Achieving visual consistency

_G001455 copy

One of the questions I’m asked also (unsurprisingly) happens to be one of the biggest challenges for a lot of people: how to achieve visual consistency across multiple systems/ cameras/ media, and across multiple subjects. Though the latter is really getting into the question of what constitutes style and how can one consistently apply it, there are still things you can do to ensure that you are in control of the final presentation: not your camera. I certainly cannot tell a client ‘sorry, it looks different because I used two different cameras.’

_M9P2_L1000243bw copy

Light and composition: the photographic fundamentals
There is simply no way to escape the fact that all photographs are visual: this means that what your audience sees are solely visual cues, nothing else; if the perspective is approximately the same, the level of contrast is consistent, and the direction of light is consistent, then it’s actually extremely difficult to tell what equipment was used if all of the images are mixed together and reproduced at a size that doesn’t exceed the limits of the lowest camera. In order to have control over this, you need to have control over the actual image making process: images must be first made in your own mind before they can be translated (made permanent, perhaps?) to a photograph. (I break down this process in a systematic way and in much more detail in the Making Outstanding Images video series.) If you consistently train your eyes to see the same way – noticing strong color contrasts, for instance; or juxtapositions of texture; or potentially low key scenes only – then you’ll find the end compositions are remarkably similar. Metering plays a huge role in this, because more than anything, the overall brightness of a scene determines the mood impressed upon your audience.

_M8_L1022658bw copy

Have an end in mind
It’s important to know how you want your final output to look: if not, the image you capture won’t necessarily contain enough information for you to get to the right end point. Then you’re going to start experimenting in postprocessing (“it’s grainy and dark so I’ll make it black and white” is perhaps one of the most common examples) which results in each image being different in mood and feel. Have a vision; I always shoot to produce the files I need to do the necessary processing to get to the end result I want. It’s also one of the reasons I will never release unfinished raw files: that’s like a chef serving partially cooked food.

_5018628 copy

Learn the idiosyncrasies of your cameras
Once you can clearly visualize the final image in your own mind, the next step is wrangle the camera into doing what you want it to. This means taking control, and being fully conscious of how the camera is behaving in any given situation: you want it to behave the way you expect, not deliver a surprise at the decisive moment. It is therefore imperative to practice, test and experiment with your gear to see how it behaves under a variety of situations; this allows you to adjust whatever settings and custom functions are necessary in a non-critical environment. Though most modern cameras are quite well behaved, the complexity of autofocus and matrix metering systems can often produce strange results in the real world if the camera happens to encounter a situation which the designers never tested it for. It’s also worth noting that the higher the resolution, the more critical autofocus accuracy and predictability becomes – otherwise your image may well be a large but garbage file if things are even slightly out of focus. Unfortunately, the only way to truly be certain of the outcome is to go entirely manual – that’s fine for exposure, but modern DSLRs are somewhat challenging to focus due to their poor focusing screens. The good news is that EVF cameras tend to actually be easier to focus due to peaking, showing actual depth of field, magnification, stabilizers etc. On top of that, they can also easily show areas of overexposure – which DSLRs cannot.

_8015135bw copy

Select your lenses carefully
There is a reason why conventional wisdom dictates spending more on glass: there really wasn’t much optical correction possible in the days of film. Even so, there are some optical aberrations we can fix, and others we can’t: distortion, vignetting, some lateral chromatic aberration, and to some extent, color, are all fine; in fact, a lot of manufacturers already apply software correction in-camera. There are some things that cannot be fixed, however: resolving power, microcontrast and tonal gradation, for starters. If you want images that have that consistency of ‘bite’, it’s best to start with lenses that can deliver those results in the first place. This is one of the reasons I have so many Zeiss lenses – they’re available in multiple mounts, and deliver a consistent ‘look’ across the board, regardless of system – we’re talking specifically about resolution, color and microcontrast. Even for a system such as the D800E for which I run a complete mishmash of lenses – I do so because the raw files deliver a fairly consistent starting point with all of the tonal information I need to get my end results to a consistent level. Granted, PS is a great equalizer in the right hands, but why make life difficult? I’d rather spend time shooting than processing…

_RX100_DSC2507bw copy

Consistency of workflow and software packages
…And that neatly brings me to the next point: even though certain specific software packages might deliver optimum results for a given camera, limitations within those software packages mean that they either a) do not work for other brands, limiting your lens and sensor choices; b) are missing mission critical tools; c) lack local adjustments. To this day, the only software package that has both the necessary flexibility and ability to deliver consistent results across the range of cameras I use remains Photoshop and Adobe Camera Raw. It does not always deliver the best possible raw conversions, however I find that it does deliver the best consistency of final output. Though you can suggest that I do the conversion in one software and then do final finishing in PS, there are tools within ACR which are lacking elsewhere – the extreme flexibility of gradients, for instance – which are very difficult to replicate afterwards in PS. And on top of that, there’s the time factor: the more time spent messing around with software, the less shooting you’re going to do. Professional demands on throughput aside, given that most of the image is created in camera – postprocessing can only enhance the presentation, not add missing light or subject elements – it would make sense to allocate more time there. I’ve spent plenty of time already refining this workflow; to save you the time, it’s available in introductory and intermediate flavors.

_G001776 copy

Postprocess in close proximity
Related to consistency of workflow is how you do your processing. Consider the output of one shoot – say 200-300 files. If you process them at one sitting, then fatigue might well compromise results by the time you hit the second hundred. If you process them months apart, it’s going to be difficult to remember what you did with earlier batches, especially as there may be a lot of adjustments to make if you want to achieve a certain look. I find that processing them in batches as frequently as you can – stop when you get tired – is the best way to balance quality, speed and consistency. On top of that, if you leave your processing for too long, you may well be unable to remember the final image you visualized at the time of capture.

_A_DSC0629 copy

Avoid filters and software effects
By using somebody else’s presets, you’re effectively handing over creative control to the software developer. Even if by some means of magic I created a ‘Ming filter’ that made any image look like one of mine in color, contrast and light, I would not encourage other people to use it: simply because the way I see ties in to the way I shoot which again leads to a certain end result. Different compositions may not suit that style. Simply, everybody sees a different way, leading to the end intention of every image being different – and the circumstances at the time of capture are always different. So how can you apply the same filter and expect a consistent (and good) result?

IMG_3203b copy

Shot discipline
I’ve said a lot about this already in a couple of previous articles here and here. Shot discipline breaks down into a few things: stability, focus, exposure, and resolution. Stability is about ensuring that you do not move in a way the camera can resolve; either by keeping shutter speeds high or having a very solid tripod. Minimum shutter speeds for handholding are proportional to pixel pitch. Focus is self explanatory, I think. Correct exposure for digital is all the way to the right of the histogram, to the point where a small area just clips; you need to do this in order to have the most information to work with in postprocessing. Finally, resolution isn’t about your sensor: it’s about using your lens at optimum apertures for both the lens’ design, as well as the pixel pitch of your sensor – i.e. before diffraction starts reducing resolution. The smaller your pixels, the sooner this happens. For instance, optimum working apertures of a 4.8-micron D800E with an f2.8 pro zoom are just f4-f8, or perhaps f11. In essence, the higher the image quality potential of your camera, the harder you’re going to have to work for it. Smaller and smaller ‘misses’ in focus, stability or exposure will be more obvious at the pixel level; I suggest picking a resolution that is perhaps one size up from your normal output size, otherwise you’re going to be frustrated every time you open a file. There is no point in dealing with larger files that contain no additional information.

_8040518 copy

A rigorous editing process
This step is the one that’s most often missed: your final control over an image is a simple one: show or not show. We as photographers are only judged on what we show, not what we shoot. If it doesn’t ‘fit’ with the intention, theme, or the rest of the set, then don’t include it. Simple, right?

Collectively, all of the items on this list make up what is thought of by the pros as ‘workflow’ – from capture to output, whether that output is digital media, print or even film/ negatives. The more defined your workflow is, the more easily repeatable you can make it; and that almost reflexive handling of situations in the viewfinder and in PS is what’s required to achieve consistency. I’m going to end with one note and a little demonstration: if you click on any of the images to take you to their flickr landing pages, you can see that none of the images in this post were shot with the same camera – they weren’t even all digital :) MT


H2 2014 workshops now open for booking – Making Outstanding Images San Francisco, Chicago and Venice; Masterclass San Francisco and Venice – click here to book or for more info


Visit the Teaching Store to up your photographic game – including workshop and Photoshop Workflow videos and the customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!

Don’t forget to like us on Facebook and join the reader Flickr group!


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


  1. Thanks for your essay about visual consistency and photographic style. I’ld like to add: a huge challenge is (to my humble opinion) to work in different styles for different clients … to keep these styles in your mind and be able to reproduce them, whenever you get a new job from client A, B, or C. A comparison of different styles of various companies shows that there exists no linear scale of quality. Shooting with a Lomo-camera on the cheapest film from the nearby drugstore can be appropriate for a company which produces … let’s say fashion for young people whereas for another client, e.g. an architect one might need large format films to show the finest tonal variations of glass and concrete. Style depends on the job. Form follows function. Best regards from an enthusiastic reader of your blog.

    • Absolutely – and sometimes they want you to match their existing style, which is the biggest challenge of all. It’s rather counterintuitive. Or, one can just try to take client which hire you for your own work…

      • Yes, in many cases clients will choose a photographer, whose style fits to their corporate identity. But it could also be a challenge to be asked for something completely new.

  2. I love the photos and great article!

  3. Apropos “.. idiosyncrasies of your cameras”: I believe this deserved the first place in numbering your above points (not that I’m even close to your performance). But it seems to start with learning what happens when one looks through the viewer as against what the camera makes of that “same” view. Before digital it took quite a while (and some wasted footage) to get used to the idiosyncrasies, now it’s not only a faster process but editing and software make it easier to match results (though you argue not to rely on these too much and i agree).

  4. I’m glad you’re freshly pressed. Your work is amazing!

  5. Stunning …

  6. Thanks for your thoughts and experience in the photographic world. i will be taking a lot of your info on board. Once again thanks.

  7. Inspiring and practical

  8. Ambitions4 says:

    Well explained. Interesting images.

  9. Your use of shadows is masterful.

  10. Love the photos! Thanks for all the helpful tips!

  11. Reblogged this on PhotosbyMIK and commented:
    Its people like this that make me feel like a VERY bad photog. The work in this post alone. Wow. Read it!

  12. Gorgeous examples!

  13. You have offered valuable suggestions and your work professes the definition of visual consistency. I enjoyed finding out some of the images are digital while others are not, I love the surprise ending in this fact.

  14. packsandroid says:

    Blog is good
    Thank you

  15. Great work. Yes I think exposure affects outcome in different ways especially when consistency is in question.

  16. good advice on various techniques…very useful for real life

  17. Wonderful text. I’m not a photographer but this article made me want to pick up my camera again :)

  18. Of late, I have begun to focus on control over the results from my camera. And your post came along. Thanks for the advise and the visual treat.

  19. Reblogged this on digitaldissecting.

  20. Reblogged this on Finder's Keeper's and commented:
    wonderful imagery

  21. Amazing images

  22. Great b&w photos!

  23. Muito lindo, principalmente a sombra do homem e a bicicleta dá um sentido de profundidade tremendo.

  24. Love your work!

  25. Reblogged this on Apps Lotus's Blog.

  26. Great post!

  27. Reblogged this on Mindful Planet and commented:
    Beautiful photography.

  28. Reblogged this on ABC's of Entertainment.

  29. brilliant…loved the second one and the one with the shadow of the bicycle and rider…too good…and tip on “Learn the idiosyncrasies of your cameras” again a good one…thanks!

  30. In the second picture, the man with the bicycle, I love the reflection of him in the window, as crisp as he is, it makes it look real and surreal at the same time. All the photos really do share a consistency. Truly remarkable.

  31. crushtonbljs says:

    old picture

  32. Reblogged this on THE PIFOSTIC and commented:
    Awesome photography

  33. A really interesting post, thanks for sharing.

  34. good working this images

  35. Reblogged this on Rotating Relativity and commented:
    Not only is this relevant, its consistent with logic and complexity simultaneously. Thank you. Follow my page as well.

  36. Very poetic. The pictures are breathtaking.

  37. I reblogged your article on
    Please visit and give your thoughts.

  38. Reblogged this on Stylish Affairs and commented:
    This is a though one for me because I don’t see how on earth can contribute to such a content especially that I know absolutely nothing about photography except what any novice can know which summarizes in just knowing to press the shoot button and that’s it! I think it is the reason why I wanted to reblog this post so that all those who face the same photography challenges like me can receive help from their fellow bloggers. I am what I call a baby blogger and field is fashion but whenever I visit other fashion blogs, I can’t help it but wonder how do they make their pictures look so clean, filtered, consistent, etc?
    So can all (or at least some) of the photography genius bloggers help me and all those who are in the same situation like me?

  39. Reblogged this on brendan joyce and commented:
    Fantastic in so many ways, many thanks for the inside look.

  40. Reblogged this on mafgrupocinco and commented:
    Esta bueno

  41. stunning shots… love the shadow and light play!

  42. Congratulations on being Freshly Pressed, although from the number of responses you have, you didn’t need the boost.

    • Thanks. It’s not the first time, though I’m guessing the reblog was because of the content/ popularity rather than to boost it…I’m lucky to have a dedicated following here. Other articles have had up to 400 replies, 50+ is average.

  43. Such helpful information and stunning photographs. I find the second image, in particular, mesmerizing. I’m still trying to orient myself to what I’m seeing and wrap my mind around it.

  44. Reblogged this on Location de benne en Alsace (67) and commented:
    I reblog !


  45. Good blog ! Thanks Ming Thein !!


  46. I’m pretty certain that you do have a ‘Ming Filter’, but I can understand why you don’t want to sell it or even admit to it’s existence ;)

  47. Hi everyone, I would like to say and express my thoughts about how great to read everyone’s opinions and their visual aspects on this work( photos and the essay itself) and I strongly believe that any work that is not standard is unique.
    Unique for everyone, and not every professional would agree with me and many others, but that is what the talent and good works comes from unique, honest, opened minded, not scared to do stuff person. Maybe it’s not necessarily “politically correct” or not standard, but it’s really awesome!!! Great job, love the way it’s done! :)

  48. NeutraL-GreY says:

    I was able to instantly determine that the last shot was film before clicking on it but for the life of me I can’t figure out why It jumped out at me.

    • I think it’s the highlight rolloff (still). Getting close with digital, but not quite there yet…

      • Maybe someone needs to invent a sensor that produces a negative image (not that this is likely), just like film. Seems to me that the highlight rolloff is at least partly due to this (ie it’s typically the shadows that get clipped on film images).

        • Now THAT’s an interesting thought…

        • Tom Liles says:

          That is a good one, Bob. I had the same idea for different reasons back when I was trying to camera-scan my color negatives with a DSLR. Would’ve been great to have a “uRAW,” RGB inverted dataset to reverse the negs in-camera (by design) as it were.

          Had never thought about highlight roll off etc though—-but aren’t we set up for that already with digital recording?
          (Most numbers in the top bits)

  49. Derrick Pang says:

    Wonderful write up and amazing photos. The 7th is my favorite. Learn a lot just by looking at the photos.

  50. Tom Liles says:

    Ok, this has been eating away at me for days: the first image is slightly off kilter, that seems quite unexpected to me. Were you tempted to level the horizon in post? It might even be that you un-leveled the horizon in post… I have no interesting thing to say or conclusion to make. I’m not sure I even want to know the answer…
    Just to note this and that it has grown from a minor “oh, hm” thing into an all-consuming meditation on intentions and their relevance.

  51. Ming, this is a nice set of images and good advice. There is enough consistency in the images to prove your point.I suppose my way of avoiding the problem is a practical one, just use only my D800.

    • Thanks – and yes, that would also work :)

      But there are of course situations in which you might not be able to use that one camera, or might have to use something else, or might not have it with you etc…

  52. Are you having sales soon? Really keen on ep4+ep5 but couldnt afford it right now =(

    • Sorry, we don’t do sales because we believe in the value of the knowledge in the videos (and so do our customers). There is an introductory bundle for Ep 4 & 5 which will end when we launch a new video, as usual.

  53. Thanks for the insightful article, and plethora of color and feelings in these B&W images, as always Mr. Thien :-)

    As far as concerns that little mr. ti , as so unwisely expresses its own self: “I never said ..blablabla”,
    mr. ti did say. a lot of words. a crap lot more than that. using letters. only.

    Let that little piece of *ti* put its $ where its little piece of mouth is, in binary format or exhibition format, proving its unlimited and utterly exploding knowledge of EETR, BS.TTR, CR.TTR, xxTR. etc…

    cheers ;-)

  54. Martin Fritter says:

    The images above are quite nice, well better than that.

    I’ve found the fish fight very informative and oddly helpful. It seems to me that Ming and Steve McCurry are almost polar opposites – at least aesthetically. Sort of Apollonian/Dionysian – or something like that. Or digital/analogue. Ming’s has embraced the most demanding aspects of the digital medium, which, like it or not, does suggest specific aesthetic problems and choices. His best work does not look like analogue photography – it also doesn’t look much like the general run of digital – it’s really on the leading edge, which makes it very exciting.

