Now I make absolutely no apologies about this being a particularly geeky post, so read on if you’d like to understand a bit about how I managed to capture,my low light Cenotes pictures that received such great feedback on the web recently.
I’ll do another post about how I edited them in Lightroom and Photoshop with a “how-to” video soon. For now though check out how I used a technique called exposing to the left ETTL to make sure I was able to record as much visible detail in a very tricky lighting situation.
Story so far…
I’m just enjoying the UK sunshine on a warm summer Sunday afternoon after returning late last night from my most recent Red Sea Relaxed trip, where once again we achieved an unbroken run with dolphins making an appearance at the Barge, and with one of our party not yet into double figures with her diving it was a particularly thrilling time.
Paul Colley has written our guest blog for this and you can see what he thought here.
So while I was winding down from another action packed trip, I was doing some research about the current thinking on exposure methodology with modern digital cameras, as you do……… well you do if you’re a weirdo photo geek like me anyways 😉
It’s a RAW war out there….
Googling away this morning doing some research for this post, I found that there were many conflicting views about how exactly to expose your pictures, with many saying left and equally many saying right.
Ok, what am I referring to here?
I’m talking about histograms, those funny little graphs displayed alongside your pictures often toggled as a choice when reviewing your shots on your camera screen, or more commonly on your particular choice of editing software, here’s an example below, screen grabbed from Adobe Lightroom.
This technique really requires you to be shooting in RAW as you need to capture as much data as is possible with your camera, shooting with a Jpeg only camera will reduce by a huge amount how effective this technique is.
Histograms are a graphical interpretation of your exposure range as captured by the camera, with either end of the graph being indicative of the maximum amount of black your sensor can manage at the right side bar, and the maximum amount of light or white info your camera can cope with at the right side of the graph.
If your image is made up of predominantly dark and shadow tones then the graph will be skewed to the left hand side, and if your shot is very bright and mostly composed of light tones then the graph will be skewed strongly to the right. The example above is a Histogram from an evenly exposed shot, that was neither very dark or very light.
Here are a couple more histogram screen grabs below of examples of dark and light pictures.
Most cameras, even quite basic ones have the ability to display the information held in the picture file data as a histogram like this, so check your manuals to see how to access it with your camera, and even if it doesn’t then as soon as you load it into Lightroom or your editing software of choice then this graph as seen above will show itself, on Lightroom and most editing software it is usually shown in the top right corner.
So why is the digital photographic community getting so het up about this admittedly quite esoteric corner of image geekery?
As far as I can make out the accepted wisdom in the world of topside,and mostly landscape photography is that you should expose to the right (ETTR) or towards the highlight area of the histogram, slightly overexposing but not clipping the data recorded, this is so that you can retain the maximum amount of shadow detail, without the darker area of the pictures becoming too noisy. This has held true for nearly ten years.
However there is a growing band of photographers who are now saying that its better to err on the side of underexposure, as you’re at less of a risk of clipping your highlight data, and sensors are much better at dealing with noise than they were years ago.
Having researched both arguments I can see the validity in both, but sometimes the picture itself and the content and composition are at direct odds with the technicalities, so I was keen to see in one specific scenario how well this newer thinking could deliver results.
Squeezing a quart into a pint pot….
The above phrase is quite apt when you consider how digital camera sensors capture and record the image data coming through your lens.
It’s all about dynamic range, a term that defines how much your sensor is able to record, and how fine the gradations, from pitch black to bright white are rendered.
As sensor technology has advanced the dynamic range of our cameras has increased and we are able to take pictures across a much wider range of light to dark tones, but the digital world still lags very far behind what we can say with our own eyes.
So the trick to getting a correctly exposed image is about assessing the type of shot we are trying to take and using this knowledge of our sensors inadequacies and placing our exposure choice in such a way that allows us to show as much detail across as broad a range as is possible at the time.
For instance if you are shooting a picture of a white cat in the snow, you would need to be careful to expose your scene with the histogram data erring to the right or white side, but without going too far as to ‘clip’ or blow out the highlights as the term is known.
This is when you overexpose too much, and there is literally no data recorded and you go out of range of the histogram on the left side, leading to ugly white areas, and if you were to make a print of this then you would actually have no ink being left on the paper, a particularly bad look.
So the trick here is to learn how your sensor behaves when shooting RAW and only just overexpose enough so that the whitest part of the cat is just within the range at the far right of the histogram.
Conversely you may imagine that if your shot was of a black cat in a coal cellar, then the opposite would be true, i.e. you’d underexpose.
This is not how sensors behave though, and they tend to produce noisy or grainy images in the dark or shadow areas of your shots, so the accepted wisdom for shots with a lot of shadow detail contained within them is to still slightly overexpose, with the thinking that you will reduce noise in those dark areas, and the sensor will best record the finer nuances of shadow gradation like this. And of course you will process the pictures with this in mind which I will show you in a processing tutorial “how-to” soon.
And this is true, this will produce finer grained, less noisy pictures with greater shadow graduations, but this doesn’t hold true for every type of dark picture. And what if your shooting a very white cat, in a very dark coal cellar?
