Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
The leaning tower of Piezo

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



Swarm Camera

A swarm of cameras that co-operate to produce a single image
  [vote for,

The problem with cameras today is that they are limited by the single lens they have.

Open the aperture too wide, and the picture is over exposed. Open it too little, and the picture is too dark. So, the camera has a little light meter to detect how large the opening should be.

Problem is, when you are taking a picture of something with a wide range of light values, some areas will be too dark, and some will be too light. You try to do the best you can to get the correct exposure on the parts you want.

I propose a new type of camera that is actually composed of many small cameras. Each has their own light meter and aperture, so they can each guess the best exposure of what they see. That way, each camera produces a well-exposed part of the overall picture.

The idea here is that it produces an image that is not over- or under- exposed *anywhere*, no matter how bright or dark certain regions might be.

It probably won't be aesthetically pleasing, but it would help when you are using photography to record information.

lawpoop, Sep 11 2003

Terraserver composite images http://terraserver-...x=27&y=344&z=16&w=1
This is what output from a swarm camera might look like. [lawpoop, Oct 17 2004]

Table of pinhole camera exposure times http://neon.airtime...inhole/exposure.htm
Range of pinhole camera exposure times: 1 second to 5.5 hours. [lawpoop, Oct 17 2004]


       And the photo lacks any contrast making it extremely boring.   

       And it's spelled aperture.
grip, Sep 11 2003

       Modern cameras use multi-point exposure metering, where they sample multiple points in the frame and calculate the correct exposure based on the entire composition. So in a sense this is baked, if you consider each metering point a separate camera..
DeathNinja, Sep 11 2003

       Do insects see like this? With compound eyesight, I guess each eye is like a tiny camera. Maybe nature uses this technique because under certain circumstances it works better.+
TIB, Sep 11 2003

       "Apature"? I like "sawrm" better. That's the way I spells it when I get to the bottom of the bottle.
pluterday, Sep 11 2003

       [DeathNinja] Yes, modern cameras do have multi-point metering, but they still have only one lense through which to allow light. So, it's not really baked. They still fail when they try to automatically guess an image with extreme lighting -- they just get an average of the whole image, which migth not look good anywhere. A professional will point the light meter at the part of the image they want to be correctly exposed, and go from that.   

       In the swarm camera, none of the cameralets would need a multi-point meter. The idea here is that it produces an image that is not over- or under- exposed *anywhere*, no matter how bright or dark certain regions might be.   

       It probably won't be aesthetically pleasing, but it would help when you are using photography to record information.
lawpoop, Sep 11 2003

       you'll have a problem because each lens would be recording from a slightly different vantage point, thus creating a "crosseyed", multiple image...
hawg, Sep 11 2003

       You got me, Bubba. I know everything in the whole world, and I was trying to pass off someone else's idea (in the halfbakery, no less!) as my own, for my own personal glory.   

       Yet I could not forsee that someone, yes someone, might also be aware of that same bit of info put out there by *a bestselling author*.   

       Bubba, you got me. You have won this battle, but not the war.
lawpoop, Sep 12 2003

       This could also be done by having two ar more frames with diferent exposures shot in rapid succession and then blended together.
bristolz, Sep 12 2003

       Digital cameras, especially when used with facial recognition do this already. By sucessively reading the pixels you can get a built up exposure (each pixel in a digital camera actually "fills up" as it is exposed, so the longer the exposure the more electrical charge it will have). Using this you can define regions of different exposure.   

       Essentially the same idea, but a 1 mega pixel CCD would therefore be like 1 million (well a little more) little cameras each capable of being used at different exposure levels.
scarkner, Sep 12 2003

       [grip]'s leadoff point bears repeating: a uniform 18% grey card does not an interesting picture make.
n-pearson, Sep 13 2003

       well, the swarm idea isn't baked yet (imho) but i get around this by using the "spot" meter function which allows exposure set by a particular point in the image ...granted that this is not always especially convenient, still an option for those who recognize exposure issues such as contrast, shadow detail, etc., etc. ...
stig, Sep 13 2003

       OK, let me repeat myself:   

       This is not for pretty pictures. This is to get DATA from a PHOTOGRAPH.   

       The problem with current cameras is that they have one global aperture and exposure time for every spot on the film.   

       Except for areas of the image where the exposure is exactly right, dark areas will show no data (i.e. they will be black) because that part of the image doesn't send enough light to expose the filme. Similarly, light areas will show no data (i.e. they will be all white) because they send too much light and totally expose the film.   

       If you have a complex image, with lots of lights an darks, and you would like information from all of it, you would need to shoot a separate shot for each area, then paste it together in a giant collage.   

       The swarm camera is like an automatic collage generator.
lawpoop, Sep 17 2003

       I understand what you're goal is, but I don't think you understand how photographs work. It takes a scene with a hell of a lot of contrast to produce a photograph in which you lose data in the highlights and even more to lose it in shadows (for negative film opposite for slide.) And that is a limit of the film (and processing), not the camera.   

       Additionally, the shading in a photograph IS information. If you could provide an example of the information you're trying to record, that would be helpful.   

       And finally, if you want universal shading of an object, there is a much simpler device to achieve that: a flash.   

       And finally again, I do like your idea of a swarm camera for producing collage photos. But that's somewhat baked by these cheap plastic cameras that take 4 successive photos.
grip, Sep 18 2003

       1/ don't take critisism here too hard - this is a site about discussion, and not all of the points made were designed to be to the detriment or 'correction' of your idea.   

       2/ get yourself a decent digital camera and a copy of photoshop. Then stop bitching.
yamahito, Sep 18 2003

       I understand how cameras work. Any photography guide will tell you that photos 'flatten images'. The totally white/black examples I used were the extremes, but just look at any photo with dark and light areas, and you will see that *in those areas*, there is less contrast.   

