What is a histogram?
A histogram is a graphical representation of a series of values. They have plenty of uses outside photography but the histogram which your camera shows is used to graphically represent the tones in an image.
If your camera has a histogram function, and most of them do these days, it’s worth making use of it. It’s actually pretty straightforward to read and manipulate.
When a photo is shown on the camera’s LCD, we examine it to judge whether it looks the way we intend and if not we take it again. As part of the process, we consider the image’s exposure – whether it’s too dark or too light. We repeat the process of “shoot – review – adjust” to ensure we have a photo that looks as good as possible before we move on to the next subject. Our judgement of whether an image has “good exposure” will be affected by the conditions under which we’re viewing. It’s very rare to have “optimal viewing conditions” – you might be out in the bright sunshine or you may be in a darkened room. The same image will appear very differently under each of these conditions. To make things even more complex, it’s usually possible to control the brightness of the LCD screen and this too can affect whether you judge an image to be too bright or too dark.
With the brightness of your surroundings and LCD both affecting your judgement, the histogram becomes a very handy feature and using it as part of your decision making process can reduce guesswork and help you to create images which need far less correction when you get them onto the computer.
The histogram is not affected by how bright or dark the surroundings are – it will not vary depending on whether the image is taken in a dark or light room – it only ever represents the light which has entered the lens and hit the sensor. Using the histogram as a guide can give confidence in decisions made about an image’s exposure, especially when working in very dark or very light environments
Interpreting the Histogram
Histograms are made up of very thin “bars” which together represent all the tonal values between black and white. The bar on the far left represents the darkest tones including black and the furthest right bar represents the lightest tones including white.
The height of an individual bar indicates the proportion of an image which is made up of the tones which that bar represents. If we create a dark photograph (eg a photograph of the night sky), the leftmost bar(s) on the histogram will reach the top of the graph. If we make a bright photograph, the bars on the right will reach the top. To see this for yourself, just try using some “extreme” camera settings like 1/4000sec, F16, ISO 50 while taking a photo of something dark, in a dark room and you’ll see the histogram will be mostly flat with a spike on the left hand side.
Generally, a histogram which has a peak near the middle and low points at either end indicates a balanced amount of dark and light tones. If the high point is over to the left, it indicates a dark picture while a bright picture is indicated by a peak on the right. By manipulating the controls on the camera, we can affect whether it gets darker or lighter which in turn affects the shape of the histogram.
The bulk of this histogram is over to the left implying the picture is made up of dark tones
Here, the majority of the histogram lies to the right hand side, the image may be lighter than really intended.
Whilst the ideal graph shape is entirely dependent on the result that you’re after, a rule of thumb is to keep the bulk of the histogram away from the edges and, if possible, adjust your settings so the graph is spread across the whole chart area. This will give most flexibility when editing the image on the computer.
In essence, if you look at your camera’s LCD screen on a bright day or if you reduce the brightness of your LCD screen, you might be fooled into thinking the image is darker than it really is, causing you to adjust the settings to create a brighter image. In a dark environment, even a dimmed LCD image can appear very bright which may well cause you to adjust the settings to create a darker picture. It’s very easy to ignore the conditions under which you’re viewing your image when going through that “shoot – review – adjust” process.
If you see a spike on the far left or far right of the histogram it indicates we’ve exceeded the range of tones which the camera can distinguish and black (or white) has been recorded instead.
If you were expecting to take an image containing a range of very dark tones but your camera settings were not quite right, you might pusha number of these tones into the “black” zone, creating a spike on the left of the histogram. Once stored as black, no amount of editing will be able to invent the range of tones you knew existed in the original scene.
Conversely, if you take a picture of a sunset and your settings mean you overexpose the image, you’ll find that an area of your image around the sun is recorded as white with the histogram showing a spike on the right hand side. When you lower the exposure on the computer to correct the image, you find that everything is the same tone and you cannot bring back the beautfiul tones and colours which you originally saw.
