Near misses

_3501970 copy

Today’s thoughts are on a slightly unusual topic, especially;y given my usual obsession with curation – after all, your audience can only see the work you choose to show, not all of the work you shot. I also realise that I say plenty about what I believe a good photograph should include and exclude, but not a lot about why some things don’t work – and worse, what constitutes the kind of near miss that I’d reject (or at very least, not show at all). So, at the risk of showing the ugly side – today’s post is illustrated with images of mine that seemed good in theory, but didn’t make the cut in execution for whatever reason. It’s probably also helpful to talk through the initial idea at capture and some of the context – this is not always obvious, and often the reason of an image’s failure to meet the intent of its creator.

Let’s take the header image, for starters: in person, there was very strong layering created on the physical background objects due to the angle and intensity of the sun; the shadows were perceived to be almost as dense and solid as the physical objects themselves. Moreover, the sense of spatial separation between shadow, physical object and reflection off the floor was a lot stronger than in the image, no matter how it was processed. I had to increase black density to give the shadows the same sense of solidity, desaturate to create abstraction and remove the distraction of certain color highlights in various portions of the image, but somehow lost that sense of spatial separation. I don’t believe it’s a tonal zone problem, because the shadow, reflection and physical object zones only overlap just enough to create continuity in the image (say I-IV, IV-VII and VI-X respectively).

The problem is actually a physiological one on the part of the viewer – both viewer of the scene and viewer of the image – in that the focal planes of the various elements are slightly different (reflection effectively further away) but the overall focal distance is quite close, meaning that 3D spatial perception from two eyes comes into play. This is a large contributor to our perception of depth and dimensionality – especially when it comes to reflections in objects, since they are further away than the (physical) foreground. Using depth of field cues to suggest separation does not work as you lose definition that your eyes have – and which creates that sense of surrealism of superimposed objects or images. Conflictingly, I am attracted to these kinds of subjects for precisely that reason; unfortunately, they rarely work in 2D capture.

_3502969 copy

The arches are not an obvious rejection on their own. The image has strong geometry and underlying structure, there are no messy cut points, the sense of abstraction is especially good in the upper-mid section where the shadows of the vaults are almost interchangeable with the real elements…so why is it a near miss? It’s a question of external context: firstly, there were other images in the set which I felt characterised the overall building or specific design detail better (i.e. this particular image was in no-mans-land: not specific enough, but also not general enough to get a solid overview). Secondly, it was the upper-mid section of intertwined curves that attracted me: but including anything below for context made them feel insufficiently prominent; excluding anything resulted in messy cut points due to the horizontal supports (two rows visible here). And changing vantage point to have these aligned in such a way as to be cleanly excluded by the bottom of the frame edge was not physically possible.

_3500231 copy

Here, we have another spatial problem similar to the first image, except there are no prominent virtual images (shadows, reflections) to take into account – there is simply too great a depth of field required to execute without focus stacking. Yet even if this was done, we would lose the sense of repetition and spatial separation from the subject being physically close to us and the two-eye 3D perspective; reduced DOF cues once again fail (and once again proving that shallow DOF is not the solution in every situation to create what most people incorrectly think of as ‘depth’). The closer a subject is to our eyes, the greater the perceived separation because left and right eyes are seeing increasingly divergent images as the difference in angle to subject becomes greater (think of a triangle with a fixed base: this being the distance between your eyes, which can’t change, but the distance to the subject can. You’d also think I’d have learned this lesson by now, but there have been enough exceptions with images that did work that I keep persisting to find them).

_3502243 copy

Conceptually, I was aiming for dark and menacing with only the headlights and a silhouette providing definition. The difference in color temperature between inside and outside was deliberately left in to create a bit more mood. I can only use laziness as an excuse: I was on my way to a meeting and thought I’d try this with my phone rather than going back for a camera; it would have been much better with longer perspective to flatten the car against the plain part of the wall so the M&E stuff on the left could be separated or later retouched out. Moreover, I’d have a bit more dynamic range to work with later to dodge the key line highlights in the car. Basically, it works, but not as well as it could had I done it ‘properly’.

