Reflective Meters/18% Grey

D

Dredd123

Messages
214
Edit My Images
Yes
Hi,

I've been trying to get my head around the whys and wherefors of 18% reflectance and the light meters built into todays typical DSLR's for a while. I wanted to put what I have learned into my own words, and for someone in the know to cast an eye over it and see if I've understood correctly. Here goes:

At the end of the nineteenth century, some scientists at Kodak looked at thousands of black and white photographs, and decided that the average tone of all of those photographs was around 18% reflectance. Therefore, they decided that if a reflective light meter was calibrated to suggest aperture, shutter speed, and ISO settings that would expose the scene put infront of it so that the average reflectance of the scene was 18%, this should cover most bases.

Have I got this right? I understand that you, as the camera user, has to deal with scenes in the real world which this methodology does not work for (snow, sand, black dog), and you have to work around this by using exposure compensation, or dialing in the aperture, shutter speed, and ISO settings directly in manual mode, but I just wanted to understand where the 18% reflectance came from in the first place.

Cheers,
Dave.
 
Dredd123 said:
Hi,

I've been trying to get my head around the whys and wherefors of 18% reflectance and the light meters built into todays typical DSLR's for a while. I wanted to put what I have learned into my own words, and for someone in the know to cast an eye over it and see if I've understood correctly. Here goes:

At the end of the nineteenth century, some scientists at Kodak looked at thousands of black and white photographs, and decided that the average tone of all of those photographs was around 18% reflectance. Therefore, they decided that if a reflective light meter was calibrated to suggest aperture, shutter speed, and ISO settings that would expose the scene put infront of it so that the average reflectance of the scene was 18%, this should cover most bases.

Have I got this right? I understand that you, as the camera user, has to deal with scenes in the real world which this methodology does not work for (snow, sand, black dog), and you have to work around this by using exposure compensation, or dialing in the aperture, shutter speed, and ISO settings directly in manual mode, but I just wanted to understand where the 18% reflectance came from in the first place.

Cheers,
Dave.
Click to expand...

That's at the heart of it, but quite a few things follow on from that assumption. 18% grey is 18% reflectance, it's a darkish mid grey, elephant grey ;)

What it means is that films and processing systems are produced to put 18% grey in the middle of the tone curve (known today as the dynamic range). This gives brighter highlights and darker shadows the best chance of falling on a good part of the curve.

Further, and given that in very broad terms film is reckoned to have a maximum of seven stops from top to bottom of the curve (dynamic range of 128:1) these assumptions put pure white at about 3-3.5 stops above 18% grey.

This works very well for slide film, but since negative film has a secondary printing stage where quite a lot of things can be modified, you can often get a better result by ignoring 18% grey and picking a key tone to pin to the middle of the curve (or anywhere else for that matter). This is the basis of Ansel Adams' famous Zone System.

Digital works differently again, somewhere between the two in practise. With post processing you can modify the result quite extensively, but the key difference is that negative film can take a substantial amount of over exposure and still retain some usable highlight tones (if not optimum). With digital, there comes a point where white just blows to pure white, with nothing left.

The point at which digital blows to pure white varies quite a lot between different sensors, and still further with the post processing that can be applied. Manufacturers exploit this by adjusting the calibration of their metering systems - some may be 18% grey, others 15% or 12%. Estimates vary and nobody is saying. It's a moot point though, as it matters less what figure is actually used, more that it's a constant. 18% grey is universally regarded as a reference point, but mainly for historical reasons.

The expose to the right (ETTR - of the histogram) technique is a digital nod in the direction of the Zone System, in that it deliberately shifts the tone range to better exploit the sensor's potential (which can be more than seven stops). The tones are then shifted back to the correct value in post processing, but with more tonal detail retained particularly in the shadows.

ETTR is a very effective technique, but it differs from the Zone System in that it only pegs the highlights to the top of the dynamic range, instead of pinning a key tone (any tone) to any position on the curve. And also, because of the limited highlight headroom, you can't go too far or you'll blow it to bits.

The big advantage of digital is that you have an LCD that shows the actual exposure - not a guesstimate calculated by the metering system and all its numerous variables (eg vignetting, diaphragm accuracy) and assumptions. Even better, you have a histogram derived from it, and the wonderful invention of blinkies (highlight over exposure warning) that shows you in critical detail exactly which areas of the image will blow. Using these wonderful tools, we can adjust the exposure pretty much to optimum (whatever that may be) without any guesswork and regardless of the meter's calibration, other variables in the system, or the tones in the scene.
 
Dredd123 said:
Hi,

I've been trying to get my head around the whys and wherefors of 18% reflectance and the light meters built into todays typical DSLR's for a while. I wanted to put what I have learned into my own words, and for someone in the know to cast an eye over it and see if I've understood correctly. Here goes:

At the end of the nineteenth century, some scientists at Kodak looked at thousands of black and white photographs, and decided that the average tone of all of those photographs was around 18% reflectance. Therefore, they decided that if a reflective light meter was calibrated to suggest aperture, shutter speed, and ISO settings that would expose the scene put infront of it so that the average reflectance of the scene was 18%, this should cover most bases.

Have I got this right? I understand that you, as the camera user, has to deal with scenes in the real world which this methodology does not work for (snow, sand, black dog), and you have to work around this by using exposure compensation, or dialing in the aperture, shutter speed, and ISO settings directly in manual mode, but I just wanted to understand where the 18% reflectance came from in the first place.

Cheers,
Dave.
Click to expand...

You are right in so far it was an empirical way to derive film speed.
Indeed many settings were used and the results judged after the best possible print was made from each. So it took into account the Printing method too.

At that time nothing was standardised. Every manufacturer had his own idea about paper grades and film speeds and film development.

What is interesting is that when the experiments were done by Weston, By Kodak and the American standards association, and also the geometric scale used by Din and Ilford and the British standard, they all fell with in one third of a stop of each other, even though the methods used differed in detail.

It was found that this Film speed correlated with a point above fog level on the Characteristic curve derived scientifically.

The 18% point was more a convenience than an accurate or scientific point.
It was found that the average reflectance of an average scene lit by sunlight did equate to about 18%. But of course that gives neither a film speed nor an exposure index. ( an exposure index is a "personal" film speed found and used by a photographer for a particular film and process.)

The 18% point give you a simple means of establishing where an average Grey falls on a given characteristic curve.

It is also used amongst other values to set the sensitivity of an exposure meter of any sort. It is in reality no more than a convenient means of establishing a standard tone.

The Histogram is quite different to a characteristic curve.
The Curve shows what a film can achieve with a certain development showing contrast (slope) and the maximum and minimum density against exposure.

The histogram shows the distribution of pixels with equal value in a scene.It is not actually very valuable in exposure estimation unless one also compares it to the scene itself. a photo of a small black cat on a white ground would show a small peak on the left and a massive one on the right with virtually nothing in between. It would do much the same for quite a wide range of exposures.

Histograms are very useful if used with thought.
 
You must log in or register to reply here.
Back
Top