How accurate is your camera’s exposure?
Knowing the accuracy of your camera’s exposure meter is critical to achieving correct exposures by choice and not by chance. You may be aware of of the various metering systems your camera has, such as evaluative, centre weighted, partial and spot. You may appreciate how one system may have an advantage over others depending on what you photograph. Without knowing how that translates into tones in the final image, it will not be possible to reliably reproduce what you’re are photographing or deliberately change its tone.
There are various ways to make fine adjustments in post processing but if the original exposure is not accurate the overall quality of the final image will be adversely affected.
When I started photographing weddings a few years ago the advice was to under expose by 1/3rd of a stop. And when photographing into the sun under expose by 1 stop. This seemed to be fairly universal advice to protect the Bride’s white dress at any expense, but resulted in dark images that were very noisy and often lacked contrast.
To rectify the problem in Lightroom I used the auto exposure function, which would increase the exposure and reduce the value of the blacks, brightness & contrast even though these are values designated by the camera manufacturer. This improved the images a lot but they always lacked pop and were noisy.
After a lot of researching I discovered the zone system, which was originally devised by Ansel Adams – a pioneer photographer in his day. The zone system is essentially a way of measuring the reflective exposure of something within a scene in a manner that will enable you to reproduce the same tone in the final print. Ansel Adams system was quite complicated and only really relevant to Black & White film and post processing. However there is a simplified zone system especially for digital photography. This spans 5 stops from black to white, which as luck would have it translates well to the latitude of current digital camera’s sensors.
The key is to measure the reflectance of a tone within a scene and to accurately reproduce it. For example, grass will give a tone equivalent to 18% grey. If your camera’s meter is accurate taking an image of grass will produce a correctly exposed image. This is because all exposure metering systems are designed to give an exposure reading averaged out to 18% grey. The reason for this is historical when hundreds if not thousands of images were all evaluated for their average tonality and the final result was 18% grey. If I remember correctly this was done by Kodak and is why all camera and light exposure metering systems operate to this value.
Not all scenes are average. Pictures taken in extra dark or bright conditions need to be altered accordingly to produce a correctly exposed image. There are advanced metering techniques with spot metering but are beyond the scope of this article.
Based on the knowledge that the camera’s meter will average whatever it reads to be 18% grey can be used to verify your camera’s meter accuracy as follows.
To perform these tests you will need an incidence light meter, an 18% grey reference card and of course your camera 🙂
1. Baseline your test with an image using your incident light reading
Place your 18% reference card at a slight angle to the main light to avoid glare making the grey too bright. Take your incidence reading directing your meter towards the light source. These were taken in my back garden so the light source was the sun, although it was a bright overcast day.
The grey card is a lastolite pop-up exposure and colour balance.
The 18% grey, being a middle grey, is half way between black or white. Therefore with a digital image ranging from 0-255, this would translate to 127 for Red, Green & Blue channels.
I took my readings in Lightroom using the eye-dropper, which gives a percentage luminance for RGB. My incidence values, once colour corrected were Red 51.4%, Green 51.4% & Blue 51.5%. This shows the grey reference is colour neutral and that it is in fact very close to a true 18% grey, assuming that would be 50% luminance.
2. Now record images of the reference card using your camera’s meter.
The next step is much easier if you set your camera into manual mode, otherwise you will have to adjust the exposure compensation for each proceding shot. Manual is the only mode to use your camera and really take control, in my humble opinion, with a few exceptions such as photographing sports.
The key to this stage is to meter only off the 18% grey reference card! This can be done using the camera’s spot meter or zoom in and use a partial metering system.
Taking exactly the same scene as before, set your camera to an exposure of 0 EV and take the image. Then repeat after increasing your exposure by 1/3 or 1/2 stop and continue until you have gone up to a full plus one stop. You will get results similar to:
You can click on any of the above images to see a larger version with the red, green, blue values.
You will observe that the first image taken at 0EV is too dark. The second image is actually the closest to true exposure with the RGB measuring around 53%.
These test shots were taken on my new Canon 5D as I wanted to verify the meter’s accuracy before using the camera for a real job.
I was surprised how accurate the Canon’s meter was, as both my 1Ds and 10D are between 1/2 and 2/3 a stop under exposing. To demonstrate how this translates to the real world, I would set my meter 2/3 a stop above the tone of whatever I was metering off. That is unless I was doing a wedding under bright sunlight as white is highly reflective and we do not want the dress to burn out. In this case I would use plus 1/3 a stop to air on the side of causion.
Now we know the accuracy of our meter we can do a final check against a white reference.
3. Repeat the above tests against a white reference starting at plus 2EV
In the simplified zone system, white is two zones above middle grey and that translates to two stops of light and hence plus 2EV on the camera’s meter to record white. Remember if we set the camera to 0EV the result will be 18% grey wether the tone measured is white or black in real life.
The following images show you the type of results you can expect:
The image above is the correctly exposed image with the camera’s meter set at 1/3 stop above normal for the selected tone.
I noticed a key difference between my 1Ds and this 5D, which is the exposure readings only go to plus and minus two stops from the normal exposure. This is perfectly fine when your camera’s meter is accurate and you’re operating within the five zones but if it is not the only solution is to count the number of additional steps required beyond what you can see.
The white reference is not completely pure but is a good representation of what you would typically encounter.
In the last two images I have included the Lightroom histogram. As you can see, it is only the last image that is effectively + 2/3 stop above the normal exposure. The white is only just on the point of clipping.
For more information on the zone system see this detailed and quite technical description. Alternatively, if you just want to learn about taking correct exposures by choice and not chance, see the Simplified Zone System. This recommends a book that is very easy to read and understand. It starts with the basics of exposure through to the process of being able to capture accurate exposures once you know how accurate your camera meter is.
If you enjoyed this post, please consider to leave a comment below, subscribe to the RSS feed and get future articles delivered to your feed reader or via email if you prefer. You can also see more photography tips.