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
"My only concern is that it wouldn't work, which I see as a problem."

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.

user:
pass:
register,


                                                     

High Contrast Camera

Use LCDs to block excessive light
  (+10)(+10)
(+10)
  [vote for,
against]

There are lots of occasions when photographing a scene, and part of that scene (e.g., the sky) is very bright, but the region of interest is relatively dim (everything below the sky).

If you set the exposure/apeture so that the foreground looks normal, the sky is washed out and white. If you set the exposure so the sky looks normal, the foreground looks dark or even black.

What if there was a clear grid of LCD pixels (like an LCD tv screen, but transparent) directly in front of the CCD?

The camera could quickly make each LCD pixel as clear or as opaque (black) as necessary, to ensure that the brightness light reaching the corresponding CCD pixel is close to the middle of the CCD's optimum sensitivity.

To calculate actual levels of brightness for each pixel, multiply the percentage of light let through the LCD pixel by the brightness detected by the CCD pixel.

goldbb, Jan 08 2009

Pixim http://www.pixim.co...chnology/technology
A simpler way [csea, Jan 08 2009]

A deep red half-grad on black and white film http://www.flickr.c...ntindale/866186383/
[Ian Tindale, Jan 08 2009]

A tobacco half-grad on colour film http://www.flickr.c...tindale/2896920112/
[Ian Tindale, Jan 08 2009]

A dark grey half-grad on black and white film http://www.flickr.c...tindale/2862581415/
[Ian Tindale, Jan 08 2009]

Sun glinting strongly off water http://www.flickr.c...262/in/photostream/
(no grad or ND involved) [Ian Tindale, Jan 08 2009]

What a black and white picture would look like using this technology http://www.nexterna...avy/images/grey.jpg
[Worldgineer, Jan 09 2009]

Dolby Vision http://investor.dol...fm?ReleaseID=358113
Related technology for high dynamic range display [csea, Jan 12 2009]

TWICE http://www.twice.com/
This Week In Consumer Electronics [csea, Jan 12 2009]

[link]






       The usual way to keep a CCD sensor from overexposure is to detect when any one element becomes saturated, and stop the integration. A company called Pixim [link] has a method that allows the integration time of each element to be adjustable for optimum exposure.   

       For examples, see their Images page.
csea, Jan 08 2009
  

       A better way is to use a half-grad ND filter.
Ian Tindale, Jan 08 2009
  

       Nice photo [IT], but what if the sun is glinting strongly off the water below the grad section?
coprocephalous, Jan 08 2009
  

       The grad rotates. But in that case, the whole of the scene represents highlights and shadows from a high dynamic range, so other than using a plain ND, you shoot or process for as low a contrast as possible.
Ian Tindale, Jan 08 2009
  

       Actually what you are describing would not result in a High Contrast Camera but rather a High Dynamic Range Camera which is a very different thing.   

       Its a nice idea however probably not a new one, the real key here is HOW do you make it work in real time, if you can figure that out call nikon and canon after filing for your patent then prepare to be a rich man.
jhomrighaus, Jan 08 2009
  

       //The grad rotates//
How fast?
Surely that just has the effect of averaging the ND-ness over the whole frame?
coprocephalous, Jan 08 2009
  

       With specular highlights such as sun glinting strongly off water (see appropriate photo linked) with no bright sky visible in the scene then you're not as such trying to reduce a part of the scene illumination in comparison to another part, the whole scene represents high contrast, so the solution is a bit different. In that case, you do things to aim toward a low contrast result (if you want to preserve any semblance of tonal range inbetween highlight and shadow, that is). For example, I put most of my scans through a (not very) secret transfer curve shape, intended to keep highlights intact.
Ian Tindale, Jan 08 2009
  

       Yes, but the original idea addressed the problem with bright sky versus the rest of the scene, and typically that is where you'd deploy an ND grad. If you have highlight points scattered around the entire scene (eg, a night scene) then a grad is no use, and you either surrender to them and let them burn out (after all, they're hardly important information) or to prevent large areas of the scene from hosting flared specular highlights, you have to address the contrast response. It's a problem even the human eye finds hard. Your idea of a parasite sensor is interesting, and will particularly address the latter kind of response error (the specular highlights distributed around the scene) but would also work quite well in balancing the sky and the land (a different sort of problem). Another way would be to modify the bayer matrix so that instead of two greens, there's one green with a modified response, and a grey with comparatively poor response.
Ian Tindale, Jan 08 2009
  

       I think it's a good idea and could save some expensive camera equipment. A friend of mine who is transisioning into being a professional photographer brought his expensive camera on his honeymoon to Hawaii and took some gorgeous shots, till he must have taken a long exposure of a sunset and fried some pixels on his digital camera. All the later pictures have a grey area where the sun was in that one shot.   