    Now for me the jury is still out on the digital revolution. But I’m old enough that I have never fully embraced digital audio. But if you’re going to do digital, do it at the highest technical and aesthetic levels, and that is what I think this site is about. Best of luck and keep at it!

    Oh, on the general topic: interesting the stylistic unity between the digital _and_ analogue work presented here. That’s quite an achievement.

    • Thank you. There are however still things that we digital photographers can learn from film – B&W highlight tonality, for one. I’ve been shooting film for that reason. Once I figure out how to replicate the tonal response – and I’m pretty close – that will probably be the end of film for me. At these sizes, digital/analog images look pretty consistent – but at larger sizes and in print, there’s still a small, but visible difference.

      • Philipp says:

        Hi Ming

        Would not be a leica monochrome exciting / something interesting for you then? Focused on B&W, great tonality, higher light sensivity…

        And what do you think about IR photography? I find it a very interesting field of photography….any intentions of you going to or trying it?


        • Nope, because I can take the colour out…but I can’t put it back in.

          IR doesn’t appeal because it’s very difficult to visualise; too much luck involved.

  55. mosswings says:

    That’s basically my point. And actually, I’d like to restate my point. On further reflection, ETTR has actually already been implemented by the manufacturers….in their evaluative metering systems..sort of. ETTR, despite its name, has to me always meant “expose to maximize data quality”, but what data quality means is somewhat subjective. In many scenes without specular highlights, the standard 2.5-3.0 stops between neutral grey and well overflow are not fully utilized…there’s room for increasing exposure if so desired. But a savvy photographer will evaluate the scene and adjust the exposure, often upwards. In scenes with highlights, 3.0 stops may not be enough, so the exposure may go down – if those highlights are important. Evaluative metering systems try to do just this – they evaluate the whole scene and make data-preserving adjustments. However, the criterion they often use is “protect highlights for JPG users”.

  56. Quoted:
    “Since I can’t reply above anymore: Ask yourself again, why this amazing idea of ETTR is not the standard way digital cameras expose pictures. Are the people at Nikon, Canon, etc. so daft not realize that they are throwing away data unnecessarily that could be captured by ETTR? Is there a big conspiracy going on? ;-)
    Anyway, sleep well!”

    I suspect one reason that manufacturers have not yet included it is because they are still trying to prevent DR issues for casual jpeg shooters. The new Olymus E-M10 is reputed to underexpose so casual shooters will not find blown highlights in their images. I would also suspect that they are currently assuming that anybody who wants to implement ETTR, and has an EVF, knows how to read and adjust their histogram accordingly. Yes, it would be a nice feature, and I suspect that it may make its way into a soon to be released camera, but as histograms have become more common, the need for an automatic feature is greatly reduced. In short, it’s a good question, but not one that I would want to bet the farm on.


    • But then if the JPEGs were too dark, all the casual shooters would complain of underexposure…you cannot have one size fits all settings for every situation. Every picture is different!

      • And that is the reason that I never took to Nikon’s matrix metering. We were each trying to outhink each other, and I never knew what the camera was, or was not, going to do. With spot or center-weighted, I have some idea as there are not primarily based on a “matching” of the scene to a database of exposure settings. Sometimes there is no substitute for experience.


  57. Tom Liles says:

    I understand the love for #2, and I love two too, but #3 is a special image for me—it’s from the M8 or M9, if I recall correctly, from London (my home country, but definitely not my home town: I’m a Northerner!) and MT, your articles on those cameras were my route to discovering this site, your writing and most of all your photography (and eventually, in time, your quick-draw GR holster).

    Seeing it again makes me think two things — two of which might be a little heretical to say on this essay of all essays — they are:

    1) Your style has evolved, noticeably, since then. I can’t verbalize it (to pick up on a recent article) but I feel it

    2) This is the sort of easily quantifiable “now THAT’s a photo!” type photo that I would have reacted to very strongly two or more years ago before I shot cameras (not to say I don’t react to it now, obviously not the case); whereas the harder to understand stuff is more rewarding to look at for me now, but that’s a little lonelier experience… I could imagine showing #3 to my friends, family, people at work and everyone needing no convincing that this is a high-caliber professional image and that it’d be good on their wall. I’d contrast that with, say, a recent fav. of mine from you, this one—where I’m not sure what the same set of people would do if I enthusiastically showed it to them. Probably give a non-comittal “yeah that’s amazing” but not really feel it as I’m feeling it. I don’t know if this is artistically good or bad, but commercially it’s surely not good—hence I understand your stress with having to balance what you want to do with what people react to. There is nothing else but to be a schizo.

    Abstraction isn’t new to you, it goes without saying, but looking back at #3 there, I feel like we can appreciate another level of tonality and perhaps subtlety in the newer work. Equipment plays a part, but that cuts both ways—if you’d had a D800E and Otus at that scene, that time and place, you might not have chosen to make this exposure, or this exposure in this way, at all…
    (This is why I never delete older photos in my scrapbook [Flickr], the past is as dynamic as the present—old photos go from good to bad to good again, to bad, to good, etc., etc., all depending on me, in the now)

    Good essay. Again.

    • Florian says:

      Hi Ming

      as Tom (Liles) mentioned you probably would not have shot the 3# image with your D800e/Otus combo….the 3# was one of your images taken with your leica m8 as you still had one…..

      Is there any cheaper camera (lens combination) which come as close as possible (closest!) to this beautiful color rendering and is highly recommendable for street photography (more discrete/ inkognito capturing/stealthy)? D800e+ 55 1.4 otus too big isnt it?

      Sony A7+ 55 1.8 Zeiss (due to Fullframe, zeiss rendering known for excellent optics/ contrasts)?
      Olympus EM1 + 42 1.2 or 25 1.4 PanaLeica (due to touchscreen af)?
      Fuji XT-1 or X-Pro with 24,35 1.4 or 56 1.2 (different sensor, color rendering, rangefinder style xpro1, manual control similar to leica)?

      Or is not any camera with similiar rendering avaiable/existing due to the fact that the m8 and m9 leica used a ccd sensor and not like the new m240 and all other named cameras a cmos sensor?
      So is the beautiful color rendering only the fact of the existing ccd sensor used in the M8? Cannot believe…..



      • Not true. I’d definitely have shot it with the D800/Otus; I used it for street photography in Taipei. Stealth is down to the photographer and his actions, not just equipment size.

        The whole point of this post is that the equipment is irrelevant, so your question doesn’t make sense. The answer would be any or all. All of the images here have the same rendering but are drastically different. What IS consistent is compositional style, quality of light and post processing.

    • 1) Totally agree – but are you saying that just taking into consideration the images in this essay, or in general taking into count every other image I’ve published?

      2) Agreed again. The stuff that is popular with the mainstream seems to be visually pleasing on the face of things, plenty of immediate punch – but somewhat lacking in subtlety and thought. I suppose it depends on your intended audience. That recent fave is a set that was shot for me as an experiment, so it’s probably to the far left end of pushing for a bit more.

      Fully agreed though: the way I saw then is not the way I see now. But I can make it look similar stylistically…

      • Tom Liles says:

        1) sure, it’s just a general feeling almost certainly predicated on recent output in total. And when we say recent, perhaps arbitrarily take that as going back at least six months. There are elements, to my eye, of how you compose and create tonality right now in the older stuff—I guess stylistic evolution is typically a gradual fade-in/fade-out, so we see some aspects bleed-in and become more prominent, others fade away. But both old and new photographs are easily identifiable as your style, I think. Better put, as “you.”

        2) interesting that the leaves and carpark was a test shot (from the battered 645D loaner!); there’s no grandiose reason I rate it, in fact it’s probably that it seemed like the kind of quiet and casual “this one’s for me” type photograph is the reason I liked it. I wouldn’t expect to see it in your portfolio; but the sketch presents a strong idea about composition which represents more, to me, where you’re at — what’s cool to you right now — than the more orthodox stuff.

        Art is definitely about the new, so I’d always rather be engaged with (for or against) the contemporary you than the old one.

        • 1) Another thing to consider is that whilst one might have accidentally made something you like, consistency and control is quite another thing.
          2) Won’t be in the portfolio, but I personally sense the germ of an idea there…

    • Hang on – a *northerner*, Tom?! And to think, all this time I’ve been reading your posts in my native southern accent! Ö

      • Tom Liles says:

        Ee by gum!

        From the grim North I’m afraid Todd—a small village no-one’s heard of that’s famous for nothing. Escaped into the big city of Manchester for university; escaped into the bigger city of London for job interviews thereafter; ran away to the even bigger still city of Tokyo to avoid working for Bayer or Glaxo Smith Kline or other big serious Chemical co…
        Held on in Tokyo for about a decade; came back to UK and had two great years in nuclear in the SE and lived in good old Rye, round the Weald from you :)
        And then came all the way back to Tokyo again. And now I own a piece of it. I can legitimately, and legally, say Tokyo is my home—I’m a Tokyo… oh, what would be the right appendation? A Tokyoite??

        I’ve been around the World, all over—but know this! You can take the man out of the North, but you can’t take the North out of the man!

  58. Reblogged this on Human Relationships.

    • Roberto_Burle_Marx_Fan says:

      Great work and essay Ming! Agree with others, the second photo is definetely the best for me as well!

      But a provocative question is it really good and necessary to have a consistent, unique style as a photographer (y artist) (like corporate identity in companies?)? Are those style elements or aspects not copied together or influenced from other artists…….? So it isnt really an own, new style…..

      I think it is very hard to find a complete new style…….

      2 cents!

      Have a unhurried night!

      • Not necessary because we all evolve anyway. However it is useful if you’re trying to make your work identifiable and thus commercially valuable…no other way to make a living as a photographer, unfortunately.

        • Roberto_Burle_Marx_Fan says:

          Yeah or you make it due to provocative actions, photos or news like Mike Dowson or Terry Richardson……the charlie sheens of photography…..the poor girls….

          • I don’t think that’s a very good idea. I’d much rather chase the impossible unique style…

            • Roberto_Burle_Marx_Fan says:

              Totally agree but they are nevertheless very successful despite their “strange methods” (hard to find the right/appropriate words for them/what they are doing confirmable with several trustwpórthy evidences/sources;)!)

              But Richardson seems get banned or kicked out working for the VOGUE US thanks to Anna Wintour…the right decision he is still dominating the whole us market due to his business network, relationships etc….

  59. I strongly disagree with your comment regarding software filters. As usual, looking to the analogue word clarifies the debate. I am not a film shooter. Yet, I know that film shooters have favorite films because of their look. So when you choose film, you are effectively choosing a filter–just earlier in the process. I suspect that is also true of choice of darkroom chemicals. So is a digital filter any different?

    My workflow is to start in ACR, then Photoshop (where I pretty much limit myself to a hue/saturation, curves, and selective filter layers, as well as an overlay layer for dodging and burning, but for black and white conversions or Leica MM files, I use Silver Efex. I find it useful and efficient to start with one of three filters that usually work for me–Base, Push Process(1), or Fine Art(1). I then make the necessary adjustments. I also have two or three film types that I like (What I particularly like about Silver Efex is that it shows you the curve associated with the different film types). In each case, I can dial down or up the filter through the individual controls, but I find this to be very efficient. When I work with color files, I often use Nik’s Color Efex software to either finish the Photoshop work or as a base starting point. I particularly like their contrast filters, particularly their dark contrast filter, which often works better than adding a vignette. Occasionally I find the Reflector Effect (Silver) is a good way to brighten the bottom of a photograph. It often works better than a gradient. I am a huge fan of Saul Leiter. I have found that one of the color filters under the Vintage Filter tab comes close to emulating his color scheme (with adjustment as needed). I got the idea for that experiment when I read that he used to buy expired color film because he liked the color cast.

    Software filters are just more tools in the photographer’s arsenal. So there you have it.

    • Tom Liles says:

      Hi Jack, I gave myself a couple of days to think about this one—hope you catch it so late after the fact.

      I’m a nascent film shooter (I use both 120 and 135, even Fujifilm instant film in a 120 polaroid back) and it does seem true that films have a look (though it is hard to separate completely the influence of the scanner in this if we’re digitizing, and more or less everyone is digitizing); though I’d qualify that a little and mention that C-41 emulsions tend to be quite uncontrasty and leave a lot of room for going this way or that in post—they are really quite close to RAW files in many ways.
      Typically what it’s hard to really change, radically, is coloration. I recently got spiked on Kodak Ektar—completely unaware of how bad it is for taking photos of people. I took a whole roll of 120 in a 6×6 camera (12 shots) of people under carefully prepared lighting, couldn’t wait to get the negatives back, and when I did and scanned them I was greeted by some very red-magenta looking friends indeed. I used these photographs as an exercise in color correction and a test of my skills with the HSL palette; the people I’d taken were co-workers, graphic designers, who all went to art-school and retouch and color things for a living. I presented my finished retouched photos to them without saying anything about having been Ektar’d, and only two of twelve of them noticed that something wasn’t right (when shifting hues in HSL, I think the whole color wheel takes a spin and other hues tend to drift away from where they should be; shadow areas go a bit strange too..). While I’d corrected the reds in their skin to a good degree of verisimilitude, one of them was wearing a red cap at the shoot and he noticed that in his photo his cap wasn’t red (the cap’s red was “perfect” [pleasingly red] on the pre-edit scan, but of course he wasn’t!). The other guy was a little more astute and picked up on slight color casts in the shadows and the color of his lips and gums being a bit off. Luckily I also shot a roll of monochrome (I did 1 shot color, 1 shot mono) and so I had some unimpeachable keepers from the shoot.
      So Ektar, for example, has a look, when it comes to color; but tonally speaking it seems a little more complex than that to me: if I do the classic thing of exposing for shadows, letting the film’s highlight shoulder roll the highlights off and scan the developed negative carefully after getting it back from development (you see how many confounding steps there are between exposure and final output with film), I can set myself up with a 16bit TIFF that looks like a bell curve [on the histogram] when opening up in Lr/Ps… this is pretty much with most C41 films. I can take that high or low key if I want, within reason, and it’s really not that much different than working with data from a digital camera from years ago. Anything more drastic and we should be envisaging that end goal at the time of shooting (though with C41, it’s not such a great idea to expose for highlights, so for a deeply chiaroscuro image, for instance—in those cases you have to know that that’s what you’re going out to get, and load the camera with E6 process film, aka slide film/color reversal. Or it could be that C41 underexposed for the heavy grain and a disheveled feel is the goal of the exercise–I know a couple of street photographers who routinely shoot print film, C41, like that on purpose)… At any rate, correctly exposed and carefully scanned C41 is a lot more flexible than I’d thought or expected it’d be—though there is the color response. That’s all I find is really so different between film stocks.
      In B&W there are also famous looks—the crunch of TriX (getting quite hard to buy now), the velvety shadow texture in Acros, the low-fi elegance in Ilford films, and so on and so on: the bona-fide film guys you know have surely told you all about it. But again, to me at any rate, that is not unique to film and to the crunch of Tri-x I’d present roughly taken iPhone shots or GRDII or III monochrome files; for the velvet in Acros there is surely no 1:1 digital analog, but there are certainly very tonally smooth and pleasing digital cameras which do a digital version (in the digital space); and so on and so on..
      I have a few brands of digital camera and getting photographs from them to look the same is not as simple as it should be. I’m too inexperienced and wet behind the ears to know the whys and wherefores; all I know is, I can set up a product shot, background lighting, the works, put the D3 on the tripod, do a “pre” white balance from a grey card, then take the photo; and then mount the A7 on the tripod, do a “pre” white balance again from the same card same lights, take the shot at the same camera settings (and focal length, etc., of course) and the photos are different, markedly so, more than slight differences seen in the histogram should account for, when opened up on the same computer and same screen. Not as marked as the difference between Portra and Ektar, say, but you get my point. Nikon and Sony both process signal differently, that’s what it must be, but both sensors there were fabbed by Sony, so it really is quite amazing how these differences are so stark. The point at hand, though, is: is what happened here so different than the films talked about above? Film is perhaps only different by degree, and not type.

      Without a doubt it is easier to process the digital images, and someone with as much skill as Ming or the people who work in my office can equalize these things across cameras or media very, very well (and conversely they are the only people who really notice the tiny differences, so the different datasets only look equalized to us laymen); but you see what I mean about different only in degree, and not type—signal processing can be just as variable as film characteristic curves.

      All this perhaps riffs with your point, since we could go a step further, a step back before post, and say the choice of digital camera, like the choice of film, is a filter. But this is to begin to make the point so all-encompassing as to be of not much practical use.

      The real thought I came up with, considering your comment Jack, was that the reason digital filters grate a little is perhaps because they are after the fact. I mentioned above that if chiaroscuro is what we’re looking for when shooting a film camera, probably the quickest and best way to get it is shoot E6 and expose for highlights with a vengeance. This decision is a kind of filter, we knew we wanted the film’s look, but we had to make that call before the exposure—and once we’d made the decision and loaded the camera with expensive film, that is it.
      Why digital filters grate at some, perhaps, is that this thought process is leap-frogged, and the photographer doesn’t have to really commit, full on, upfront. I can see how that offends the aesthetic, or plain ethic, sense of photographers; I guess it looks too much like a free lunch.

      I’m not tonally careful enough to be bothered to use filters, or not. But I don’t have a problem with their existence at all—as you say, it’s just another tool. But I can well see the other side of the story, and have a good deal of time for the idea that it’s dangerously close to out-sourcing some of the big ticket creative aspects of post-processing — though as your workflow shows, it isn’t as simple as that, as simple as pressing a button and then going “next!” — better put: it seems like it’s a holding up of the hands and relinquishing of control (plus the idea that anyone can get the result, and everything that that implies). It’s certainly true that in film days many photographers just took the photos, sent negatives to a trusted print-maker who did a ton of dodge and burn, etc., in the enlarger and made the iconic images we all know today and attribute to the photog (with no mention for the poor old printmaker!). Photography’s closest thing to a Godhead, HCB, never printed anything himself—he did exactly as above. His printmaker was his filter (another filter, after his choice of film and lens :) ). So perhaps digital filters really bring it full circle, more in line perhaps with how it used to be. But I just have this nagging feeling that it’s better to go cold-turkey and do as much as is possible ourselves. If not just for the know-how. It may be that I’ve been out East too long and the confucianism is seeping into my bones, but I feel like if we didn’t carry the process out fully, if we were permitted to cut some corners (which is how digital filters are viewed, like it or not) then the end result is irredeemably tarnished or not at its full value.

      Sorry for the ramble. Hope you catch this, Jack.

  60. Great read. Thank you.

  61. Brett Patching says:

    Great article Ming.

  62. Very interesting tips and good pictures. Thanks for this post

  63. This is like a refresher/concise summary of the big topics you dealt with in the early days of the blog and the great thing is it’s making sense to me! Style/consistency is hard to achieve as an amateur and I still have some way to go for being totally consistent but being conscience of what you want to do before taking a shot is a big thing in achieving this.

    • The big topics are still there in the archives. Frankly, as a pro it’s even harder because most of the time your clients have very definite ideas about what they want; it’s almost always in the style of somebody else or not what you think would work best visually.

      • Wow, that was um interesting above….anyway i still look back at the archives – hopefully it’s still there years down the line for others to discover and not get lost in the Internet after you sign off…..

        • We shall see. I’m thinking of doing a massive ‘book of everything’ in print, with large images as an endgame…

          • I think we’ve discussed the book before ;) – but such amounts of information should not be allowed to get lost. The biggest problem I see with the Internet, legacy data or useful stuff just disappears if you don’t keep paying your subs which is a bit worrying in the record of our era

            • Yes we have – and we also talked about how I’d have to rob not one, but several, banks to make it happen.

              That said, perhaps I could just do a HD ebook version – what I don’t like about that is how easy the images become to reproduce afterwards.

          • I would buy that one immediately!!!

            • You might not after finding out how much it’d cost…I was shocked by how expensive gravure costs. And knowing that it’d probably be a 4-5 volume thing the size of Modernist Cuisine…

  64. Mmm, meaty… My favourite kind of article on this site. Any new videos on the horizon, Ming?

  65. siddharthajoshi says:

    Beautiful play of light and shadows…

  66. Nice pics. I don’t understand, though, why you are recommending to expose to the right. All the photographers I had the chance to speak to, including a certain Steve McCurry, always advised me to expose to the left, since shadows can more easily be recovered in post and colors become more vivid. What am I missing here?

    • Thanks for the interesting post. Getting the style consistent is something I still struggle with. I got Lightroom, but prefer to use Photoshop instead and adjust images individually which gives variation, but perhaps the variation is the result of an inconsistent shooting style instead. Will try and keep that in mind next time. Thanks!

      • It’s not that: simply, each image is different, therefore the optimum processing for each should also be different. The easiest way to achieve this is through a robust and efficient set of tools; PS is the best of what we’ve got at the moment. It’s probably overkill for most things, but you can always ignore the bits you don’t need.

    • More information is captured by the sensor when you ETTR (Expose To The Right). You do need to be careful not to blow out any highlights though. It is an often discussed technique, and a quick search will bring up lots of opinions about it. Here is an article about ETTR with the E-M5 and Lightroom: .


    • Shadows can be recovered in post but with a significant noise and dynamic range penalty. Sorry, but McCurry should stick to film – that rule holds true, but not for digital. Digital has to be shot very differently to achieve optimum results.

      • To be fair on McCurry – I’ve seen him give that advice too, but only in relation to film. I have never seen him give any advice on digital, and think that it’s Mr. T attributing this view to digital, not McCurry himself.

        • Something lost in communication then. That statement about exposing to the left for digital will yield utter crap.

          • Well, McCurry made that comment in relation to digital. And to tell him to go back to film is a little condescending, don’t you think? No offense, but it’s like me telling Prince how to make music. But at least I understand now, why your colors always look somewhat washed out:


            • Tom Liles says:

              Mr. T,

              With due respect to the howgreenisyourgarden blog and Steve McCurry—I don’t think either of you are fully in control of the concepts at hand (and that’s not saying that I am). The best explanation is Emil Martinec’s, and in particular I’d invite all of us to read this passage, S/N and Exposure Decisions.

              This paper was linked to on here in one of our discussions a week or two back.

              I don’t think it can be refuted.

              • @ Tom, thanks for that article. But McCurry would probably say that caring about noise is as bourgeois as caring about sharpness ;-) After all, he is mostly interested in getting color right (in addition to actually making a good photo).

                @ Ming: Your software doesn’t allow me to reply to you last jab. I asked you a polite question in the first place to which you replied in a condescending tone. Of course you can do whatever you want on your website, but you just come across as a little smart ass. But it’s a free world, do whatever you want…

                • Tom Liles says:

                  Come on, let’s have a fish fight then…

                  Mr. T,

                  You certainly didn’t read the paper I linked to else you wouldn’t have replied like that. Sock-puppet McCurry can hypothetically say whatever he wishes in your head, but real world McCurry uses a camera and cameras work on Quantum Electrodynamics and basic electrical engineering (which is just a subset, as most things are, of Quantum Electrodynamics)—all the aesthetic spiel and faux blue-collar-heroism in the World cannot circumvent the fact that electron wells saturate at a certain photon flux, that upstream read noise of an ISO gain amplifier is about five times higher than downstream read noise at low ISOs, for instance; that below 30db life isn’t great, and below 20db we’ve lost the ball game; that shot noise is in the photons themselves, not your sensor; or that linear ADCs work in a certain way… and so on and so on, and so on. These are physical facts of the tools we use. You seem to want to dismiss how yours work because it isn’t a simple a task — as you’ve figured out — and the easier thing to do is be a contrarian and when that doesn’t work, a pseudo-pragmatist–“forget ETTR nonsense and lets think about real world photos!” etc. Well, unfortunately, Mr T, bona-fide pragmatism is understanding your equipment, really understanding it, and trying to extract the best data collection possible, because in digital photos that’s all it is—numbers. Thereafter we can get our art on, hell yes. Ming’s already written above the line [the big bit with photos in it] that he exposes with an end goal in mind… not exactly the high priest of ETTR… which you seem to have missed so I honestly wonder where your ETTR-takedown comes from. You’ve made a fish fight out of nothing really, and your main points so far seem to be: “Steve McCurry said it, I agree, and that settles it.” That’s not the best of rhetorical platforms to go from. Anyway, it’s a storm in a teacup Mr. T, you expose how you want, MT will do it his way, McCurry his, and so on and so on—we’ll judge on finished images in the end; I’d think MT isn’t doing too shabbily, don’t you?

                  And condescending could be something you’re guilty of too: you know you are talking to this guy, yes?
                  (check the date on that posting)

                  • It actually was a sincere thanks from my side. But I enjoyed your little rant without content. And to be honest, I prefer to be a fan boy of McCurry rather than Ming Thein ;-)

                    • Tom Liles says:

                      No problem Mr. T, but sincere thanks usually aren’t bookended with a “but…” clause. It’s your birthday, your wife gives you a pressie, you say “Thanks, but…”

                      Not going to go down well is it.

                      Anyway, don’t take it too seriously Mr. T, we’re all photogs, this is just the internet—-I think everyone has been very live and let live and said “ok, you shoot it your way Mr. T” Can’t say fairer than that. You’ll have a job convincing many here, though, that that link is Gospel and ETTL and reclaim shadows later is a cunning plan for malleable data (from which we can make the pictures we want) or has a great part to play with awesome saturation results (question: what is “saturation,” Mr. T—I’m not trying to catch you out, but I bet you few photogs can explain this, with rigor and brevity; if I give you a blue channel curve in Ps/Lr etc., what would you do to it to saturate the channel?).
                      Anyway, just to give you an idea how it looks from over here: the only evidence you’ve presented so far for “washed out ETTR” is that howgreen… article and because Steve McCurry said so. I’m persuaded more by the brainiacs at the research labs on this one (on how digital data works); but most of all by what computes for me—and that’s pretty much as the braniacs describe it.

                      And stop with the low-blows, Mr. T. I’m just a fan; drop “boy.” I’m 35, I run a house, a wife, three kids, a car an intensive job and do it all in a second language in a foreign country; I can tie my own shoelaces and fill my own tax-returns out, etc. It’s a shame if you think you’re generating this heat because of some circle-the-wagons response from me or others… You’re bein silly if you think you weren’t pushing buttons on purpose above but forget all that; we’re just pointing out that we don’t think you — or should we say Steve McCurry — is in full control of the reasons why they do what they do. And a-OK, underexpose ahead Mr. T. It’s your world!

                      Please do give that Martinec paper a go someday though; I think I was where you are, and it blew everything I thought I knew and understood away. It’s not easy, but worth it; and a start—then you won’t need anyone’s opinion but your own.

                      I’m off to bed now!

                      — just joshing with you Mr. T; pitty the fool :)

                • WordPress limits the number of nested replies, not me.

                  I answered your question. I’m not the only one who’s said the film ways of working are incorrect for optimum quality out of digital. Far from it. You chose to argue and be condescending and believe that fame = truth. That’s not my problem, and if you don’t wish to see alternative points of view, then you’re the condescending smart ass.

                  If you care about accurate colour, slide film is NOT the place to start.

                  • I never said that you are only one believing in ETTR. I also never wrote that I am interested in “accurate” color (I’m looking for impactful color). And finally, I don’t care about fame, I just mentioned McCurry, because everybody knows him (as I said, I have spoken to other photographers, too).

                    The problem is that YOU are not willing to see an alternative viewpoint since YOU know (in your mind) what CORRECT exposure for digital is (implying that every other approach is wrong). For the last time, it would be the easiest thing for camera manufacturers to automate ETTR if it really was THE correct way to expose. But they don’t do it because it has negative implications, too, weak color rendition being one of them. And this is a fact you were simply not aware of before, if you are honest.

                    • I shouldn’t jump in, but what the heck … Suppose that you expose someone’s face to be brighter than an 18% grey reading. How does the JPEG engine (or RAW processor) know whether you intended for the face to be rendered high-key, or that you were using ETTR, so it should reduce the JPEG’s brightness to make the face be whatever brightness the meter measured? The camera and software doesn’t know, which is why it has to be a manual process.

                      Olympus and others also offer a pull-processed ISO setting that’s lower than base ISO, so one could argue that they are giving you one form of ETTR. Black Magic’s Pocket Cinema Camera has a mode that adjusts the aperture so that no pixels are clipped, but you still need to adjust the rendered brightness afterwards manually to reflect your artistic intent.

                    • Sorry, have to put my last reply in here, since that is bollocks. JPEGs are simply a rendering of the raw data and the in-camera JPEG engine could adjust the ETTR Raw data like you do in Lightroom or Photoshop, too, to get the overexposed Raw image back to a “correctly” exposed JPEG one. By default, that would result in better JPEGs if your theory were correct.

                    • It would, except the in camera JPEG engine is still limited by the same lack of knowledge about the subject and artistic intention of the photographer. So even if it is technically possible, the creative decision the camera makes will almost always be different.

                      And to be clear, we are not talking about absolute overexposure: that’s when the highlights are clipped. There is no real recovery taking place here, it’s interpolation of the data. If it’s brighter than you intend it to be – which it almost certainly will be – how does the camera know what your intention is (and how much to bring exposure back by)?

                      It’s not ‘bollocks’, as you so eloquently put it – it’s a creative limitation that the camera cannot decide on behalf of the photographer.

                    • Final, final reply: Don’t you think that the camera software would not be able to automatically expose so far to the right so that it stops just before the first pixel clips? Don’t you think the software would not be much much more precise than what you can read off the histogram on the crappy little screen of your camera when manually exposing to the right? I think you have to face it that your arguments simply don’t hold water.

                    • Yes, it could. But if that first pixel clipping is so far away from the rest of the image – a single point light bulb, for instance – it would leave the rest so dark that you’d see nothing. The artistic intention is compromised and you just have a dot. Sometimes very small parts of an image must be overexposed to look natural, and only the photographer can make that decision.

                      As to my arguments not holding water – well, if you choose to ignore every piece of evidence and logic and twist my words out of context, then you can poke holes in anything…

                    • No, it’s not technically possible to implement ETTR metering consistently for the reasons I listed earlier. It also doesn’t yield usable final JPEG results, which is what most of the population wants. Not everybody shoots raw or post processes, and I said I wanted accurate colour.

                      Again, you’re taking my comments out of context. This is the last I’m going to say on the subject:
                      1. ETTR is the correct approach for maximising raw digital data for post processing later. NOT for jpeg and final output straight out of camera.
                      2. Underexposing is correct for slide film and possibly JPEG if you have no intention of further processing.
                      3. ETTR metering is not easily implementable because the camera cannot tell if you’ve got faces or lightbulbs (see my other comment). And it’d never be done anyway because most camera users are JPEG shooters, for which ETTR just looks underexposed most of the time.

                      And let’s not talk about honesty: I’ve got far more information and transparency about my personal background available freely online than you do. However, it’s very easy to be an a** if you don’t even bother to share your real name.

              • Anything can be refuted if the emperor has new clothes :)

            • mosswings says:

              Careful…we’re getting dangerously close to trading slaps across the face with a wet fish here. Ming tends towards a more restrained and “clinical” rendering style that has nothing to do with where the exposure was set. McCurry loves saturated colors, as do a lot of viewers. The reality may be somewhere in between.

              It sounds like what that blogger is talking about is what happens when you shift up or down the tonal curve from the manufacturer’s calibration point. If one doesn’t make the right corrections for a change in exposure level the colors should change. “Unthinking ETTR”, just using the histogram off the camera’s JPG rendering, doesn’t reveal channel saturation effects, and in those situations the codec will try to reconstruct color from the channels it has. The result isn’t perfect Practitioners of ETTR have in the past employed UniWB to represent the actual signal strength in the RAW channels, but dealing with the Greenies is annoying.

              I think that the blogger’s “Expose for the Key Tones” admonition is more what contemporary ETTR has evolved into, because of the practical limitations of luminance histograms. Look up the advice of many other knowledgeable photogs and they will always say to use the RGBL histogram if your camera has it, and pay attention to each channel.

              None of this is “Expose to the Left”. That does make sense when you can’t separate the rendering curve from the recording encoding (i.e., film), OR if your recording characteristic is dropping codes in the highlights (for example – Nikon’s “visually lossless” encoding scheme, which can result in lost highlight detail in extreme cases).

              Now one of the things that I think is true is that underexposing by a small amount – 1/2 to 1 stop – deepens the apparent color. But that may be just the way our eyes work, and not any hocus pocus going on in the camera.

              Forgive me for saying it, but there’s a lot of hand-waving in the referenced article that detracts from the good information that it presents. The author should have just stopped at the color chart and noted that ETTR should be EFTKT. The technical “explanations” therein are about as hand-wavey as mine are here.

              • Lower exposures appear to be more saturated because of the way digital displays render output at higher luminance levels. You can still have more saturation at higher luminance levels if you increase it afterwards in post processing. That has nothing whatsoever to do with exposure.

                • mosswings says:

                  Yes, and this is different than the reasons you might want to underexpose slide film. They get back to the rendering curve characteristics, which are controllable in digital photography and not in film.
                  Tom, the links articles are classics and excellent, but they don’t specifically cover what greengarden is discussing, which is tonal shifts with departures from the rendering curve setpoints. I still contend that what is being seen is simply a consequence of the rendering curve used and secondarily channel saturation effects.
                  But greengarden’s post has me scratching my head…digital recording doesn’t work that way.

                  • Rendering curves are inseparable from saturation – or at least perceptual separation, because we are increasing/ decreasing the luminance of the individual colour channels in applying the curve. However, you can change luminance without saturation easily by using Lab mode and just working on the L channel curve…

                    • Replying to your 7:24 post: You are right, with ETTR you get lower noise and better gradients, but the tradeoff is weaker color rendition, which you still do not seem to acknowledge. Hence, what maximum IQ is, is subjective (as so often in photography). Do you want lots of detail in the shadows with low noise or do you want to have good looking colors instead. Hence, there is no right or wrong.

