Using The GIMP (or Photoshop) to Improve Photos?

Nom du Keyboard asks: "Is it possible to use The GIMP (or Photoshop) to improve my digital photos? I have a mid-range 7.1MP Olympus camera capable of shooting in Raw mode. When I inspected a section of clear blue sky on a bright, sunny day (which I've long believed to be relatively good reference of uniform color and brightness) I was surprised (disappointed, since I expect digital perfection) at the variance in adjacent pixels. It's also a quick way to identify any bad pixels. Surprisingly, actual photos from this camera look pretty good despite this variance so far. Moving on from that point it led me to wonder that, if you shot a uniform white surface, perhaps blurred as much as possible to avoid any imperfections in the surface itself, could a correction (adjustment) layer be created in GIMP or Photoshop exactly tuned to your camera that fixed the variations in your CCD sensor and improved the image quality in the process. Any thoughts?"

Comments

try it

[tverbeek]

I don't know. Why don't you try it?

Interesting idea

[Aladrin]

That's an interesting idea, but it assumes some pretty clean conditions. The light has to be absolutely the same over the entire surface, it would probably need to be blurred as you said, the surface would have to be absolutely the same color everywhere (no dust, no marks), the surface would to be completely non-reflective, and probably some other things that I haven't thought of. It would be extremely hard.

It also assumes that the variations are always the same, and that the variations in your photos are from defects and not from the natural color differences in the real world and the digital camera's attempt to map them to a very restricted color palette.

Clear blue sky != monochromatic

[Gothmolly]

a) If you are using anything above ISO50 on a cheap digital (like yours), you will get ISO noise
b) blue sky is not really blue, you can't expect 7.1 million pixels to all agree
c) there may have been microscopic dust on your lens

Basically, you're looking for your camera to be Adobe Illustrator, and it isn't.

Good thing you don't shoot with film

[Overzeetop]

You'd be just devastated if you blew a film image up to the level where you could see the grain.

Here are two questions for you:

1) Do you find that you are printing your images at sizes larger than 12x18?

If you are, then you probably ought to have more pixels (i.e., a better camera). I'm okay with digital pictures down to about 150dpi, others swear that you need 300+. Then again, there are people who swear that $3000 unobtainium coated silver strands wrapped in virgin PTFE and assembled when the planets are in alignement make their music sound better.

2) Presuming you are actually printing at at least 200dpi, can you really see the difference without a loupe on the final prints? I'm not worried about your monitor, because I'm going to bet that if you have a consumer-level camera, you're not doing photoediting on a 7.1MP monitor.

You see, if you can't tell, don't worry about it. Let your geek side go and spend more time in the field and less time in the darkroom. Seriously - unless you have significant image problems you can see in your final output, the camera and imaging is good enough. Go take some great pictures, and worry a bit less about having digitally perfect pixels.

Why do you care for photographs?

[dlevitan]

Yes, your $1000 digital camera is not going to have a perfect CCD. There is no such thing as a perfect CCD. And I don't understand why you care unless you're trying to do science work with it. Look at it this way, no one is ever going to look at your picture and say its horrible because one pixel is slightly different than the one next to it. You look at the content of the whole photograph, not three pixels.

If you are trying to do science, then a DSLR is not what you need. DSLRs use Bayer interpolation to create a color image. This inherently kills your accuracy since not every pixel in the image is actually a pixel on the camera. CCDs used for astronomy (which cost more than your whole camera) do not do this and they still suffer from the effects you mentioned. Every exposure used for scientific work goes through a whole data reduction process that tries to remove as much noise as possible. Others have mentioned most of the process (bias frames, dark frames, and flat fields), but most astronomical CCDs also have an overscan region which is part of the CCD that is not exposed to light and is used to record the thermal noise on the CCD. This changes from exposure to exposure and from temperature to temperature (and yes I am a researcher in astronomy).

In short, there's no reason for you to care about this, and there's no chance of fixing this completely (CCDs are not digital - they're analog). There's also no way of applying the same solution to every photograph (and CCDs can change over time). Don't worry about pixel-to-pixel variations and just take photographs for their content. If you're really interested in how CCDs work, read the Handbook of CCD Astronomy by Steve Howell. Its a great introduction to CCDs and how to use them for astronomy.