       I was thinking about this same idea for a welding helmet years ago when LCD shutter helmets first came out. The idea was to only darken the brightest area. It's a lot harder on a helmet as there are two eyes to block and relative position of such would vary, but there may be a market out there.
MisterQED, Jan 08 2009
  

       csea, thanks, Pixim is very similar to what I had in mind! True, it sets the per-pixel exposure time instead of per-pixel apeture, but it's very close.   

       jhomrighaus, thanks, you're right, it's high dynamic range instead of high contrast.   

       Ian Tindale, your point about pixels burning out due to too much is something I also had in mind, though I forgot to write it. And, I think it's a very important point!   

       I think the best solution would be to start with pixim's design, of having per-pixel digital processing, but instead of shorting the per pixel exposure time, darken the pixel's LCD to reduce the amount of light reaching the pixel's light sensor.   

       This would prevent the pixel light sensors from being burnt out due to too much charge, or UV exposure, but it wouldn't prevent pixels from being destroyed by heat damage, though, since a darkened LCD absorbs the light that hits it, generating heat.   

       One solution would be a cooling system (air, liquid, thermoelectric, etc.,) but this could get complicated, and expensive.   

       If someone could invent an LCD-like device which reflects light instead of absorbing it, then heat damage could be prevented without needing any cooling.   

       Does anyone know if electroreflective materials exist which are transparent when off, and reflective when on? (or vice versa)
goldbb, Jan 08 2009
  

       //electroreflective materials exist which are transparent when off, and reflective when on? //   

       Some variant of DLP technology might be able to do the job. Conceivably adjustable angle of reflection could fill up sequential CCD "bins" for a type of "autoranging."
csea, Jan 09 2009
  

       It is inconcievable that you [goldbb] are suggesting a pre-modification of RAW-type data to a CCD. It presumes that a set of instructions is being placed on the recieved data to limit the received data. This kind of architecture often results in positive feedback loops. And I don't mean "positive" in a positive way!   

       And if you are trying to save a few pixels in your CCD from exposure, it is a bit late for that, don't you think? Or is the underlying innovation a first, completely translucent, CCD feeding information to a processor, then a LCD, which blocks certain things from the end CCD? Are you even vaguely familiar with causality, and its speed limit?   

       When capturing the data, why not have your fancy algorithm save it as another file? With a 16 bit per pixel device it is possible to histogram everything you want out, or in, or anywhere in between. 32 bits per pixel and above are there for the same reason a dog licks its balls, it can! Or is that the real problem? 16 bits per pixel a little to much for your puny human eyes?   

       You want to save your CCD, put a filter on the lens. You want to manipulate images for the best picture, your best bet is *all* the data.
4whom, Jan 09 2009
  

       The LCD can momentarily shut off while the circuit is calibrating, so it is not necessarily a positive feedback loop.
Spacecoyote, Jan 09 2009
  

       So the LCD shuts off in what? Opaque, or tranlucent?" Positive" was not a reflection <cringe> on what the LCD does, just what it does with the information from a point that has past.   

       So let's get to the crux of the matter. You want to reduce "glare" from images, or, you want to protect your semi-conductors. If the former you would do best with all data possible. If the latter, there is no provable way to assimilate information and act retrospectavely. That is all I meant, as that is what I understood the idea to propose.
4whom, Jan 09 2009
  

       Dolby made public at the Consumer Electronics Show in Las Vegas a related technology for high dynamic range displays [link].   

       This uses adjustable LED backlighting behind an LCD display to achieve darker darks, brighter highlights.
csea, Jan 12 2009
  

       can anyone recommend a good news source for what has been happening at CES? BBC is pretty poor, and Popsci is a little too much like an editorial.
miasere, Jan 12 2009
  

       TWICE is the official publication... [link]
csea, Jan 12 2009
  

       ta
miasere, Jan 13 2009
  
      
[annotate]
  


 

back: main index

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