Exposure

It's an excellent article explaining what "exposure" actually is and the difference between it and image lightness.

You can vary the image lightness in camera without altering the exposure.

It also explains why iso is not a direct cause of actual noise.

Alan,

One of the more thought-provoking things I've read recently, at least in my own mind was, "Exposure is the process, not the end result."

Steve Thomas

Nice write up. However, nothing is new. This process goes back to the early days of photography because of the electron trapping properties of silver halides where in an emulsion of fogged by polished silver plate fogged with a halide fume. Photons cause exchange energy with the silver halide - resulting in an energy transfer and a release of an electron. Photography cannot be adequately explained without QED.

www.imaging.org/common/uploaded%20files/pdfs/Papers/2004/AGX-0-606/30729.pdf
link.springer.com/content/pdf/10.1557/S0883769400062862.pdf

That's an interesting way of looking at it.

Which process are you referring to?

For example, the process I generally follow is to maximise the exposure within my dof and blur requirements without clipping important highlights.

I then set the final image lightness in post.

Reading through Bob's article he summarises very simply how to minimise visible noise, especially in low light situations.

"Concluding, the noise in images with a very small exposure, typically low light or ‘high ISO’ images, is produced not by electronic amplification, but by the structure of the light itself. The way to decrease noise is to ensure as large an exposure as is possible – hence the popularity of techniques such as ‘expose to the right’, which seek to maximize exposure."

where exposure was defined earlier in the article.

In layman's terms - exposure is the amount of light striking the sensor per unit area during a shutter actuation.

Bit of an odd article, again, if I'm honest. The opening is a little strange, if I go to an old book then the meaning must be right because it's been used for a long time. Hardly a scientific proof that the definition as applied to film can be applied to digital. It then jumps into mathematical equations, then into the quantum nature of light and how that affects noise. What the article seems to do is address Bob's concerns about the common technical misconceptions members of photo forums make when they get too scientific for their own good.

What it doesn't really do is give a clear and simple answer to what exposure actually is. Not to say that you can't also provide the detailed and correct scientific explanation, just that the article only seems to address the one, not the other.

Exposure is the amount of light that falls on the light sensitive media, generally expressed as per unit area to separate it from the idea of total light per picture independent of unit area. So generally we need a device that can control the amount of light that falls on a light sensitive media, and we do that by controlling the period of time and the aperture through which we allow the light to pass. We also in, the case of the camera, make the assumption that we have a lens on that is focussing light into an imge and so each expoosure is a range of different intensities that fall within the range of sensitivity of the media.

But exposure is not the settings we use on the camera, it is not f5.6 @ 1/250sec, if we confuse this with the amount of light we let in then it suggests a single value for the incoming light. Those are just the aperture and the time by which we let into the camera a range of different intensities of focussed light.

Thing is, to make this useful as a photographic process you need two further things; you need a material that has a progressive response to light, and you need to calibrate that material so you repeat the process with the same result.

Calibration is a simple concept, if we take our light sensitive media and say: if you follow process Y then X is the amount of light you need to produce 18% grey in the final image, this is basically your ISO standard. So if we photograph our 18% grey card then we must ensure that the same amount of light must always fall on our media as defined by our calibration value (ISO) and we must then always follow process Y. Of course due to ambient lighting conditions such as weather/time of day etc. our 18% grey card is always reflecting different amounts of light, so we use the aperture and shutter speed controls in the camera to limit that light and ensure that in all conditions exactly the same amount of light hits the light sensitive media, and then it will always reproduce as 18% grey.

However film photographers such as A Adams found it far more useful to use the film's threshold as the calibration point rather than the 18% grey. The film's threshold is the amount of light needed to produce the first reaction in the emulsion, if you like it is the number of photons needed to strike a molecule in the emulsion to change it and allow it to be developed. What this did was to separate the exposure X from the process (development) Y and allow you to calculate exposure X plus process Y. Or, if you exposed with a calibration value of ISO100 and know how many stops above threshhold your 18% grey card is then these are the range of development times and these are the finished values you will get for our 18% grey card in the finished image.

In digital things got both more complex for the mathematician who needs to understand the absolute process and easier for the photographer who just wants to take photos. For the mathematician the process Y became programming based on sensor output while ISO remains little more than middle grey calibration in the output JPEG. A lot of the absolute is hidden and becomes the choice of the programmers.

For the photographer though it means exposure is simplified basically to highlight and noise control. Because digital sensors have a saturation point there becomes a maximum value they can record and because of the quantum nature of light, the lower the level of light that hits the sensor then the greater the proportion of noise in the finished image. It really means optimum exposure is one where important highlights are maintained while maximisung the amount of light hitting the sensor, highlight and noise control.

Unfortunately "optimum exposure" often becomes the aim if you concentrate on the technical understanding which will produce the sharpest images with the most detail and the least noise. If it's the best the camera can do then it's the camera that dictates the output to some extent rather tha the photographer, highlights cant be clipped, noise must be minimised and no values should be clipped in the histogram!!! But rarely does this ever coincide with the images we as humans find to be the most compelling. :-)

That's pretty much what the article shows.

Especially in low light scenes minimising visible noise is a high priority for me.

It's highlights important to the photographer's aim for the image that should not be clipped, not all highlights.

Only if you restrict definitions to photos of actual physical objects do highlights need to be "preserved". If the photo is of the impression of light or other visual effects then it's the effect that becomes important.

Everyone will have their own goals for their final images.

In any case, even if the goal for the final image is some kind of effect I wouldn't clip important highlights for the final image when taking the photo