Myth 2: ‘If you do find scenes that exceeds the DR of your sensor, then just do multiple exposures and blend them in post processing’. Many subjects are not perfectly static and so are ill-suited to multiple exposures. Documentary, reportage, portraits, wildlife photography, weddings and street photography come to mind. Even in landscape work, you get leaves moving in the breeze, moving water (changing reflections on its surface as it moves). The you have night issues, where it can again become impractical. If you are shooting cityscapes at night, for example, the moon moves, clouds move, water moves, vehicles move etc. If each exposure is one minute, then after four or five minutes, the moon is not in the same place as it started in…
Myth 3: ‘Dynamic Range is just meaningless numbers for measurebators to play with and has nothing to do with ‘real world usage’. Again, this alludes to technical as well as historical ignorance. So, what is the historical context?
It goes right back to the dawn of photography. All photographic recording media has a Dynamic Range and this was raised repeatedly by Ansel Adams, who was undeniably a technical master. Ansel talked at length in his books about how he used water bath development to expand the effective dynamic range of film when he encountered a scene that had a brightness range that was wider than his normal film and developer combinations would be able to cater for. Water bath development would ensure that, assuming healthy shadow exposure was given, excessively bright highlights would be retained within a printable range by limiting the availability of active developer to these heavily exposed parts of the film. I have seen first hand some of the prints made from such negatives and they are both beautiful and natural. Looking at the light in the image, it is easy to see how the image would have be fatally changed had Adams not expanded the Dynamic Range of his materials. He would have been forced to choose between either throwing the backlit subject matter into deep shadow, or allowing the delicate highlights to blow out.
Myth 4: ‘If you need more than 12 stops of DR, you are just shooting stupid subjects that have no real world value’. Wrong – see Myth 3. I have given examples above where one can easily outstrip a 11.7 sStop sensor and find 14.8 stops wanting. The fact that a 14.8 sensor is still *very rarely* inadequate is not a fair argument to use to then suggest that 11.7 stops is no worse. This is because 14.8 stops is precisely 3.1 stops less inadequate and that can make a huge difference!
Myth 5: ‘Images that take full advantage of 14+ stops of DR, or where large shadow or exposure raises have been done look unnatural and over processed’. Again, this is a fallacy that is a product of people seeing poor quality processing. Nobody complains about Ansel Adams’ prints made from water bath developed negatives, or Roman Loranc’s use of Pyrogallol developed negs, in which bright highlights are naturally held back by this tanning developer. Back when I shot nothing but film, many photographers spent a lot of time working with tanning and staining developers (Pyrogallol and Pyrocatechol based). They did this because these developers had the potential to allow negatives exposed under an average brightness range to develop normally, but to retard development in hot highlights, when scenes had a much wider brightness range. Such developers, when used with experience, have potential to manage brightness ranges far beyond the likes of the 14.8 stop Nikon D810!
Dynamic Range & Pattern Noise/Read Noise/Banding
Dynamic Range and read noise are not the same thing, but they can be connected. Patterns/banding appearing in heavily underexposed parts of the file that are subsequently boosted in post processing are examples of read noise. Canon sensors typically show this in the form of lines running from top to bottom of a file when viewed in landscape format. When viewed in portrait format, the lines will now show travelling from left to right. Some camera’s files will show patterns in the form of a lattice, or grid, when dark areas are heavily lifted in post.
How are Dynamic Range and Read Noise related?
Well, a sensor with lower DR will show shadows as darker than a camera with more DR, as explained earlier. For a given rendering of highlights, a sensor with 14.8 stops of DR will show separation of values 3.1 stops further down into the shadows than the sensor with 11.7 stops. To make the shadows on the 11.7 stop camera look as ‘illuminated’ as those on a 14.8 stop camera, you will need to raise the shadows by 3.1 stops. Not only will this result in more visible noise (as the shadows on a shot exposed at ISO 100 are now effectively at ISO 800), but trying to raise detail in parts of the file that are at or very close to the noise floor may expose read noise. Unfortunately, these patterns can ruin an image very quickly and can be difficult to remove (often requiring additional software). Such ‘banding’ (in the case of Canon) will be accompanied by lots of chroma/colour noise and look plain horrible because in simple terms, the sensor has little juice left in the deep shadows. So as you can see, limited DR can result in not only bringing about lots of noise in raised shadows, but the emergence of read noise, so its a double-whammy. In contrast, the file from the high DR camera may not need any shadow or exposure lifting at all and, if lifting is required, much less noise will be visible and more detail will be retained as well.
Dynamic Range & Brightness Range at Capture and Output
The brightness range measurable in a print has no relation to Dynamic Range at capture and this is often misunderstood by those asking ‘who needs more than 11.7 stops of DR when a print is only capable of showing 8 stops in brightness range’? At capture, Dynamic Range determines what information the sensor can record without losing anything. The density range the output media is capable of only describes how bright or dark the highlights and shadows will appear when viewed and measured as a print (or on screen). Matt papers are capable of less density in the shadows than glossy papers and so have a lower density range compared to glossy paper, even when the same file is printed.
Dynamic Range and Black and White Photography
Slide or transparency film had very little dynamic range – much less than 11.7 stops! If you shoot now how you shot then, you’ll have no issues either way. However, if you are a committed B&W shooter and were used to being able to maximise your materials to capture very wide brightness ranges, 11.7 stops will feel much more restrictive than 14+. This is not only due to capture, but limitations in post processing.
For you B&W fans out there, have you noticed how your finished B&W files look when converted back to colour? They often appear super-saturated and very much more processed looking than what you would desire out of a colour file, right? This is because the file has taken a lot more punishment! I also find myself doing more dodging and burning in B&W then in colour and tend to find that a B&W file converted back to colour can have the shadows dropped more to look ‘right’. Perhaps this is because with color, we have colour to help separate shadows out. With B&W we only have density/tone. This all means that I am placing much greater demands on the DR of the sensor and the flexibility of the files in the shadows in post-processing with my B&W work.
Advantages of More Dynamic Range
1. Less shadow noise, because you are less likely to need to raise shadow exposure.
2. Less likely to need graduated ND filters.
3. Less likely to run into banding/pattern noise.
Less likely to need multiple exposures and the hassle or impossibilities associated with that.
4. You can shoot using simpler rules of thumb. When I had to ‘shoot and scoot’ with some of my documentary work, I could ensure that I provided enough shadow exposure and worry about recovering highlights later on in processing and printing. I could do this because I knew the film would retain even horribly overexposed highlights. With digital it is the reverse. We can expose to prevent serious clipping in the highlights and then undertake huge shadow recovery in post. This simply is not an option with cameras in the vicininty of 12 stops of dynamic range, because (with Canon at least) you can run into terrible banding and noise within two stops, because your subject matter may already be at the noise floor. With a 14+ stop camera, you have loads more room for manoeuvre in the shadows and great things can be done with areas of the image underexposed by 3-5 stops! Thinking like this with 11.7 stops is just not possible. I often find myself having to balance exposures on a knife edge with my Canon 5D III, whereas I could obtain far better files (of the same scene) with careless technique using my Sony A7 or A7R!
Disadvantages of More Dynamic Range
Umm, I can’t think of any, unless you want higher contrast out of camera files and absolutely refuse to do post processing.
I have written this to help people seeking understanding to have a better grasp of what Dynamic Range is and where is becomes a factor. Whether it is important to you comes down to the sort of imagery you shoot (and its brightness range), the type of output you choose, as well as your personal standards. I have also written it to dispel certain myths promulgated by those who do not fully understand the issues at hand and whose assertions are often misleading.
I am of the opinion that 11.7 stops is ample for most people most of the time, but it can be woefully inadequate for some photographers. I have just done a quick scan of my May 2015 trip to Iceland and can honestly say that 15% of the images would have been beyond reach had I shot them on my 11.7 stop Canon 5D III (I shot this trip on my Pentax 645Z, which has 14-15 stops) and that’s just regular landscape use where conditions were more often gloomy than high contrast.
Don’t forget, extra dynamic range gives you the scope to push digital files around with the same violence one could dodge and burn a well exposed and gently developed regular old B&W film.
For what its worth, go and have a look at Salgado’s digital work (he shoots Canon). You will notice that the range of lighting conditions he has shot under has changed compared to his film work. He seems less aggressive with contra jour lighting and more keen to seek lower brightness range scenes. He still produces stunning work and the printed results are still (mostly) amazing. However, you can see that he (or more likely his processing and printing team) have had to work their behinds off more often than they might…. and we are of course not privy the files that just could not be saved.
But that’s the point: Salgado is a master and has vast support behind him to apply enormous effort and skill to processing his files and making his prints. Most people don’t have the supreme talent, the time, the money, or support he has. High dynamic range just makes things easier, safer and more ‘doable’. I’m sure even Salgado would have appreciated that when pointing his camera upwards towards the rainforest canopy during a monkey hunt. At a minimum, there would have been a lot less swearing amongst his image processing team, of that I am sure.