It appears that there is a degree of misunderstanding of what a histogram actually is and exactly what information it displays.

I will reference this to jpeg images for the sake of this thread and for all those ready to correct me on unhelpful technical details please allow me some licence in order to keep this as understandable for as many people as possible.

A histogram is an x,y representation shown as a series of columns representing black on the left brightness with a relative luminance value of 0 and white at the extreme right with its luminance value of 255 while the space between is filled with the remaining 254 columns representing the luminance values of all the shades graduating from black through to white.

The 18% much talked about is “middle grey” which, of course, falls in the centre of the histogram and holds the luminance value of 127.

The important thing to remember here is that any histogram is not an absolute. Any given histogram is tightly related to a particular parametric circumstance. What that means is that a histogram display in a camera displays the range of luminousities within the limits of the dynamic range of that particular camera.

If one camera has a dynamic range of twelve f/stops and another is capable of six and both were used to photograph a tonal sweep, say a surface with an luminance range of fifteen equivalent f/stops in sunlight with exposure set for an 18% grey card then both histograms would look much the same even though camera (a) recorded six stops either side of the middle whilst camera (b) only managed three.

The histogram on both cameras would show that values were recorded in all the columns 0-255 because in these cases the information they display is limited to the actual dynamic range of each sensor.

The 6 f/stop sensor in this case would record as black a luminance value which the 12 f/stop sensor would record as 9% grey.

Likewise the 6 f/stopper would record a shade as being white whereas the 12 would only record the same luminance value as being around 60% grey.

In other words the 12 stop sensor would in one shot record the same range as the 6 stop sensor could if shot as a +3,0,-3 f/stop HDR.

The 12 stopper would have twice as much shadow and highlight detail even though the histograms would probably look much the same.

That’s probably enough for now, I’ll let this sink in and settle for a bit before I continue. Rob.

Pleeeease.... I knew this would happen!!!

I asked that people give some latitude so we don't all get bogged down in the detail. This subject is confusing enough for so many so let's just please stick to the subject CONCEPTUALLY and we can all build on people's knowledge with a minimum of confusion.
The "13%" to which you refer is actually 12% and is the level set as middle grey as the median sensitivity of the sensor to compensate for the non-linear way the sensor reacts to a linear progression in light levels so that relative brightness levels recorded by the sensor more accurately mimic the reaction to light of human sight.
I used 18% not because I do not know what I am talking about but because I was referring to the reflectance characteristics of standard objects which is not the same as parameters selected by designers for sensor calibration.
I hope this clears that up. Regards, Rob.