       As people have said above, "just get a light meter reading of that area where you're losing contrast". OK, then you are just changing the global aperture setting for *all areas of the image*, not just what you are trying to correct for.   

       So, if you have an image where the center is fine, but there is a dark corner, and you meter the dark corner, then the center is too bright.   

       So, the solution is to take lots of little photos (more than 4 ) to create a composite image with good contrast everywhere in the image.   

       Check the link I made to the composite satellite images. You'll notice that each image had a different exposure time than the others. This is what I envision a swarm camera's results to be like. It's not 'great art', but it is useful for getting data out of the image.   

       [yamahito] don't take my replies to comments too seriously. ;)
lawpoop, Sep 18 2003

       Additionally, indendendant cameras can also have their own focus, so you could have a more of the image in focus.   

       A single camera has a global focus which applies to all areas of the film. So, you either have to have things far away in focus, or things close up in focus. With a swarm camera, you can have both.
lawpoop, Sep 19 2003


       I stand by the second statement, however: my camera has an auto-exposure lock (I think it's a fairly common feature to mid-range digital cams) to make sure the global exposure isn't too high. I realise this does not yet address your given aims, but an important thing to realise about digital cameras is that they actually incode a lot of information in pixels that are too dark for us to resolve detail, but lose detail at high exposures. If you set your exposure such that the most exposed part of the photograph is not too bright, you can change the contrast of the rest of the photograph in photoshop - with a great deal more control than the swarm camera would appear to offer...   

       With regard to the idea of different focii, there is a concept in photography called 'depth of field.' This depth of field is the distance around the focus point which is still in focus. Any (decent) camera will allow you to adjust the aperture width to vary this, to acheive exactly the effect you have just described. For close up shots, you can set a low depth to blur the background and make the foreground more striking - for landscape shots, you can set a near-infinite depth of field to get crisp detail for the whole image.   

       Of course, one effect that the swarm camera could produce that conventional cameras cannot (afaik) would be to have something in focus in the foreground, out of focus in the mid-ground, and back in focus in the background.. but then again, setting up a swarm camera to do this would probably still be harder than 'faking it' in photoshop...
yamahito, Sep 19 2003

       I gotta say that you don't know how cameras work. Most lenses have an adjustable focus and 'infinity' is a setting. Meaning everything in a scene CAN be in focus with a single lens.
grip, Sep 19 2003

       I have to say you've never used a camera. When you get your hands on one, try this:   

       Set the focus setting to infinity. Point it at something close up.   

       It will be out of focus.   

       Now, put that close up object in focus. Now point the camera at something far away. That will now be fuzzy.
lawpoop, Sep 19 2003

       The 'infinity' setting on a (decent) camara will not be talking about the focal length, but the depth of field - it is the mode where the camera sets the aperture as small as possible so as to mimic a pinhole camera as closely as possible.   

       If it does this well, or even if one were to use a pinhole camera, everything, no matter how close up or far away, should be in perfect focus.   

       Where did the other bit I was about to reply to go?
yamahito, Sep 19 2003

       Christ [lawpoop], if you can't take advice from someone who understands the underlying technology of your invention better than you do, why are you posting here?   

       When you look through the lens of your typical SLR camera, the focus will NOT show what will be recorded on the film. Why? Because the aperture stays wide open (meaning very little is in focus) until your press the shutter, THEN it closes down to your setting and opens back up after the shutter opens.   

       I will say I should have been more specific in my last anno. If you set your lens to a smaller aperture (depends on the lens f/16 on mine), and the lens to infinity, you have have almost everything in focus (on mine that's 1 foot from the lens to infinity).
grip, Sep 19 2003

       Jesus, Mary and Joseph, [grip]! Why shouldn't I trust anybody on halfbakery? Obviously it's safe to assume that *everybody else* here knows more than me. Heck, why shouldn't I trust you?   

       I'll shoot a roll of things with the focus set to infinity. If you're right, I'll trust you. Otherwise, why should I? You've contradicted eveything *I thought* to be true.   

       I did the same to you, and you don't trust me. I'm following the same rules you are.
lawpoop, Sep 19 2003

       Don't pinhole cameras have to have exposure times of *hours* in order to properly expose an image?   

       A swarm camera would do this in normal SLR times, like 2 seconds or less.
lawpoop, Sep 19 2003

       //Don't pinhole cameras have to have exposure times of *hours* in order to properly expose an image?//   


       Well, I suppose it depends on how bright the subject is. But I don't think so.   

       For the record, I'm strictly an amateur when it comes to photography.
yamahito, Sep 19 2003

       Shoot away [lawpoop]. But you may if you use a 2 second shutter speed your subjects may be quite in focus, but VERY blurry. 1/60 of a second will be just fine.
grip, Sep 19 2003

       Now I got it! Here's the problem: exposure time. The inifity setting only works when you can define the exposure time. With a regular camera has a global aperture and exposure time for all regions of the film.   

       With the swarm camera, each region of the overall image that one cameralet covers has it's own exposure time (and focus setting), so the swarm camera has the potential to do the whole image in focus in shorter time than a regular SLR set to infinity.   

       Ok, that might not be totally correct, but if you're trying to make a photo with everything in focus, there has to be some difference in exposure time between an SLR with a global aperture and a swarm camera.   

       So, the reason the pinhold camera takes so long to expose things is that it lets in so little light. When you shrink the aperture, the exposure time grows.   

       The swarm camera would compensate for this, since the exposure time for the composite image would only be the time of the longest exposure of one cameralet.
lawpoop, Sep 19 2003

       The latest link shows a range of exposure times of pinhole cameras beign between .5 seconds to 5 hours 20 minutes.
lawpoop, Sep 19 2003


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle