Techniques, Reviews and Commentary
Why a Monchrome Digital Camera Will Never Float My Boat
Original colour image
I feel that even if say, oh.... Fuji brought out a “mono only” camera in the sub thousand dollar range I still wouldn't jump on board, not that I am saying they will of course.
Allow me to explain why.
I love monochrome images, over the years I have played with all sorts of analogue developing concoctions, printing papers and techniques. That love has endured into the digital age and too a degree been heightened by the increased flexibility that digital offers to the medium.
Monochrome is to me is all about tonality, composition and texture. In short using a mono only camera would heighten the textural characteristics and detail, shouldn't change composition at all (that’s a characteristic of the photographer) but will degrade the tonal options significantly.
For me, extra texture and fine detail always help in mono but the loss of total overall tonal control is far too high a price to pay for the gains and if I really need the extra detail I can create stitched images and employ other methods as most of my shooting is static in nature.
Colour to Greyscale version
To explain this for those who either don’t fully understand how monochrome is created and to clarify for those with questions marks over my stance lets take a little walk through monochrome history.
Early film or even earlier glass plates were not equal in their response to different colours. Most were biased towards making reds appear lighter and opposite colours, ie cyan skies appear much darker tonally than you would normally expect. Such films, often referred to as “orthochromatic”, render skin tones lighter and are sometimes still used even for portraiture. In fact there have been quite an array of modern films with extended red sensitivity, especially in the low film speed arena, Kodak tech pan being an example.
All things being equal, when presented with Red, Green and Blue areas of colour at equal brightness the human eye will see green as being lighter, red a little darker and blue quite a bit darker. BUT here there is a disconnect between how film and digital sensors record and render images and how we actually see them.
Colour to Desaturated version
Virtually all films show some exposure bias towards one part of the colour spectrum or other, those films that are generally equal in their colour response are call Panchromatic films. Hence the Kodak Panatomic X or Ilford Pan F nomenclatures. With a panchromatic film one controls the mapping of colours to tones via filters placed over the front of the lens. The operational principal is simple enough, a coloured filter will lighten its own colour and darken opposite colours, with intermediate effects on the colours that lie in between.
The most common monochrome filters were, green, yellow/green and orange, with red and blue being far less often used.
There is however a limit to the flexibility this use of filters offers. What if you want to darken the sky and also darken red tones, but at the same time you want to lighten yellows etc. Short of doing an inordinate amount of dodging and burning you can't. Basically you use colour theory to work out the best starting point as far as filtration goes, shoot carefully and go from there with the dodge and burn tools under the enlarger.
Colour to Monochrome gradient version
So what about a mono only digital camera, well it works exactly the same way. It records “brightness information” only and the overall tonal response to colour can be jigged by attaching different colour filters to the front of the lens. One point a lot of proponents miss here is, that whilst a “mono only” sensor is better a capturing light due to the lack of “on chip” colour filtration, once you start adding strong colour filters to the lens to control the mapping of colours to luminosity values much of that sensitivity advantage evaporates. The filtration could soak up 2 or even 3 stops, so your 10,000 iso monochrome wondercam is now 1250 ISO in practical terms. Red filters take the most, light yellow/greens probably the least.
In theory a mono only sensor should capture more detail because each pixel is recording real detail, unlike a colour sensor where much of the detail is synthesized by the cameras processing using the information from the differing channels. Put it this way, with a regular Bayer Sensor the camera records only 25 percent of the final image pixels in the red and blue channels and 50 percent of the green. All the rest of the data, pixels or whatever you wish to call it is invented through the clever interpolation processes built into the cameras jpeg engine or in the RAW processor on your computer.
In theory and practice, that will mean the mono only sensor will have an advantage, even if you put a filter in front of the lens. However the improvement in detail whilst real is not as great as the statistics might have you believe. Those interpolation algorithms are pretty damned clever and getting better all the time.
Green Channel only version
Some Leica Monochrom users will say, “well I don't use any filter”, and of course you can shoot sans filter, but the trade off might be a bit too great.
A couple of examples are in order.
Lets say you are shooting a subject that has red and green tones that are equivalent in luminosity. In real life (colour) they will be obviously different but in grey scale they will likely appear almost identical. Separating these tones in post processing will prove quite difficult because there is no easy selection method as there are no colour channels to use as a basis for the process. You will be stuck with making manual selections around the objects to isolate them out before applying tone curve adjustments etc, that may not be so bad if it is just an isolated object but in practice it usually involves lots of objects.....messy to say the least.
Red Channel only
Another example? Take a landscape with a lovely clear cyan/blue sky, but the sky appears too light, which is normal. You want to darken the sky, but this is going to involve making some type of selection based on drawing around the sky or using some other selection method. This will prove to be quite hard without colour to guide the selection and it is likely you will pick up halos and other inaccuracies where trees and rocks etc cut into the sky.
You could shoot mono with a polarizer filter to darken the sky but remember you lose two stops and the darkening effect is not actually the same either,the resulting sky will likely look quite uneven.
Of course sometimes you could shoot filterless, scenes which have strong tonal contrasts and limited colour palettes will likely work quite well sans filters, but I found this situation quite rare with film and likely no better with monochrome digital. In fact most of the time I find scenes with good strong colour are the ones that work best in monochrome, but I definitely need sophisticated ways of controlling how those colours map into luminosity values.
Blue Channel version
How do we convert colour images to monochrome in the normal colour digital world, well there are several ways, and the methods I use depend on the photo and the look I am after. Most conversions are however of the channel mixing variety, which simply put means we mix differing percentages of the three colour channels to arrive at a monochrome tonal look we like.
The closest method to what happens with a “mono only” sensor would be to convert the image to Lab mode and keep only the L (luminosity channel) or perhaps do a simple “convert to greyscale” process. Either way I think for most images you will find these two options render a pretty bland looking monochrome photo unless you do a fair amount of dodging and burning.
Interestingly “bland and flat” are two words a lot of Leica Monochrom critics use to describe the look of the files from the camera, I am not sure I agree at all with that but I know where they are coming from. It needs to be said that from all accounts once edited, the images do not look bland and flat.
Anyhow lets not knock the Leica, any camera using a monochrome sensor would produce files that looked that way, unless shot through filtration and post edited.
Ultimately if you shoot with a monochrome sensor you have two core options, shoot unfiltered or shoot filtered, and in either case your final post process options with regard to individual colours mapping to tonalities will be limited to dodge and burn style processes.
Now, what then are your options if you shoot in colour?
Use channel mixing and you are not just limited to fixed values of standard monochrome filters, you can use an infinite array of intermediate values.
- Simply convert to greyscale.
- Convert to lab mode and discard the colour channels
- Convert via a black and white gradient map, which can then be fine tuned.
- Convert using layers and adjust tonal mapping using the saturation tools
- Convert using one the advanced monochrome conversion tools such as the “Black and White tool in Photoshop, which is a sort of automated and understandable version of the one above.
Final version derived from combining elements from all of the above colour to monochrome conversion methods using layers
But all of this dear reader is only the beginning, because the real power comes when you start using layers and combining different channel mixes for different parts and the image, or combining the effects of channel mixing and gradient maps via layers etc. I could go on and on, personally I could show you probably 20 different ways to convert to monochrome and all of them will give a different look.
But I can tell you this, virtually every monochrome image I create is done via multiple conversion methods and layers using, desaturations, gradient conversions, multiple channel mixes and more.
And of course I can still dodge and burn to my hearts content!
And here is the good bit, I still have a colour original should the mono version not work out!
Shooting in mono only would simply make my mono life very much harder and in many cases I would not be even able to get close to the same level of subtly I can achieve with my current colour derived method. But granted my images may indeed be a bit more detailed via the mono capture route.
Some of you will no doubt be saying, well Brad your still looking at a colour screen on the camera, that’s not a true monochrome experience. Sure you are using a colour screen, but you don’t have too. Why not put your camera into mono mode, so long as you are shooting RAW you will still have a colour file to work from and the JPEG will be Monochrome, and if that helps you pre-visualize the monochrome image then you can be a happy camper
There are some deficits in shooting in colour and converting to mono in terms of quality, these have to do with noise and detail in particular but without giving too much away for now let me just say they are totally solvable and there is no practical reason why a converted colour derived image can not very closely match that from a specialized mono sensor......trust me it just takes a little know how.