Why Dynamic Range Is Important (Especially for B&W!)
Discussing Dynamic Range is not easy, because doing so often leads to debates ranging from ‘tetchy’ to outright flame wars. It seems that some of those fighting about this topic can have their vigour explained by a lack of understanding. However, those who genuinely understand dynamic range and are still shouting from the rooftops that ‘it doesn’t matter’, well, they must assume that everyone has the same needs as they do. I encountered this sad reality the moment I posted an article on Nik Dfine (which is very useful for correcting banding with Canon sensors) a while ago (on a forum), but its still an important topic and so here I go again!
Unfortunately, because Canon has yet to produce a sensor with very high dynamic range, some people often dynamic range debates as a threat to the status of their brand. However, the relevance of dynamic range to imagery has nothing to do with Canon and this connection will cease as soon as Canon produces sensors with more than 12 stops and a propensity for banding. I wrote about the forthcoming Canon 5D III replacement (5D IV or 5DX) here.
My interest in dynamic range extends no further than its implications for me as a photographer and the results I wish to achieve. I also believe that understanding this topic is pivotal if you want to produce the best file/print quality possible under the widest possible range of circumstances. In film days, it would have fallen under the category of ‘understanding your materials’ i.e. film, developer, paper etc and nobody sneered at that, did they?
I have chosen to illustrate this article with film images, precisely because that medium gave me the flexibility I needed for the way I shot. It is also the reason why I did not start shooting B&W digital until using cameras with much more dynamic range and shadow recovery potential than my Canon 5D II and III cameras.
So starting at the beginning…
What Is Dynamic Range?
Dynamic Range (often abbreviated to DR), describes the range of brightness a camera’s sensor is able to record and is usually stated in terms of the number of stops (of light). If exposure is made such that the brightest tone/value in the scene is just short of clipping, the more DR a sensor/camera has, the further into the shadows it will be able to ‘see’. This will be shown in the form of record separation between dark tones. The result is that cameras with more dynamic range often appear to produce out of camera RAW files that are lower in contrast, because if a person exposes to avoid clipping, the resultant shadows are less dense and better separated. High dynamic range cameras always produce RAW images that appear ‘flatter’ than a lower DR camera.
At present, the lowest commonly available big brand cameras (M43/APS/FF) have a dynamic range of about 11.7 stops (e.g. Canon 5D III) and the highest dynamic range commonly encountered is 14.8 stops in the Nikon D810 (Sony Sensor), with most APS-C or larger Sony sensors pushing out about 13.5+ stops of DR.
Just remember, Dynamic Range effectively determines the luminance range (from highest in a scene to lowest) that a camera can capture without clipping (highlights) or blocking up (shadows). It has no relation whatsoever to output i.e. prints and their brightness range (more on this later).
Dynamic Range is a product of the sensor’s design (the largest influence) plus the camera manufacturer’s processing of the imaging data provided by the sensor (which has a much smaller but measurable impact). This is why the Sony A7 and Nikon D610/D750 have slightly difference Dynamic Ranges, despite using closely related sensors (the same goes for the A7R/D800/D810).
What Impact does More or Less Dynamic Range Have?
Less Dynamic Range means that if the highlights are just short of clipping (and so record properly), cameras with more DR will be able to show greater separation down into the shadows. If a scene has a limited brightness range (such as a Welsh summer day, with heavy overcast skies and lots of rain), you don’t need much DR to be able to capture this, as the difference between highlights and shadows may only be four stops. Any camera will be able to cope with this perfectly and so a camera with more DR offers no advantage. In fact, it has a slight disadvantage, in that the file from such a camera will be even lower in contrast than one from a low DR camera and need more adjustment in post processing. This can be very quickly adjusted, however and I can only see this being significant issue for those who shoot JPEGs and are averse to post processing (especially those who live in the cloudy and rainy UK!).