The camera is capable of recording only a limited range of tones between black and white so if you expose for the very brightest areas, you might well find that some areas are recorded as black while adjusting your settings to record details of a dark subject might well mean you just store just white for the sky. Its up to you as the photographer to decide whether this is acceptable or whether you need to recompose or perhaps use a technique to brighten the dark areas (lighting) or darken the bright areas (filtering).
How do I find the Histogram?
Each camera is different of course and some cameras may not have a histogram option. Some cameras are capapble of displaying a histogram of what the camera is seeing before a photo is even taken.
Generally speaking, the histogram is part of the playback menu in your settings.
For Nikon cameras, use the multi-selector button – while in playback mode press the up and down arrows to scroll through the available pieces of information. One of these should be the histogram unless you’ve switched it off in the menu system. If you zoom into the image, you’ll find the histogram changes to reflect the tones within the area shown on the screen.
In playback mode, it’s a case of pressing the info button. You may need to press it a couple of times to see the histogram
For Sony cameras, you’ll generally find the histogram when you press the “DISP” button. Again, you might need to press it several times. If you dont see the histogram, be aware that it’s possible to turn off the histogram view option in the menu system.
for Olympus cameras (eg the OM-D EM10ii), you’ll likely find the histogram by pressing the Info button numerous times. As with the other camera makes, it’s possible to switch off the histogram option so bear this in mind if you cannot find it straight away.
If you’re lucky enough to have colour histograms on your camera, this is an added feature which allows you to see the brightness of each of the 3 basic colours (red, green, blue). It allows you to check colour balance. This may be useful if you happened to be viewing the image while using tinted glasses or perhaps when you’re in a room lit only by coloured light bulbs. Most of the time this is not the case but be aware that if, for example, you saw a spike in the red histogram, off to the right hand side, you may well be losing detail in the very brightest reds of your image while the other colours may be fine. You could in this case consider reducing your overall exposure.
Histograms are not perfect
At the time of writing this, I dont think anyone has incorporated a histogram which represents the RAW image data. What you see always represents a JPG version of your image. This means, if you’ve told your camera to use a vibrant shooting style or perhaps a style with increased contrast, your histogram will represent THAT image while the RAW data remains unprocessed.
Expose to the right (ETTR) is a technique which involves choosing camera settings which cause an image to be slightly overexposed and the histogram to be “pushed” over to the right hand side but not going so far that highlight details are lost.
The theory is that the camera records fewer distinct darker tones than it does lighter tones and therefore, you’ll have more chance of recovering distinct tones from an overexposed image than you would from an underexposed image.
It’s suggestted that each new bit-level in a camera’s storage capacity is used to represent a new exposure stop of tones – since each bit doubles the number of storable values, double the number of tones can be stored each time a bit is added. Today’s cameras are capable of using 12 or 14-bits when capturing an image.
There are conflicting arguments about this theory which suggest that no matter how many bits are used, the range of tones between black and white is always spread evenly across the range. This argument suggests that it is best to acheive optimum exposure on the camera rather than aiming to over-expose. This approach makes more sense to me… but perhaps the argument for using fewer slots to store darker information has not been fully explained to me. I’d be interested to hear a technical explanation of why it works this way so please leave a message in the comment section if you have an idea.
Noise (Electromagnetic Reciprocity)
Whatever your opinion on the ETTR situation, noise is something we should aim to avoid in order to give us flexibility when editing. The wrong values can be recorded when either too much light hits the sensor and photos are left to “splash” from one area of the sensor to another or when there is such a slow trickle of light coming in through the lens that the sensor is “fooled” into recording static or recording the wrong values (noise). There’s a lot of information on reciprosity on the internet if you’re interested in finding more.
In essence, to avoid “bad data” it’s worth trying to keep your histogram in the mid-tone area and avoid creating spikes at the far left or right of the available box because this may be where information is lost.
Comments on this post are very welcome!