                      But enough now of our Cambridge Union debate. Let every man seek heaven in his own fashion.

                    • I don’t acknowledge it because I do not agree. I think it depends on your definition of ‘weaker’. If you are talking about saturation, then that can be fixed easily in post. Accuracy cannot – there’s less information the lower down the histogram you go; how can you measure something accurately if you ruler has fewer divisions? Since by your own admission you are not concerned with the latter, that leaves only the former. And it’s easily correctable with little to no penalty, if you so choose.

                      This is not a Cambridge Union debate. There, you at least know your opponent by name, and can sort out differences informally over a beer afterwards. This is a one sided shouting match into the dark.

                  • Tom Liles says:

                    Agreed mosswings—but taking a look at the howgreen… piece and Mr. T’s thoughts it seemed like we might be getting ahead a little bit; and as I’d too recently fell flat on my face on the basic premises of ETTR and what we are really up to (and what sensors and circuitry are really up to) I passed the links onto Mr. T. I’d venture that before getting any further into ETTR it’s necessary to go through the Martinec paper and have a handle on what ETTR is in the first place.
                    Charles Poynton’s site would be my pick for a link about the concepts involved in color. It’s really for video, but all the same concepts apply—since we both output to monitors a lot…
                    Back to Martinec, since this paper was mentioned the other week (it turns out I’d had in my bookmarks from a while back but had forgotten all about it) I’ve spent a good deal of time going through it… and a week later, honestly, my exposure approach is completely up in the air now… It seems like the best balances for exposure (wrt SNR) are to be found in the range native < best < 1600. As Ming mentioned in the article, having a plan for how we want to process the image after the fact plays a huge part in choosing that balance.
                    But just this simple fact, that if aggressive shadow reclaiming is on the cards, shooting a particular scene at 640 instead of 320 might yield the less noisy end product. That was a bombshell—the two main points: 1) it all depends, and 2) the mantra that all beginners (like me) are indoctrinated into, that “shoot at base ISO for best IQ” isn’t strictly correct and has to be questioned (by the photographer himself, at each scene).

                    On color… Your comments were helpful though I confess I couldn’t make that howgreen… article make sense on this (and hence assumed a trip back to the bedrock might be a good suggestion). Color is a massive topic, and I mentioned Charles Poynton’s excellent work on video and digital-video color before—a brief look through the work there, and elsewhere, illustrates how in-depth this topic can get. This is before we mention that color itself is a perceptual property of our eyes and brain and so getting strung up about rote-systemically reaching “perfect color” is a hiding to nothing—just treat each image as it comes, edit till it looks cool. Stop. Next image. I don’t think it needs to be much more difficult than that. Though what’s for sure is that it is more difficult, theoretically, than any photography websites linked to here try to propose.

                    And now, for bed!

                    Cheers mosswings :)

                    • mosswings says:

                      Mr. T: I would propose that the standard way digital cameras expose is to balance avoiding clipping highlights against proper subject exposure. Try using evaluative metering on a scene with a strong specular highlight component to it and the camera will often lower the exposure. Dynamic range enhancing camera features, like Nikon’s A-DL, lower exposure by as much as a stop in extreme dynamic range scenes. The reference points that camera manufacturers have been using of late have tended to provide greater highlight headroom (not much, a fraction of a stop). Olympus has been praised and damned in the press for exposing to preserve JPEG highlights but underexposing for RAW.

                      ETTR – or more appropriately EfTKT – is situationally dependent, and like all exposure techniques, is somewhat subjective. Proponents of ETTR will readily admit that if your intent is to preserve specular highlights, then ETTR can easily become ETTF, particularly if the R or B channel has a very strong signal. ETTR is not an exact science. It therefore is not something that would work well with a camera on full-auto.

                      We have to respect Mr. McCurry’s exposure choices: they meet his needs and personal tastes. But his choices are not your or my choices, and we must always determine for ourselves whether they suit our needs. Delving into the exact science of digital sensor color recording and rendering is definitely beyond my pay grade, and I fear greengarden’s as well. We need either an expert 3rd party to weigh in, or for one of us to inquire further of Mr. McCurry.

                    • @ mosswings: I agree with you, everybody has their artistic choices and can expose in whatever way they want. However, let’s remind ourselves what Ming actually wrote: “Correct exposure for digital is all the way to the right of the histogram”. He did not write that ETTR is his personal choice or that he does it in certain situations. He apodictically wrote that ETTR is CORRECT for digital. And by now, I guess we have established that this is simply not the case…

                    • And once again you manage to take my comments out of context.

                      It IS correct if you want a raw file (or any file) with as much information as possible to work with and post process to maximise image quality in your final output. I’ve said that countless times on here, too. Your final output exposure may not be the same as the captured one, but pushing the lower bits of DR in a digital capture always results in noise; pulling the higher bits does not because of the way sensors work. Every subsequent higher bit/ stop doubles in the number of tonal levels it has.

                      How about this: you take a jpeg, underexpose it by a stop, and compare it to a raw file ETTR’d and pulled back to the same exposure. Ignoring detail, which looks better? Which has smoother tonal transitions?

                    • Since I can’t reply above anymore: Ask yourself again, why this amazing idea of ETTR is not the standard way digital cameras expose pictures. Are the people at Nikon, Canon, etc. so daft not realize that they are throwing away data unnecessarily that could be captured by ETTR? Is there a big conspiracy going on? ;-)

                      Anyway, sleep well!

                    • There is a very simple answer to this: because the pre-exposure meters in DSLRs do not read data off the sensor and do not have the same dynamic range.

                      For mirrorless/ direct sensor readout, its because how much ‘clips’ – one pixel? Two? Ten percent of the image area? – is an artistic/ creative choice as much as a technical one. The camera manufacturers have to make an arbitrary decision and leave it there. It might work for most, or not, because it can easily be fooled by a situation in which the tiny bit that clips is the bit that matters – think a of spotlighted actors in a stage performance. The camera has no way of knowing what the content is or the exposure intention. By luminance map, it could just as easily be street lights at night, and whilst you’d want to keep detail in the faces of the performers, you probably don’t need it in the underside of the light fixtures.

            • He is incorrect. I don’t care who he is, it’s factually incorrect. To believe a statement is right just because the person is famous without checking facts is positively dangerous.

              My colours look washed out compared to the over saturated stuff that populates most of the internet because it is a conscious choice and there is a separate HSL adjustment step in my workflow for it, not because I expose to the right. I could go the opposite way and saturate if I so chose. It could also be because of your monitor calibration. Sorry, but that’s another factually incorrect statement.

              • Have you actually read the article I linked to that explains why you never get the colors back in post? Apparently not… Also, McCurry advised to underexpose by 1/3, i.e. not a huge amount. And from the snail photo in the article, I can see why he is doing that. Also, McCurry is one of the best color photographers in the world and surely knows how to get the best color out of his gear instead of caring about superfluous detail in the shadows. Sometimes it’s helpful to learn from the great…

                • Read what I just read: I adjusted the colours afterwards. Consciously. It’s NOT the same as the reasons in the article you linked. Postprocessing changes everything; maybe it applies if he’s shooting JPEG only – but even then, it makes no sense because you are capturing less information – colour and luminance, which is equivalent to spatial detail. Underexposing by 1/3-1/2 stop is only applicable for slide film.

                  In any case, do whatever you want…

                  • I understand that you additionally desaturate your colors in post to make them look even paler. That’s your prerogative and you can also do whatever you want. But like the article says, if ETTR is such a great idea, why do cameras not do it automatically or at least have an option to do it automatically?

                    You are a good writer and some of your pics are also interesting (a lot of them are not). The only thing that bothers me is the high horse you seem to be riding on…

                    • You’re contradicting yourself. On one hand, you’re attributing a lot of weight to an article that doesn’t make a lot of sense, but ignoring empirical evidence (i.e. images) presented by me in another article. You accuse me of making boring images – fine, I’ve always said preferences are subjective – and then being arrogant, but you do even worse by refusing to consider another point of view except those you choose. Sorry, but I provide the resources of this site for free; it’s an enormous amount of work, and frankly, it’s my prerogative to say whatever I want or engage who I want (or don’t want to). I set up and run the site as an educational resource. If I am not confident of my facts and cannot back them up through experience and experiment, it’d be pretty darned useless. However, it seems it’s still useless anyway if you’re ignored and accused of being arrogant and wrong!

  67. mosswings says:

    What an absolutely glorious 2nd shot. Passersby reflected in the polished ceiling over their heads, with the camera upside down? Escher would have doffed his cap, Ming. Chapeau!