In short, Dynamic Range only becomes an issue if the scene you wish to record has luminance values that are further apart from highlight to shadow (broader in range) than your camera is able to record, due to its Dynamic Range. More specifically, it is only an issue (a problem), if you need to retain tonal separation into the deep shadows, or full highlight information. If you are happy for either the shadows to block up (which may result either in slabs of black in shadows, or possibly silhouettes) or for highlights to blow out, then it is again not a problem. Sometimes we want these effects, but more often we do not. Besides, if we took the same shot on a camera with sufficient DR to record the entire brightness range of the scene, we could always process it in such a way as to throw some of this information away in post processing. It’s much easier to throw a subject that shows separated tones into black than to recover detail from where there is none.
What Sort of Scenes might prove demanding in terms of Dynamic Range?
Anything with very high contrast/brightness range. Typical examples (and this is a very brief list) would include:
- Building interiors featuring windows. There is a high difference in brightness between the interior (especially if there is no interior lighting) and the daylight flooding in. I encountered this a great deal when shooting documentary work inside bombed out buildings in Afghanistan, for the series ‘Russians and Royals’ and ‘Afghan Heroin: Not For Export’. Personally I do not like bleached out squares for windows, but with the film I was shooting, I could usually expose for the human subjects inside and then pre-flash paper in the darkroom and battle to get something out of the heavily exposed highlights. With digital, we might try to expose to ensure no clipping in the window areas and then lift shadows, but our ability (and need) to do so will depend on how much dynamic range the camera possesses. For those who see the tests online in which shadows are raised by four or five stops, these tests *are not entirely unrealistic*. While some people never take shots like this, I took a lot and had I shot the series digitally, the difference between a Nikon D750 and a Canon 5D III would have been absolutely enormous in terms of the image quality I would have obtained. Brides standing beside windows, or doorways would be another great example of where a massive brightness range might exist under ordinary circumstances.
2. Backlit trees or buildings (or anything). The sun is very bright, after all.
3. Hard shadows such as caused by direct angled sun on sidewalks and streets, with areas of deep shadow caused by buildings. Sunny day + city = huge brightness range, due to hard and large shadows.
4. Nighttime cityscapes: Light bulbs are very bright, although this is less the issue than any walls or structures that receive a ‘hot spot’ of light from a nearby street light. Its OK for the bulb itself to blow out (you can just add tone in post), but when the wall or pavement, or lamp post, tree, or billboard blows out, you cannot recover the texture and it looks awful IMO. This means having to expose less to reduce clipping in these areas, resulting in in very deep shadows in those areas far from direct illumination. Whereas a camera with high DR may show a clear separation between the bottom edge of a bridge and the underside structures and dark water below, a lower dynamic range camera may seriously struggle here and just result in a flat black block, which will ruin the image.
Some people never shoot these sort of scenes and that’s fair enough, but for those who do so regularly, higher DR makes things much, much easier at the capture stage and/or leads to far superior results. At the capture stage, one can just dial in a bunch of negative exposure compensation until clipping is gone on the histogram (or close to gone, in the knowledge that this is a JPEG histogram and highlight recovery may be possible in post), knowing you can boost exposure in post. I have to be far less aggressive with shadow recovery with my Canon 5D III compared to my other cameras.
Myths about Dynamic Range
I would like to discuss two areas where misunderstanding and misinformation often crops up online:
Myth 1: ‘If you can’t expose properly then….’ This is the usual slur on some forums (especially Canon orientated ones) against anyone wishing for more DR in their camera. Sadly, such comments only go to show that the author does not know what he/she is talking about. If one understands what Dynamic Range is, it should be clear that no amount of ‘correct exposure’ can alter the fact that a scene can be broader in brightness range than the sensor is physically able to record. While bad exposure can compound limited dynamic range, ‘best possible exposure technique’ cannot overcome insufficient dynamic range for the brightness range being recorded. When using colour film, everyone recognised that transparency film had less dynamic range than negative film, often necessitating lots of graduated neutral density filters by landscape photographers. The allowed photographers to reduce the brightness range of the subject to within the capabilities of the film being used.
Click Below for Next Page….