Which is very much like shooting sunbeams and divers in a very dark Cenote.
Capturing the impossible…..
I had had the luxury of two years earlier shooting at the exact same locations and wrote a blog about my experiences. And if you’re thinking about visiting Mexico for a diving trip, I thoroughly recommend it by the way, then check out my report overview here.
At the time I had had no experience of shooting in the Cenotes, but I did have previous photographic experience shooting in similarly ‘impossible’ lighting extremes.
So I simply applied the same techniques i’d learnt many years ago shooting spotlit musicians against dark stage backgrounds, and broadly speaking this worked and got me the images I wanted on my first attempt.
This time though on a never ending quest to push the boundaries I decided to try another technique that I had been reading about, and that was to rather than pick my ISO, shutter speed, and aperture combination, based on the minimum shutter speed I thought was feasible shooting relatively slow moving subjects, i.e. from 1/8 sec to 1/30 in most cases and with an aperture of just above the maximum with my particular choice of fisheye lens, the Panasonic 8mm at around f4.5 to f5.6, my reasoning being that most lenses perform better when not wide open, but wide open enough to let in as much light as possible giving me tolerable shutter speeds at my choice of ISO
My ISO I tried to keep down to between 500 and 1600, on the previous trip I had pushed things to above 6400 in some cases just so I could get a visible onscreen image, but since then I have learnt that the manufacturers just tend to boost the sensor in camera which introduces noise at source, so I was keen to see what would happen if I cared less about this and simply picked the shutter speeds and apertures I felt I could get away with, paying little regard to how underexposed I may end up being. This was all a big leap of faith as i’d never shot like this before.
Fumbling around in the dark…..
This was all experimentation and went against a lot of perceived wisdom, I am shooting with a relatively small sensor, it’s a micro four thirds on a mirrorless Panasonic GX7, not even the latest tech in this format, so by rights this shouldn’t work very well at all.
However I had been reading about something called ISO invariance, which is simply put, a baseline ISO that the manufacturers build into their tech which dictates at what ISO the sensor is natively operating at, and above and below that ISO the software in camera is applying leverage to allow us to use lower and higher numbers above that level.
Which means that we can decide to shoot at that point and then if recording in RAW apply the necessary corrective measures within our software of choice rather than in camera, which will allow us greater control over dynamic range and shadow detail.
I didn’t actually know if my sensor is ISO invariant so this was literally and metaphorically fumbling around in the dark and hoping for the best in the challenging circumstances, but all my recent Cenotes pictures were shot like this so it wasn’t without it’s successes, borne out by the response to the earlier blog post across our social media.
The major downside to shooting like this is that it is quite disheartening at the time of capture, as your screen is showing you a mostly black image, with only the very brightest highlight details being visible, and with no realistic way of knowing at the time that you’ve actually recorded anything worthwhile.
And as is the nature of these trips I rarely have time to assess all my shots at the time, as even with small groups the day is eaten up with the more important matter of the guests pictures.
So it was only when I got home that I realised that i’d achieved a modicum of success with this technique.
And will endeavour to fine tune it for my particular camera on future occassions.
The Ups and Downs of Left and Right
For shooting in the extreme conditions of the Cenotes this technique definitely works, and even with a relatively small sensor like mine, so I imagine with a larger or full frame sensor this technique could be pushed to the full, so those of you using Nikon D800’s or 4’s and 5’s or Canon or Sony full framed sensors you’d get much better results technically than I did.
I just urge you to be brave, your pictures will be more noisy , but I’ve looked at them in close detail and the noise isn’t too bad even with a smaller sensor, and for web or smaller prints up to A4 (canvasses will go even bigger), then this technique will deliver results.
They are definitely less noisy than if id shot at much higher ISO’s to start with.
Sometimes I think we worry too much about excess noise and should care more about the composition and artistic impression conveyed, and the feedback I received about these pictures from photo geeks and normal divers reinforced that opinion in my mind.
So go on be brave and lean over to the left and embrace the dark side.
Don’t be afraid of the dark
I’d suggest you give it a go,and don’t be afraid of the dark, in a literal and metaphorical sense. See how your camera reacts when shooting RAW and try underexposing by successively greater amounts, to see what happens to the noise when you then increase exposure in software, by comparison to shooting at higher and higher ISO settings to achieve the same combination of shutter speed and apertures.
This is taken very seriously now on the various photo review sites like DPReview, and they give side by side comparisons of different sensors, whenever a new camera turns up.
For all of us shooting in the less than brightly lit world of underwater photography, who encounter low light shooting scenarios more than most, then this may well influence your next camera purchase, particularly if you shoot wrecks and other low available light subjects.
Don’t worry if you’re struggling to understand any of this, it is definitely lodged in one of the odder corners of camera tech, but by all means come up and have a chat about this at this years upcoming diveshow where I will be helping to man our busy stand.
I will leave you with an onland shot, that I took on the same trip which should also show you how much of a difference this technique can make for a more regular type of photo that most photographers could identify with.
Check back soon.