    • Thanks!

    • Hi Ming,

      I have to mention this too…..I needed several minutes to get it right (constellation/image formation) and how you probably shot it!
      An image where you have to stand still for a moment and think about…. Definetely a masterpiece shot!!! Congrats! Should be easily worth as an ultraprint if you captured it with the D800e and not the EM1….after you have the right lens/stuff for getting the most out of the d800e (zeiss 135, otus 55) as you mentioned, the em1 gets dusty in the drawer/case/vitrine or is just doing the video stuff , no stills anymore…..poor em1;)

      By the way what do you think about the Sony A99 (FF-Amount System) and the SLT concept?? Have you ever been played with it? The successor of the Sony A99 – the II could be a gamechanger even for the new d800s same 36mpx sensor as the Sony A7r but with IBIS, built in wifi etc. and if they get it done that the A99II will only perform slightly less in low light (due to the translucent mirror design) it could be very interesting….the ibis can probabyl reduce the difference too… and for the sony a mount there are a lot of zeiss lens on the market, especially the Sony Zeiss 135 1.8…maybe even optically better than the Zeiss 135 ZF2 and with AF and 1.8……


      • Equipment is only interesting if it lets you do something you couldn’t do before. The D800E/ Zeiss lenses let you print larger. The E-M1 is small and flexible. The A99…is what, exactly? It doesn’t address any problems that weren’t already solved, has so-so ergonomics and overall performance still lags considerably behind the D800E. Meh.

        • Just thought the only one with IBIS and FF still….(next maybe the Sony A99II and Sony A9 FE Mount)….

          Would it not be amazing and useful having a D800e/Sony A7r (FF camera) with IBIS…..??

          So you would use the OMD system solely if you dont do any ultraprints despite less IQ? Cannot believe….no Nikon….

          • We have VR lenses.

            No, I wouldn’t use the OMD system solely. There are no tilt shifts, and those are mission critical for product/ studio/ macro/ architecture, which is the mainstay of my commercial work.

  68. An excellent example of mastering your craft. I just wish your videos were also tailored to LR, as I transitioned to that instead of PS when it was introduced (and having previously used other PP software rather than PS). I know that you have chosen PS as your primary tool, but I cannot help but wonder how many folks who are using a workflow based around LR (for whatever reasons) could benefit from some of your teaching and guidance.


    P.S. I keep looking at that second image in your post. I am still not sure what to make of it. :)

    • Thanks. Sorry, LR is just too limited – there are things that PS does that LR simply cannot, and frankly, at $10 a month – its a no brainer. Complexity/ slowness of PS is not an excuse: it’s much faster to do things right in PS than LR, assuming you can do them at all.

      • I guess my frustrations with earlier versions of PSE turned me off on wanting to bring PS into my workflow. I will give it some consideration, as I am a big believer of slowing down so I can get there as fast as possible. And while I do not disagree with your assessment about LR, I suspect that many users could still benefit from additional lessons. We are blessed with great PP software, but for some of us, it can take a while to really master many of these programs. I was hoping to build on my base of LR knowledge rather than go back to square one with PS, but there is usually no substitute for a lack of proper/desired tools.



        • Well, PSE is probably the only thing that’s worse than both LR and PS! It has the limitations of the former and the complexity of the latter.

          Sorry, but how do you expect me to teach something that I don’t use myself? Isn’t that hypocritical, not to mention pointless since I don’t have anywhere near the same level of expertise?

          • Given my frustrating experiences, I am glad to hear that PSE is not a good example of a stripped down version of PS! And, I certainly would not expect you to offer any direct lessons for PSE, as it is somewhat of an ill-suited program for good PP workflow, so I apologize if it sounded like that was a request. I only referenced PSE as the reason why I did not take to using PS, not that you should tailor your teaching to it. But, I thought you had once mentioned that you were going to release some instructional video(s) specific to LR. I may have been mistaken, but in either event, PS is the preferred tool, and the basis for your videos.

            Sorry for any confusion,


            • Not at all. LR has been asked for many times, I’ve spent far too much time trying to make it work, and in the end I always have to go back to PS – 90% of the time what I want to do can be achieved easily in PS in a few seconds, but LR requires too much messing around for an inferior result. I find it very difficult to justify releasing any LR training since it isn’t what I’d use myself; it smacks of a lack of integrity. I wouldn’t want to mislead people into thinking I can achieve the desired results with LR, when it’s clearly not possible. That would surely be dishonest and rather pointless, no?

              • I understand. I appreciate honesty, even if it is not the news I wanted to hear. And even if a switch to PS is not immediately in the works for me, these posts/discussions continue to be a very valuable resource.


                • I’m some where in between both PS and LR although I would say I’m a little more familiar with LR. If it helps, understand how PS treats things and how things like layers work in general. Step through common functions you would like to use and go through them, gradually you’ll get a understanding of how things work in PS. Also Google, YouTube, a few good books on the subject doesn’t at all hurt :D

                  • In the past I have used LR and PS to apply Ming’s techniques. There was an occasional color shift changing between LR & PS. This may have been a configuration issue on my side. The videos are all in Bridge/ACR/PS. However it is fairly easy to map to LR yourself if you are computer savvy/don’t mind translating. However after watching EP4-EP5 of making outstanding images I switched to Bridge/ACR/PS. I really like the combination. I added Photo Mechanic into the mix a I shoot a lot of photos and Photo Mechanic is helpful in selecting images that are keepers. I then go to Bridge/ACR/PS for the keepers. Since Bridge and ACR are included in PS there is no cost associated in switching away from lightroom..

                    • It was the color space. EDIT -> Preferences . External editor tab. Choose: 16 bit pro photo RGB.. This is the recommended setting by Adobe.

                    • Thought ProPhoto is larger than ARGB, there are no monitors that can display it – so I’d actually recommend using ARGB as it’s easier to visually see what’s going on…no second guessing outside the gamut.

                    • Are you in the same working colour space in LR and PS? That might account for the colour shift.

                    • I too switched to ACR/PS fully after EP4/5. Too bad you’re not earning commission, Ming ;)

                    • Sigh. I did if you’d bought it in full via B&H or Amazon, but now that they’re subscription only…nada. :P

                    • I’m still a LR/PS person, and sometimes I don’t even go into PS if LR is enough. I know: heresy. I use PS if I need to straighten stuff out with distort instead of just tilting the frame, and if there’s any dodging and burning to be done. LR’s gradients are the best thing ever, though, and that was one of the best tips from Ming’s various workflow videos.

                      I’m also on the verge of printing myself, and LR’s print management is very handy and much more capable than PS’s. Of course, the day I decided this, my local printer decides to offer Hahnemuhle’s baryta paper for no extra charge …

                    • I agree with you on LR’s print management. We never print from PS…

                  • Thanks. I appreciate the encouragement. It’s just that right now I am a bit overwhelmed, and taking on something new that requires a lot of attention is not as easy as I would like. But, I am sure that I will continue to move forward as time permits.

          • marciok01 says:

            I will probably have to give PS a try. Started with Lightroom, especially because of its workflow / interface – I can select / discard / edit a 300 shot session to a 20-30 pictures output in about an hour; it’s file management is very good. But it’s tools are somewhat lacking, especially in sharpening and noise reducing. Tried DxO Optics Pro too: it’s lens correction profiles are amazing, and with some pictures the output is much better than LR, but it’s interface is a completely mess; complicated, the fonts used made it hard to separate the developing sections, and (for me) the workflow is MUCH slower than LR. Sharpening is not very good, and its PRIME noise reduction really kills a lot of noise, but details are gone too – and I dig the intention to only shows some image enhancements only viewing the picture with >75% enlargment, but for me is a major annoyance.

            Will giva a shot with PS, but don’t know how PS deals with shoots with a lot of pictures (have no clue, in fact). But I know that it’s denoiser is good, and it’s sharpening is top class.

            • You can do the selection justas fast in Bridge, and if you don’t need PS, it’s possible to save directly from ACR – again, just like LR. But if you need the power, PS has it…

              Quantity of images? I have to have very high throughput otherwise I’d get buried with the post processing. A heavy shoot for me might be 5-6 thousand in three days; I’ve been known to change cameras in the past because I wore out the shutters. In the space of a year or two. And I have multiple bodies…

  69. Really strong essay, Ming, and a really solid set of images, too! Thanks!

  70. thuy pham says:

    stunning pics


  1. […] suggestion made by knickerhawk in the comments section of an article a little while back on achieving visual consistency, here’s both a little test/ exercise for you and a little more expounding on the idea of […]

Thoughts? Leave a comment here and I'll get back to you.

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 35,785 other followers

%d bloggers like this: