maximum of 220 sub-cameras to be installed, (giving a 120deg circular field of view). With all the cameras installed the overlap between sub-camera's images and elimination of poorly illuminated data would cut its theoretical 3 gigapixel capability back to around 2 gigapixels - around the number that its 16mm aperture would diffraction-limit it to, the group says.

The team says the image quality of the current camera is reduced by the use of injection-molded plastic relay optics in front of each sub-camera but that they believe this restriction can be overcome with the use of high-refractive-index plastics to improve this performance.

However, it's the electronics that the group says currently prevents the camera being made smaller - the optical system accounts for just 3% of the camera's volume - with the rest of the space taken up by the associated electronics and cooling required to dissipate the 430W expended every time the camera takes an image. With smaller, more efficient electronics, hand-held gigapixel cameras may become an everyday reality, they say.

An example image, shot in Seattle, showing various crops taken from a 0.96GP image

Comments

Total comments: 132
12
ashish7t9
By ashish7t9 (Jul 2, 2012)

Wow, great to see the advancement of the technology. Cool....

1 upvote
Harveydad
By Harveydad (Jun 27, 2012)

How exciting!!! I wish I were 40 years younger..........

0 upvotes
robogobo
By robogobo (Jun 27, 2012)

THe coolest thing about this tech is the possibility for perspective correction by adjusting the angle of the sub-cameras. The challenge will be in redefining "RAW" output, since the image is software stitched and compensations must be made.

0 upvotes
JackM
By JackM (Jun 26, 2012)

looks like they need to get the exposure consistent across all the sub-cameras.

Hmph! I'll wait for the Mark II. ;-)

0 upvotes
Betopmodel
By Betopmodel (Jun 26, 2012)

When you arrive here in Brazil, is a fortune because the taxes are absurd!

0 upvotes
Phodeaux
By Phodeaux (Jun 26, 2012)

This camera is rumored to be included in the next iPhone to be announced in the fall. Apple will release the 2 gigapixel version though, not the sub gigapixel shown here.

1 upvote
Alizarine
By Alizarine (Jun 26, 2012)

Can't help but think of it as the new competitor for Olympus.

"The Mountdoom Sauron G-1, the world's first commercial 35mm gigapixel camera, with its accompanying spherical kit lens the Gigagon 35mm f1.8"

lol... everything has to be sold eventually on "everyday" level, really?

0 upvotes
W Sanders
By W Sanders (Jun 25, 2012)

Question: Is it cheaper than an M9?

Comment edited 19 seconds after posting
2 upvotes
RDMPhotos
By RDMPhotos (Jun 27, 2012)

Probably, I doubt it would retail even close.

0 upvotes
kb2zuz
By kb2zuz (Jun 28, 2012)

If it was developed for the US DoD, the price tag should be around $5 million.

0 upvotes
Litebright
By Litebright (Jun 25, 2012)

Let's see, that's 20 MP per KG, uses more power than my TV, and produces enough heat to fry 1 egg with each picture taken. I just realized if I use bailing wire on the one I'm building, it will cut all of those numbers in half --or more.

0 upvotes
NigelMoore
By NigelMoore (Jun 25, 2012)

" small, efficient electronics are the key to being able to miniaturize the camera". No s**t, Sherlock!

Comment edited 12 seconds after posting
3 upvotes
PaulSnowcat
By PaulSnowcat (Jun 25, 2012)

This is really impressive. but I guess the uses of this camera are scientific and security mainly... Still... The main complexity of the lenses in our cameras is making the DOF plane to be actually a plane, not a half-sphere. If a camera can accept a spherical DOF, using a spherical sensor, maybe the lenses will become drastically cheaper...

Look at Tamron 17-50 - it has spherical DOF and it's cheap :) Not too good for current cameras, but for a camera with a spherical sensor that would be OK.

2 upvotes
historianx
By historianx (Jun 24, 2012)

what's next, Exapixels?

0 upvotes
kb2zuz
By kb2zuz (Jun 28, 2012)

I'm holding off for a Yotapixel camera.

0 upvotes
MP Cevat
By MP Cevat (Jun 24, 2012)

Produces good 4x6" prints. And 4x6 meter too.

2 upvotes
huyzer
By huyzer (Jun 24, 2012)

So that's how a fly, mosquito, dragonfly, insect, etc. sees the world with all those lenses. :P I don't know. I'm guessing. :D

0 upvotes
time traveler
By time traveler (Jun 24, 2012)

I hope this camera can shoot Full HD video and work with my EF-mount lenses (at least with adapter). Oh one more thing, I need a bigger camera bag... :)

0 upvotes
KoKo the Talking Ape
By KoKo the Talking Ape (Jun 24, 2012)

Definitely, a bigger camera bag. I would also definitely need that carbon-fiber tripod I've been eyeing... :-)

0 upvotes
Artak Hambarian
By Artak Hambarian (Jun 23, 2012)

The principle is not clearly stated. Looks like a larger format camera, but with a composite sensor. The only real difference is the ability to control the exposure at a given camera/area. How the lens wrks at such a wide angle is mistery yet - probably the trick is in the innovative lens? Still thinking.

0 upvotes
ScooterNC
By ScooterNC (Jun 26, 2012)

It uses a spherically symmetric lens to form an intermediate image which is curved (with only a 4-element lens), it then samples this curved image with the hundred-ish micro-cameras (basically small relay lenses like an eyepiece of a telescope)

0 upvotes
Max Thunder
By Max Thunder (Jun 26, 2012)

That's it, just like taking a composite picture of the world reflected on a large curved mirror with 220 cheap Coolpix L24, then rebuilding the image in Stitcher... :)

0 upvotes
Trollshavethebestcandy
By Trollshavethebestcandy (Jun 23, 2012)

Eye of Soron?
I will only buy it if it's made by Olympus. I'm a label loyalist.
Can they make it pocketable? ;)

2 upvotes
Funduro
By Funduro (Jun 23, 2012)

If this is out in the open, the camera DARPA is currently using and keeps secret is ten times better.

3 upvotes
Paul Farace
By Paul Farace (Jun 23, 2012)

We can get very good images of the outside of Russian and Chinese military sites and missile bases with this ... as well as mud-brick huts of Taliban strongholds.
The Russians and Chinese on the other hand will just obtain every one of our secrets from WikiLeaks... And then the Japanese will build a Canon Rebel equivalent so that Soccer Moms and Hockey Dads can photograph their little superstars at play with Giga-Gazillion-byte size files that they will try to email to relatives instead of a 45 kilobyte size photo file.
That is why I need therapy and a bottle of Jack Daniels... :o)

7 upvotes
azmish
By azmish (Jun 23, 2012)

I suspect you might have needed therapy prior to this announcement . :)

2 upvotes
maboleth
By maboleth (Jun 23, 2012)

Yeah, Paul is the person who buys everything US mass-media delivers. Makes his government happy.

0 upvotes
Paul Farace
By Paul Farace (Jun 24, 2012)

On the other you can be the fool who doubts everything put in front of him. I have two eyes, a brain, and 50+ years of experience in dealing with life and politics.

0 upvotes
CaseyComo
By CaseyComo (Jun 23, 2012)

Why doesn't DxO support this camera already? They are sooooo lame!

2 upvotes
chrisnfolsom
By chrisnfolsom (Jun 23, 2012)

I can't wait until google updates their vehicles making google maps with these...

This is similar to my photo stitching I do, but really cool - I love the idea of having the sensor in a circular/dome orientation!

Nice to see some basic science being used - manufacturers are doing a song and dance (some pretty cool dances though) around the same technology although CCD's and backlighting have helped. I like seeing some diversity as in the Foveon, or some of the Fuji hdr stuff, or even that post focusing camera that just came out - even if it's not "commercialized" it still offers opportunities and I welcome the new ideas as they come!

1 upvote
J R R S
By J R R S (Jun 23, 2012)

Difraction be dammed... mobile phones and compacts are working way over the diffracttion theretical limmit and still more Mpix gives more detail... they could push it further i'm sure!

Still nice work... although those plastic lenses are very obviously holding the system back!

0 upvotes
Lawn Lends
By Lawn Lends (Jun 23, 2012)

Does it steal souls too. If so, I am in.

4 upvotes
Lenny L
By Lenny L (Jun 26, 2012)

Wait for Mark II. This version merely sucks the money out of bank accounts.

1 upvote
jmmgarza
By jmmgarza (Jun 23, 2012)

Gotta get me one.

0 upvotes
snackwells
By snackwells (Jun 23, 2012)

They are going about it the wrong way. I would prefer they take a simpler approach: gouge out a human eyeball (or perhaps an "eagle eye"), wire it up digitally, done...they can call it the DARPA's ALIVE-2 camera...maybe in a few years...just have to figure out how to keep the organic parts metabolizing...and the whole system would only require 5mW of power instead of 430W. I would donate an eyeball or two but mine have been experimented on long time ago.

Comment edited 2 times, last edit 6 minutes after posting
3 upvotes
mbpm
By mbpm (Jun 23, 2012)

I am sorry that I am nitpicking here, but "North Carolina" is spelled wrong in the second sentence of the introduction of the article.
I should know, I live in NC.
-_-

Comment edited 1 minute after posting
3 upvotes
lost_in_utah
By lost_in_utah (Jun 23, 2012)

I'm surprised you can read then! Very impressive!

4 upvotes
Douglas Keene, M.D.
By Douglas Keene, M.D. (Jun 23, 2012)

Pardon my ignorance, now do you spell it? It looks ok to me.

0 upvotes
bborowski000
By bborowski000 (Jun 23, 2012)

You obviously like to nitpick even when nothings wrong.......... Proof!!!! it is spelled properly!!

0 upvotes
Halocastle
By Halocastle (Jun 23, 2012)

Borrowabrainski000

Uh, 'cause they fixed it. Ok, this "page," the article, it's not paper, ok. It's electronic data stored as ones and zeros in the form of HTML code, the language of the web. Now, this may be hard for you to believe, but the code can actually be changed--no, really. They actually designed the system so mistakes can be fixed.

2 upvotes
mbpm
By mbpm (Jun 23, 2012)

OK. First off, let me say that I absolutely love reading the articles and comments in dpreview.com. There are great opinionated people with experience in photography and the arts here. Unfortunately this website, like every other website, is inundated with Trolls...

Lost_in_utah: haha, u too funny bro. I like your sense of humor. Honestly. I am personally giving you one like.

Douglas Keene, M.D: It was spelled "Caroline," but they fixed it since I've commented on the matter.

Bborrowski000: Seriously, I don't like to be that "smart alec" in the corner who thinks he knows it all. I don't crave for attention, and I almost regret posting the correction. But, hey, this is a professional article in which I personally hold high regards when reading it.

Halocastle: Thank you. At least someone understands.

Comment edited 4 times, last edit 8 minutes after posting
3 upvotes
Peiasdf
By Peiasdf (Jun 23, 2012)

Is that Saturn at the upper right side? Impressive that they can capture the planet so clearly.

BTW, wasn't there some scanned large film that produces 1GB per 4"x5" film?

2 upvotes
maboule123
By maboule123 (Jun 23, 2012)

Do I need this technology to take better pictures of nude girls?
Am I missing something here?
HELP!

1 upvote
aris14
By aris14 (Jun 23, 2012)

Yes you do! The girls!
lol

4 upvotes
matthewhuck
By matthewhuck (Jun 23, 2012)

Don't you mean millions of years, CameraLabTester?

2 upvotes
Tariag
By Tariag (Jun 23, 2012)

My biggest complaint is about exposure. Why did they went for an individual exposure? It makes like a "bubble" picture!
With same exposure for all subsensors, they may be less DR but a far better looking result imho.
Then I also clearly see an optical resolution limit (don't know if it's the main lens or the sublenses), but I think this can be improved too...

0 upvotes
CameraLabTester
By CameraLabTester (Jun 23, 2012)

The humble Housefly had a better design for thousands of years. What a brilliant creator.

7 upvotes
JoeAmateur
By JoeAmateur (Jun 25, 2012)

yes, evolution is the best way to make anything, but the fly's eye actually took millions of years, not thousands lol. I'd wager the R& D on this guy was significantly shorter.

Comment edited 49 seconds after posting
1 upvote
bborowski000
By bborowski000 (Jun 26, 2012)

Yeah Righ...........t I've never seen anything make itself yet

0 upvotes
monkpilots
By monkpilots (Jun 23, 2012)

how many fingers am i holding up?

2 upvotes
K_Photo_Teach
By K_Photo_Teach (Jun 23, 2012)

The Nikon 808 has a tiny lens and doesnt seem limited by the lens. Some tests here: http://www.gsmarena.com/pureview_blind_test-review-773p3.php
compare is favourably with even FF DSLRs.

0 upvotes
CaseyComo
By CaseyComo (Jun 23, 2012)

Nokia 808.

0 upvotes
K_Photo_Teach
By K_Photo_Teach (Jun 24, 2012)

hahaha yeah thanks!

1 upvote
chriscotec
By chriscotec (Jun 23, 2012)

Does it come with a camera strap?

3 upvotes
ManuelVilardeMacedo
By ManuelVilardeMacedo (Jun 24, 2012)

Yes, a wrist strap.

0 upvotes
gasdive
By gasdive (Jun 23, 2012)

I've been suggesting something like this for a while, but without the central lens. If you had something like an iPad but with the whole back coated with close spaced cameras. You could fit about 2400 cameras in that space. Then with some processing you could do things in post like 3d with variable intraocular distances, variable effective aperture, refocus after shot, variable depth of field,variable point of view, pixel drizzel hyper resolution with stacking for increased effective iso and reduced noise. All these artistic decisions could be made *after* the capture.

2 upvotes
air_on
By air_on (Jul 23, 2012)

I hope you know how much of an overestimation you're making. You're doing a downsampled, polyphase filtering and some of those methods you're talking either does not make sense (intraocular distances) or will need additional physicality sensors or you will need to alter the surface shape on which your cameras are placed.

I hate to be a pessimist. Good news though is military development is far ahead of what you can imagine. This gigapixel camera is old news.

0 upvotes
iae aa eia
By iae aa eia (Jun 23, 2012)

I already stopped to think that a curved sensor could make possible to acheive high resolution images using less complex optics, just like an eye does, and now this is becoming a reallity.

0 upvotes
G Davidson
By G Davidson (Jun 23, 2012)

This does seem something like an insect's eye. Designing digital cameras from th ground up to be more like eyes, with round sensors or individual ameras workin together seems like a very interesting route. With enormous resolution and dynamic range, the focus on optics would move completely to the electronic realm, with cropping, HDR, 3D and refocusing possible after the event. Interesting that the military is leading here as with so much else, with consumer designs no doubt to follow.

0 upvotes
Soulhunter
By Soulhunter (Jun 24, 2012)

Actually its just like a human eye, one front lens and lots of sensors in negative sphere arrangement behind. Insect eyes work the other way around, lost of lenses in sphere arrangement in front and sensors right behind leading to the center...

1 upvote
CFynn
By CFynn (Jul 6, 2012)

@iae aa eia
I think perhaps you underestimate the costs that would be involved in making curved digital sensors.
However you could try out the idea of a curved sensor by slicing ping pong balls and coating the inner surface with photographic emulsion - and building a simple camera to hold this "film".

0 upvotes
Jogger
By Jogger (Jun 22, 2012)

the resolving power will be limited by the lens methinks, even in the samples provided, the 100% are useless except for surveillance

0 upvotes
gasdive
By gasdive (Jun 23, 2012)

If instead of putting it all through one lens you had a flat board coated with lenses in an array you would have a much higher resolving power. If you had a board about the size of an iPad then the resolving power should be the angle in radians equal to the ratio of a wavelength of light to the width of the board. So the resolving power should be about 0.0003 degrees or about 1 arc second. To give you some idea of how fine that resolution is, Neptune is more than 2 arcseconds wide viewed from Earth. Another way of putting it: If your eyes were that good you could read the 8th row of the eye chart from 1200 feet.

0 upvotes
JoeAmateur
By JoeAmateur (Jun 25, 2012)

I'm sure the thousands of physicists and mathematicians at DARPA probably never thought about or discussed anything similar to your idea, right?

Probably none of the tens of thousands at Sandia, Los Alamos and Lawrence Livermore ---OR UC Berkely, Cal Tech, JPL, MIT that support DARPA did either.

AND, let's not forget how scientists HATE to disprove one another's theories! Nobody among those PhDs wanted to one-up that design, no sir!

You're AMAZING!

0 upvotes
gasdive
By gasdive (Jun 26, 2012)

no they've thought of it alright. Google "Light Field" and see how much work is being done on it. (it's not "my" idea). It just seems much more relevant to "photography" than this idea does.

0 upvotes
KoKo the Talking Ape
By KoKo the Talking Ape (Jun 29, 2012)

Wow, JoeAmateur, that was uncalled for. gasdive he was just throwing out an idea. He wasn't showing off or claiming credit for creating that idea. Bitter much?

0 upvotes
Ultan
By Ultan (Jul 1, 2012)

gasdive - no, the resolution is limited by the sizes of the individual lenses, not the size of the whole board, so your resolution would be more on the order of arcminutes than arcseconds. They are using 1 big lens rather than the conventional camera-array setup for just this reason. (Over long distances the fluctuating distortions in the atmosphere are the limiting factor, anyway. With sufficient framerates and processing this can actually be turned into an advantage.) The real problem with high resolution cameras is the vast amounts of data that have to be stored and processed. 1 gigapixel x 8 bits x 30frames/s (which is just an 11x11 array of quad-HD (QFHD) cellphone cameras) = about 30 gigabytes per second raw data, 3.7TB/hr - which can be reduced by a factor of 10 or more for storage, but each byte needs hundreds of operations done on it, so that's on the order of teraflops. Even very power-efficient processors still will need hundreds of watts.

0 upvotes
idbar
By idbar (Jun 22, 2012)

So where's "HowaboutRAW", asking his question? That's probably a massive proprietary file.

4 upvotes
Georgi Bonchev
By Georgi Bonchev (Jun 22, 2012)

Finally some competition for the Nokia 808 :D

4 upvotes
Sam Carriere
By Sam Carriere (Jun 22, 2012)

My bet is that it will be in the stores before the Nikon D800.

4 upvotes
calijente
By calijente (Jun 23, 2012)

Is B&H taking preorders?

0 upvotes
Randy Colwell
By Randy Colwell (Jun 25, 2012)

Yes, and B&H will get their stock approximately 12 months after everyone else. ;)

0 upvotes
mathew crow
By mathew crow (Jun 22, 2012)

Should i get this one or wait for the next one? This was announced at 17:24 so its getting a little long in the tooth by now. If they can fix the orbs, vertical banding and increase the iso I might pre-order one. I need something like this for pics of my cat, flowers, and brick walls.

20 upvotes
ManuelVilardeMacedo
By ManuelVilardeMacedo (Jun 22, 2012)

I found a gallery of a guy who likes to photograph cork pads. The cork grain must look awesome at this resolution...

Comment edited 40 seconds after posting
6 upvotes
jameshamm
By jameshamm (Jun 22, 2012)

I have thought about something similar. If you look at the human eye, it accomplishes an incredible feat with a single lens, and lots and lots of specialized receptors. I was wondering why no camera manufacturer has looked into the possibility of a curved sensor with a simple lens, since most of the work the lens has to do in our current technology is spherical to plane mapping without aberrations. This camera here is along the same lines even if it's not really applicable to artistic photography. What would make more sense is a miniaturization of this concept into a simple lens projecting on a curved sensor (or multiple sensors, or multi-layered sensors for detecting light and contrast).

Comment edited 2 times, last edit 3 minutes after posting
3 upvotes
meanwhile
By meanwhile (Jun 23, 2012)

Agreed, but I think the production of curved wafers for the sensors is problematic.

3 upvotes
Biological_Viewfinder
By Biological_Viewfinder (Jun 22, 2012)

I always find it so fascinating when technology just completely blows the doors off everything most of the people here believe consider to be some sort of speed limit.

This is just the beginning, our children may be getting cameras that connect directly to our human optics.

Technology knows no bounds. There is no ISO limit. There is no MP limit. There are no LIMITS!!!!

1 upvote
ianimal
By ianimal (Jun 22, 2012)

The laws of the Universe sets the limits. But I am not sure we know all possibilities yet.

2 upvotes
Hugo808
By Hugo808 (Jun 22, 2012)

Hmmm, I'm gonna need a bigger SD card...

8 upvotes
Tom Zimmer
By Tom Zimmer (Jun 23, 2012)

Not really, you should be able to get somewhere between 1 and 3 shots on your 8GB card. Just use a different card for each shot, the camera could have a stack of them, and it could spit one out after each shot is taken. Kind of like a huge polaroid camera,
It is of course the post processing that is going to kill you. Can you say 30 minutes of computer time before you get to look at an image? On the other hand, you won't need a huge disk drive, because you can just leave the original image on the card and that becomes the archive. This could be a lot of fun!

Comment edited 1 minute after posting
0 upvotes
tkbslc
By tkbslc (Jun 22, 2012)

This is obviously more for recon and intelligence gathering than it is for artistic photography. Big Brother wants to be able to determine every face and license plate in a 2 mile radius.

Comment edited 39 seconds after posting
8 upvotes
Richard Butler
By Richard Butler (Jun 22, 2012)

I think that's a reasonable assumption, given how few photo contests DARPA enters. However, if it offers a route to higher pixel counts, then landscape photographers might appreciate it in the long-run.

8 upvotes
cgarrard
By cgarrard (Jun 22, 2012)

LOL :) Yep very few indeed.

Big bro has been watching for a lot longer than most realize, and a lot sooner than 1984.

C

4 upvotes
acidic
By acidic (Jun 22, 2012)

Can it do 23240p video?

12 upvotes
dohdoh
By dohdoh (Jun 22, 2012)

Does Gigapan make a mount for this camera?

0 upvotes
KoKo the Talking Ape
By KoKo the Talking Ape (Jun 22, 2012)

I notice that the image captures quite a wide angle. I imagine that is because that spherical lens is designed that way. Is that required, or could you get that high resolution in a narrower field of view? Getting a very high pixel image is not all that groundbreaking if you get those pixels by taking a bigger (wider angle) image.

And is there a theoretical limit to pixels per solid angle in an image, aside from lens limitations? How many pixels are actually in the light within a given field of view?

0 upvotes
rbryll
By rbryll (Jun 22, 2012)

Assuming a lens with low enough diffraction limit, the atmospheric blur disc size for landscape photography would probably be close to what they get in astronomy - about 1-2 arcsecond at best. For 45 degree square field of view (standard lens), that would give 162,000 x 162,000 pixel image, or about 26 gigapixels. Without the atmosphere, the lens diffraction limit and aberrations define the highest possible resolution.

0 upvotes
JorgeLima
By JorgeLima (Jun 22, 2012)

The number of pixels per solid angle can be any. The question is are they resolving more detail than fewer? Maybe you what to know if there is a theoretical limit for the resolving power of an optical system. The only theoretical limit I can remember is the diffraction. The diffraction limits the angular resolution in a way which depends on the light wavelength and diameter of the aperture. Nevertheless we should not see it as a limitation of the lens, because even a perfect lens (i.e aberration free) with an infinitely large aperture has an angular resolution which is just 3 times better than a diffraction limited F/1.4 lens. However, in practice the angular resolution is limited by aberrations, not by diffraction. The aberrations may come from the lens itself, but also be introduced by atmospheric turbulence. The later is almost always visible when using long telephoto lenses, specially when looking close to horizon.

2 upvotes
KoKo the Talking Ape
By KoKo the Talking Ape (Jun 24, 2012)

@rbyll, thank you for that information. Very interesting! So already the DARPA approach, which allegedly should work for up to 50GP, could approach practical resolution limits. I wonder if atmospheric blur for landscapes would be greater than astronomic photos because the air is polluted, heated by the ground, full of weather, etc., or if it would be LESS because astronomic photos are taken through through the entire depth of the atmosphere, albeit much attenuated at the higher altitudes.

0 upvotes
KoKo the Talking Ape
By KoKo the Talking Ape (Jun 24, 2012)

@JorgeLima, thank you for the comment. You are helping me refine my fuzzy thinking. So let me ask a different question. Is there a theoretical limit to the *information* that a sensor could receive in a a given solid angle? And like rbyll, you mention the diffraction limits created by optics. So just for fun, suppose we do away with optics. How much data could an infinitely refined and sensitive sensor capture in a given solid angle? That might be a quantum physics question.

0 upvotes
mike16
By mike16 (Jun 26, 2012)

Indeed this is the real question, its not about taking nice wide angle landscapes its about directing the camera to get very high resolution of some object or part of city, effectively with a very zoomed in view like a battery of telephoto or macro lenses. Its much less clear how this kind of design would be able to deal with that.

0 upvotes
Cy Cheze
By Cy Cheze (Jun 22, 2012)

Meanwhile, most security cameras have low resolution, mediocre lenses, and shoot thousands of uneventful frames that no one has the time to look at, and which are unlikely to capture anything actionable before a heist or attack. A Breivik would be able to park his loaded car and stroll away, without drawing any attention--unless, of course, he made himself very conspicous by lumbering about with a big DSLR slung on his neck.

Cheapter to boost security posting dozens or hundreds of mock cameras near a vulnerable site. Some might be functional. More important to make perpetrators worry they are being watched. People imagine spooky-looking pod captures hyper-resolution images. They've seen movies where the protagonist tells a geek, "Enhance," and the computer makes a blurred distant object turn into a crisp license plate or finger print. Of course, the counterstrategy is simply to employ folks prepared to perish in the act.

0 upvotes
EmmanuelStarchild
By EmmanuelStarchild (Jun 22, 2012)

Most photographers can't handle 18MP.

8 upvotes
cgarrard
By cgarrard (Jun 22, 2012)

LOL!

2 upvotes
Tord S Eriksson
By Tord S Eriksson (Jun 23, 2012)

I think my limit is 10MP, just now. But more seriously, if it one day is possible to make cameras similar to the one here, that just uses one little chip, lens designing suddenly becomes much, much easier.

But my, what a monster - no way my SD cards are big enough!

0 upvotes
Total comments: 132
12