Histogram Tutorial

Understanding Digital

Part 1 – The Histogram (that funny thing on your camera LCD that you don’t understand!)

I’ve lost count of the number of amateur photographers I’ve spoken to who can’t get to grips with what the histogram is, or the important information it can give you ‘on the spot’ so you can make quick exposure adjustments.

Firstly, what is a histogram? Well basically it’s a map or chart which shows the distribution of the recorded luminance values for each pixel on your image sensor.

This is a typical RGB histogram:

In order to be able to make use of the histogram there are a few very simple things to understand first of all.

Firstly, the X-axis is numbered 0 to 255 for a reason. On the RGB colour scale Black is 0 and pure White is 255, it’s that simple. 255 units of Red + 255 units of Green + 255 units of Blue give us the brightest white that can be displayed in the RGB colour palette.

So with 0 at the left we can easily see that this is where our blacks should be, and that this left side of the histogram is the domain of Shadows.

As the extreme right edge represents the very brightest of whites that we can see, it now becomes apparent that the right side of the histogram is where our highlight areas belong; but there’s a catch here; if highlights actually reach a value of 255 then they will appear ‘BLOWN’ and without any detail – more later.

So that just leaves the chunk in the middle, and this is where all our wonderfully descriptive and highly detailed mid-tones should fall.

That’s exactly where the largest percentage of luminance values fall in this particular histogram, and this area can be regarded as the ‘Tonal Range’ of the image.

Where is this histogram from?

Well it belongs to this image:

Say ‘Hello’ to Rene – cute isn’t he?!

The histogram we’ve been learning from is his, and he wants it back so we have to be careful with it!

This image was taken on a very dull and overcast day using balanced fill-flash at -2/3EV just to give that little bit of crispness and contrast to the image.

There is a slight fault in that the SB800 flash used has put a second catch-light in Rene's eye and this is right at the very limit of the histogram.

Can you see that little ‘blip’ at the extreme right – that’s it circled in red.

Now if we look at the image below we can begin to see where everything goes.

The blackest pixels in the image are those in the blackest parts of Renes’ eye, and those dark pixels represent a very small proportion of the total picture area; so they occur at the left edge of the tonal range due to their colour and low luminance, and are at a very low pixel density because there are so few of them.

The background and Rene himself contain a lot of brownish tones and hues, and make up the largest percentage of the overall image, the lighter in tone or luminance they are, then the further to the right they are positioned, and the more of them there are then the higher is their peak.

So that, basically, is a good histogram – captured to your computer and opened up in your edit software this image would present you with very little work to do in the way of either shadow or highlight recovery.

Okay then, now we know the basics and where all the bits go we can now have a look at what a histogram can tell us out in the field about the image we have just captured.

Let’s consider these next couple of images:

In this first one we can see that every pixel in the image is crammed over into the left half of the histogram, and the blacks don’t fade out in pixel density, they INCREASE in density due to the sudden spike at the extreme left – classic underexposure – in this case, by around 2 stops or slightly less, say 1.7.

Quick fix – dial in +1.3 or +1.7 stops of exposure compensation, and keep your eye on the highlights on that foreground branch to the left of the histogram (circled).

In this Tufted Duck image we can see in the histogram that the blacks and other dark tones of the duck are contained, for the most part, within the ‘hump’ on the left; and the lighter tones (excluding the whites) and all the tonal variations in the water are within the "spikier" portion on the right. But the whites are blown completely - they are clearly ‘clipping’ the right edge of the histogram, and no amount of work in Photoshop will bring back what isn’t there – all the highlight details are burnt out.

Quick fix – dial in -1.3 or -1.7 stops of negative exposure compensation. You need to be careful here because the more you under expose here the further left you are going to move the ‘hump’ with all the black detail of the duck's head and neck, if you push it too far then you will ‘block’ or ‘choke’ those tones and lose the details contained in them. If you have the very latest Nikon or Canon DSLR you have the option of switching to 14bit RAW files; this can have the effect of increasing the cameras ‘Dynamic Range’ slightly and so you might be able to get away with slightly less negative compensation and save the darker tone detail.

Lastly, this image shows the exact opposite of the duck image. Here exposure has been calculated with the 6mm spot placed over that hugely reflective highlight on the fish's gill cover, and though not very clear here, the exposure is spot-on and renders all the superb iridescent colours on the head of a Brown trout.

But you can see that the blacks are massively underexposed, as is a good 70% of the entire image and the image is just ‘choked to death’.

No amount of adjustment will get this corrected for one simple reason – when you boost anything in your workflow you are in effect amplifying that particular signal AND THE NOISE contained within it. Yes you could improve things here quite considerably, but there is about 5 stops-worth of under exposure in the deep blacks; boosting them back to somewhere near where they should be is going to increase both the luminance and chrominance noise levels to rather unacceptable levels.

Quick fix – turn the subject a bit more towards the available light, or you could use a reflector or mirror to bounce light into the shadows, or as a last resort, but possibly the quickest, some fill flash balanced with the existing available light exposure value.

I hope this article has shed at least a small amount of light on what is in fact a very useful tool which most of us have but few use to its fullest advantage

That’s all for now.

Andy

Originally written for the website http://www.wildpixels.nl/