_PF04592 copy

The banana image has a good variety of shapes, textures and a harmoniously warm color palette, but the level of abstraction I was aiming for between the organic and inorganic forms doesn’t feel strong enough; it needs a bit less structure and more abstraction. There are also fairly large empty patches that do balance (white bowl side, dark pink wall, bottom horizontal band)  – but at the same time concentrate the subject interest into the centre of the frame. I realise this could have been solved by using a square aspect ratio to better fit the key elements (banana, bowl, chair) – but then you’d land up with a very literal still life-type composition and lose the dynamic line of the table edge vs the bright reflection. In hindsight, this one needed a fully flat perspective to work – a slightly longer lens, higher camera position and some tilt.

_3504098 copy

Compositions involving water always seem more dead in stills: the water is simply not moving, or blurred; that sense of rapidly flickering patterns is nigh on impossible to capture (it can be done, but requires something reflected or behind the surface to reinforce that impression beyond just the ripples themselves; they’re not enough). I also find it tough to balance the partial reflections of the object in the surface from frame to frame; because of both the ripples and larger scale undulations, the length of the reflection continually varies from instant to instant, meaning that it’s almost always either too short or in this case, too long – there’s not enough space between the perceived end of the reflection and the edge of the frame relative to the top of the boat. Long exposure would be a solution to get an averaged length of reflection, but then your subject won’t be static and there is rarely just enough motion to look deliberate – usually the edges just look smeared and out of focus.

_3503338 copy

In person, this cloud formation was extremely imposing: very large, very dominant, and with intense edges. Here we don’t have a 3D spatial problem (subject effectively at infinity, left and right eyes seeing the same thing) but we do have both a translation problem and a lack of physical cues given to the viewer of the image as opposed to the object. The translation problem is one of very bright objects: our eyes don’t tend to clip to white until much higher levels of brightness than a camera sensor or monitor. Even if we set the highlight rim of the clouds at absolute white (255/255/255), the only way to produce sufficient contrast to be perceptually the same as reality is to significantly darken the rest of the scene. Yet this does not feel right either, since the whole scene was blindingly bright – there is simply a limit to how bright our reproduction media can go, and even the brightest monitor isn’t sufficient in this case. Secondly, the sense of size is lost not because there’s no scale – we could print the image very large – but because of physical sensations; you don’t have to bend your neck upwards to see the scene, or scan from side to side or left to right while viewing it – we can take it all in at once, and worse, in a postage-stamp sized chunk on a mobile screen.

_8525625 copy

Usually, the challenge with aerial images is getting the color right: the atmospheric layers and haze, pollution etc. tend to filter out a lot of the warmer tones, leaving you with nothing much but blues and low contrast; when you try to enhance this, you’re left with an extremely saturated mess that is stylistically interesting but has little resemblance to reality, and loses the personality of the place you’re trying to capture. In this case, the airplane window and skies were clear enough and the altitude low enough that I didn’t have this problem and was able to add a warm tone to the shadows to restore the reds; the limitation was one to do with a window of opportunity. I fly this route often enough and have enough control over which flight I choose that I can usually get the right time of day and the right side of the plane; the clouds were an interesting bonus that made this unusual (in my mind, cities peaking through clouds conjure all sorts of romantic notions of travel that simply don’t exist in today’s world of low cost carriers) – but the same clouds happened to be in the wrong place over the city, dropping key landmarks into shadow, and there simply wasn’t enough wind for them to move away in the time the aircraft passed by. Sometimes, all the elements are present, but luck isn’t on your side – but that’s why we keep trying, I suppose… MT


Visit the Teaching Store to up your photographic game – including workshop videos, and the individual Email School of Photography. You can also support the site by purchasing from B&H and Amazon – thanks!

We are also on Facebook and there is a curated reader Flickr pool.

Images and content copyright Ming Thein | 2012 onwards unless otherwise stated. All rights reserved


  1. Wayne Crauder says:

    Very insightful. I have been told that by photographers that my lack of depth perception (due to an injury) made me a better photographer. One contest I came in second at, I lost to a photographer with one eye. And, I still take plenty of near misses that I spent too much time trying to process.

    • It would certainly help parsing the 3D world into a 2D representation, though I’d rather not give up depth perception if I had a choice…

  2. Stephan says:

    Very interesting read. It’s rare that pro photographers let us see what we usually are not supposed to see.

    I would be very happy if this became a new series with a new article every now and then. I guess there is a lot to learn for hobby photographers like me when someone who knows his stuff talks about the fine details that seperate a keeper from a near miss.

    • I got the inspiration from the Magnum Contact Sheets book – a really good look at both the rejects and the curation/thinking process, as well as providing background context to the action. Highly recommended.

      Sorry, no intention to make this a series. I’m sure you can understand why I’d rather not make a popular (read: high search result) bunch of articles with bad images attached to my name! 🙂

      • Stephan says:

        I fully understand 🙂
        But thank you very much for the book advice. I will definitely read it.

  3. My best photos, are still pale in comparison to your misses.

  4. Jos Martens says:

    Congratulations on your courageous move to show us your decision process on hit or near miss.Failure and near failure are for sure good instruments of learning. Excellent teaching.

  5. Thanks for this behind-the-scenes post! Re the first and third images: Can you clarify the spatial separation issue a bit more? For example, why wouldn’t a wider lens with tilt-shift (assuming one exists) have addressed the issue? Is it because it would’ve weakened the layering/superimposition effect?

    • No, because you can’t recreate the effect of stereoscopic vision with only two dimensions. Fundamental physical differences between human eyesight and the way a camera works, unfortunately. It can be an advantage or disadvantage depending on how you compose; it just takes some training to see that way (or literally close one eye).

  6. I will be coming back to this exercise several times to better absorb it. Knowing that an image is jut not quite right is good but knowing why is even better. I usually know why but hopefully studying this exercise will help me see things sooner, like before I snap the shutter. 🙂

    • And even once you know it isn’t worth hitting the shutter, it’s still worth curating the ones that do make it through to achieve a consistently higher level overall…

  7. Very thoughtful, as usual. I appreciate your discussing your “near misses,” it’s helpful for my own ongoing learning how to visualize & how to translate that to the image – and why it might not work.

  8. Pierre Lagarde says:

    Hi Ming. Maybe we all have our own criteria with curation though and it probably change thru time. Anyway, as always, it’s a thoughtful article with some inspiring images from you.

    • Pierre Lagarde says:

      “changes” sorry, too early sunday morning 😀

    • I find “The bar” definitely raises over time – a bit because of experience, a bit because you don’t want to make the same images you did before…or if you do, they have to be a lot better to justify keeping them!

      • Pierre Lagarde says:

        Right, and maybe also when you’ve been using the same specific media for years, your audience is more and more “picky”. Though, I keep an eye on needing to relax sometimes ;)…

        • That depends on whether you are producing for yourself, or a ‘client’ – be that an audience or a paying employer, I guess…

          • Pierre Lagarde says:

            Well, I didn’t even think of working for a client ;)…

            • Not necessarily a paying commercial client per se, but ‘client’ in the sense of audience: who is the image for?

              • Pierre Lagarde says:

                Well, I must say in my mind, client means only paying client ;). Anyway, it’s a quite different purpose indeed.

          • Wayne Crauder says:

            I curate to different levels. There is my level of something I want to hang on a wall or give to a friend to hang on their wall. Then there are prints I can sell through the gallery. Both have to be something special people want to live with. The third level is social media where they will see it once and it will give my audience a good experience for a minute and then they will move on. The final level is personal memories whether of my travels and family of working on a wedding crew. The photo will have an emotional memory that goes with it and it is more important to capture the moment than to create the emotion from the composition and light and processing for the other 3 levels. Although adding emotion to those photos will still make them better for others who see them, they still have a head start. Sometimes though it becomes like a formula and it hard to get myself involved. If I am not involved it what I shoot, the client may be happy, but I seldom produce a result I have pride in. Then I start experimenting and the failures teach. But over time, those failures give rise to something new I have learned.

  9. Beauty and poetry.
    Well done !!!

%d bloggers like this: