To Reveal or Conceal (Dennis Aubrey)


This post originated from an interesting conversation that I recently had with Robert Consoli, who has a site called Squinchpix. Robert has as passionate a dedication to documenting and archiving European archaeological and architectural sites as PJ and I have in our more narrow field of specialization, the Romanesque and Gothic churches.

The conversation centered around post-production processing, in particular High Dynamic Range Photography (HDR). We have discussed the use of such techniques in a previous post, and have stated our intent as follows: “When we use post-processing techniques, our objective is to create an image that is otherwise impossible to represent with a camera but reflects what we see with our eyes.”

Robert concluded as follows: “You say that you ‘tend to like the shadows and dim recesses’. You are certainly entitled to your own vision but I can’t resist noticing that photography should reveal, not conceal.” Robert added, “I painted for many years and the painter’s motto is ‘full color in the shadows’. Photography’s no different.”

I thought about what Robert said about revealing and concealing and decided that the issue required more reflection. And from that reflection came three perspectives on why PJ and I shoot the way we do.

The first perspective is technical. PJ and I typically shoot with the smallest aperture of the lens that provides the requisite quality image. On my Canon EF TS-E 17mm f/4L tilt-shift lens that setting is usually f16. This means that we shoot long exposures, which keeps the entire image in focus. I feel that this is important because our human vision does not blur what we are looking at, only what we are not looking at. We can mimic that in photography so that when we look at any part of an image, we see it in focus. That can cause a problem in the overall aesthetics of the image, however, because the two dimensional canvas appears flat. We use shadow variation to separate the different elements of the image such as the foreground and background, thereby enhancing the sensation of depth in that canvas. In the image that follows, PJ has an interesting take – the foreground is shadowed and the background is light.

Basilique Notre Dame, Paray-le-Monial (Saône-et-Loire)

The second idea is architectural. PJ and I are preparing a book called “Light & Stone” about the Romanesque churches of France. She came up with the title because it is the interaction of these two elements that characterize all the churches that we photograph.

Even the history of the development of the churches is about these two elements. The medieval builders searched for a way to create a structure constructed of – and roofed by – stone in such a way that would allow light to enter the interior. Because of the immense weight of the stone vault, the walls were thick and permitted only the smallest amounts of fenestration. It was only when they developed techniques of channeling the thrust from the vaults down ribs into localized areas of buttressed walled surfaces that the builders could create the walls thin enough to allow for large windows. This, of course, marked the transition from Romanesque into Gothic.

It is our belief that the interaction of light with the spaces that we photograph is an important component of each image. Sometimes it might yield the “God-light” effect.

Basilique Saint Austremoine, Issoire (Puy-de-Dôme) Photo by PJ McKey

In any event, the interiors of Romanesque churches are shadowed and subdued, reflecting the meditative and contemplative intent of the monastic builders – ut in omnibus glorificetur Deus, “that in all things God may be glorified”. I read once that even if these monks possessed the technology to build a Gothic church, they still would have built in the Romanesque style because it suited their spiritual intent. Our photography explores that intent as it is visible in these structures eight hundred years later.

Abbaye de Bénédictins de la-Transfiguration-du-Seigneur, Saint-Guilhem-le-Désert (Hérault) Photo by Dennis Aubrey

The third idea is aesthetic. It seems to me that it is wholly appropriate to tell the story of this architecture through the medium of light, the medium of photography. In some cases, the church is not the subject, but the setting. In such a case, the drama of the photograph supports the drama of the story.

Basilique Saint Hilaire, Poitiers (Vienne) Photo by PJ McKey

This sometimes results in an image that is not completely intelligible, or is even abstract, but it is still telling the story.

Candlelit procession of the Festival of Sainte Foy, Basilique Sainte Foy, Conques (Aveyron) Photo by Dennis Aubrey

In support of this abstraction, I can do no better than to quote Angelico Surchamp once again; “Don’t you think, gentlemen, that abstract art, by transferring the sense of reality, promotes access to the sacred?” (“Ne croyez-vous pas, messieurs, que l’art abstrait, en transférant le sens de la réalité, favorise l’accès au sacré ?”)

So the issue of “revealing” has larger meanings than simply to show the church. What might be interpreted as “concealing” can also mean selectively “revealing” both physical and non-physical aspects of the church.

There are probably more ideas lurking in here somewhere, but this is what came up immediately in thinking about Robert’s comments. And to be fair to Robert, his purpose in the exchange was not to provoke this kind of intellectual diversion: “I used to maintain a blog about HDR and low-contrast was one of the topics I used to address. I brought that picture into Lightroom and looked at the histogram. Sure enough all the tones were below the middle except for a thin spike on the bright end (the windows). Classic. We should discuss this topic because lots of people (and I’m sure among your readers) have a hazy understanding of this (including me) and we ought to hash it all out if there’s an opportunity.” Maybe that’s another post!

Note: I found an interview with Robert at this site.

20 responses to “To Reveal or Conceal (Dennis Aubrey)

  1. I hesitate to weigh in on this discussion among three brilliant photographers and aesthetic theorists. I know little or nothing about the techniques of photographic art and read this discussion as a kindergarteer piano student listening to Beethoven or Bach. My comments come from the different discipline of theology. Light and darkness, revelation and concealment, sight and blindness are central not only to my Judeo-Christian tradition but to vertually all religious traditions.

    What is “revealed” – i.e, what we can see – is a pittance compared to what remains “concealed” from human eyes. To be human is to shudder before the light and before the darkness/the unknown. “Immortal, invisible, God only wise, in light inaccessible hid from our eyes, Most blessed, most glorious, the Ancient of Days, Almighty, victorious, Thy great Name we praise.” For each of the three Abrahamic traditions, YHWH, Allah, God remains beyond human grasp. Another expression of it is the story where YHWH (the NAME that cannot be pronounced) “passes by” Elijah hiding int he cave, but Elijah can only see God from the back. He cannot look upon God’s face.

    What does that have to do with photography and art itself? I don’t know.I leave that to you, Dennis, PJ, and Robert. But, it seems to this kindergartner that the shadows are as important as the light. Thanks for your fine work and for this thought-provoking post.

  2. My problem is that I talk too much. People who live in glass houses shouldn’t throw stones and my house is even glassier than most. Dennis and PJ don’t need to defend their artistic vision at all. Period. Particularly when we consider how they have realized it.

    The idea of shadows in image-work, painting or photography, is interesting. Shadows are crucial. It’s shadows, not color, that give form as every painter knows. The best writing I know of on this topic is Marcia Hall’s Color and Meaning in Renaissance Art. In that book Ms. Hall discusses the four major modes of Renaissance painting in terms of lightness and dark, or value. Fashion in painting (and photography) swings back and forth on exactly the topic of how to employ value. You can produce pictures that emphasize extreme lights and darks (‘chiaroscuro’); pictures that emphasize middle tones (see any photography by Joyce Tenneson for examples but the paradigmatic example is da Vinci’s sfumato); or pictures that strive for a sane balance between the two extremes with fully developed mid-tones and full color in the shadows (Raphael is often used to illustrate this point). The chiaroscuro paintings are usually dark and employ flashes of light for drama. Caravaggio, Rembrandt, and Goya are painters of this type.

    Caravaggio is an illustrative case of changing fashions in light and shadow. Before his time painters often painted clearly over the entire surface of the canvas. Any fifteenth-century International Gothic altarpiece will illustrate this idea. I think of this as the fully articulated surface. In many of the productions (not all, to be sure) of the fifteenth and sixteenth centuries an even, anodyne light washes over the entire picture frame and illuminates the subject without unnecessary drama or fuss.

    With the coming of the Reformation the Roman Catholic church reached out for a different approach. Their general idea was that these new upstart Protestant sects were improvised religions with no real tradition or legitimacy. In order to get that across to potential converts they needed a kind of pictorial propaganda that would shove the major characters and events of Catholicism in the face of the viewer in an insistent message of the legitimacy and long history of that religion. In the stripped down, darkly dramatic, and simplified canvasses of Caravaggio they found exactly what they were looking for. Many (not all) of his canvasses can be understood in the same way that we understand the ‘Uncle Sam wants YOU!’ posters of WWI. His dark or ‘tenebrist’ manner took the world by storm and dominated painting for the next fifty years. This ‘partly articulated surface’ has the virtue of directness; everything extraneous to the message is unceremoniously pushed overboard.

    We of the 21st century are used to this kind of presentation. We are accustomed to being propagandized to the point where we don’t even notice it any more. Page through any magazine and really look at the ads; much of the time they’re stripped down, punchy, and to the point. That’s a look that we’re used to.

    The improbable and startling thing about HDR is that, because it can deliver full detail in the shadows and the highlights, it has reintroduced, after four hundred years, the fully-articulated picture surface. That’s part of what makes its look so odd. This is revolutionary. A picture with a fully-articulated surface has to be approached, by the viewer, in a completely different way from a picture with a partly-articulated surface (that is, the kind that we’re used to).

    It comes down to more than my fat-headed remark about concealing versus revealing. Dennis speaks in praise of the activity of revealing and concealing as an effective way of conveying the spiritual atmosphere of the various church buildings. He’s right to do so. I’ve said similar things myself here: http://robertconsoli.blogspot.com/ (in RHC10 on stained glass). If Dennis’ and PJ’s vision is of the contrast between the darkness and light then HDR with its uniformly developed surface might not be the appropriate technique for them. But I reemphasize that they are the best ones to judge.

    • What a discussion this has turned into – Caravaggio and chiaroscuro. Bob, there is no issue about talking too much and your remarks were anything but fatheaded. This was not a defense of the artistic vision, but an articulation brought about by the issues that you raised. It was a pleasure to begin the conversation.

    • In this overly-long post I think I stepped on the punchline. My point was that HDR has, in effect, converted chiaroscuro into a kind of mandatory sfumato. Dennis speaks up for chiaroscuro and as a result HDR may not be the right technique for this kind of photography. That was my intended point. Now I’m done!

      • Bob, I am hesitant to pursue this line because I have never painted, but you’ve got me going. There is no question that we prefer the chiaroscuro approach (and by the way, I loved the way that Leonardo referred to light-shadow as a single entity) but I’m trying to understand the idea of HDR as sfumato. There are two classical ways to create gradual or imperceptible transitions in paintings. Sfumato, as I understand it, is a way of creating those transitions between different color tones and unione refers to a way of creating those transitions between light and shadow. Do you agree?

        HDR is the process of combining multiple exposures of standard 16-bit digital images into a single file that represents the aggregate dynamic range of all the source files. In order to express that expanded tonal range, HDR processing expands the bit depth of the new file to 32-bits. This leads to an infinite theoretical maximum dynamic range, and corrects the limitations of the image sensor in trying to capture high contrast scenes. The problem is that it is impossible for monitors or printers to display 32-bit images, so the HDR file must be tone-mapped to display the expanded range of 32-bit luminance values in a 16-bit display environment. Tone mapping algorithms generate the maximum contrast in the displayable range while preserving the appearance of the color and the details of the HDR image. What this means is simply that what we call an HDR image is actually a standard 16-bit image representing more tonal range and detail than was captured in the original camera file.

        So fundamentally, the way I see it, the tone-mapping process allows us to make choices in creating transitions between light and shadow (chiaroscuro), which does make it a legitimate technique for our work. What I am not sure of is whether tone mapping allows us to create the imperceptible changes associated with unione or sfumato. On this matter, I need help.

        Sorry to get geeky, but my previous career was as as founder/CEO of a software company that worked with fractals in digital image space 🙂

      • Let me start by clarifying what I meant when I used the terms ‘sfumato’ and ‘chiaroscuro’. These terms have been used in different ways at different times so I’m going to pick meanings and stick to them. ‘Chiaroscuro’ refers to paintings that are predominantly characterized by extreme high and low values with relatively few or no middle tones. The histogram of such a picture is characterized by a peak at the high and/or the low frequency ends and a pronounced dip in the middle. An example is Sebastiano del Piombo’s Pieta in the Museo Civico at Viterbo. Its histogram is one with a huge spike on the low end of the spectrum with tailing off long before the middle and barely-represented high frequency tones. You can try this yourself if you care to. The result of ‘chiaroscuro’ is usually dramatic in the extreme. It is often like looking at a scene ‘by flashes of lightning’. In the language of photography: ‘high-contrast’ and when we see a ‘high-contrast’ picture (whether or not it makes sense to characterize it as ‘chiaroscuro’) we know that the middle values will be relatively poorly represented in its histogram.

        ‘Sfumato’ is the opposite. It’s characterized by softly modelled mid-tones with few or no genuinely high or low values. The typical histogram in ‘sfumato’ is often characterized by well-represented mid-tones with tailing off long before either end. The Mona Lisa is the example usually given of this technique. The over all effect is often peaceful and calming. I took a photo by Joyce Tenneson and looked at the histogram; it’s characterized by values above middle
        gray but still tailing off well before the bright end of the scale. The Mona Lisa is lower keyed but still has poorly represented extreme darks or lights. The general look of sfumato pictures, if we characterized them in the language of photography would be to describe them as low-contrast’. When we see a low-contrast painting or photo we know, just by looking, that the extreme values will be missing in the histogram.

        Now, before I go any further, I emphasize what you already know. There are infinite styles of pictures in the world with infinite gradations between them. I’m speaking of general categories, not hard and fast rules. Most visual production falls somewhere between these two extremes but it’s still true that when we see high-contrast or low-contrast pictures we can easily tell the fact and make good guesses about the resulting histograms. And the reverse is true. When we see only the histogram we can usually tell whether the picture that produced it is high or low contrast.

        Now the terms ‘chiaroscuro’ and ‘sfumato’ are my short-hand way of referring to these ideas. I should confine myself to ‘high-contrast’ or ‘low-contast’. I think that this makes more sense in discussing HDR (even though I myself introduced the terms from painting).

        Bob, I am hesitant to pursue this line because I have never painted, but you’ve got me going. There is no question that we prefer the chiaroscuro approach (and by the way, I loved the way that Leonardo referred to light-shadow as a single entity) but I’m trying to understand the idea of HDR as sfumato.

        There are two general cases which might lead to the use of HDR processing. There may be a single image with important high-value or low-value detail and which you want to reproduce on a target medium.

        Secondly, the scene itself may contain interesting detail in value ranges beyond the capability of the camera sensor, e.g. photographing good detail in stained glass while wanting, at the same time, to get a good representation of the interesting art work on the wall underneath the window. This requires more than one image.

        I start with the single-image case. The basic idea behind HDR is to map a source image (with a ‘large’ value range) onto a target (with a ‘smaller’ value range). The target is usually a television or computer screen or ink on a piece of paper. This mismatch in tonal ranges is a problem. Without some fancy maneuvering the bright or dark values (usually both) in the source will be clipped to whatever pure black or white the target is capable of representing. We would
        automatically have a high contrast picture with large over-exposed and under-exposed areas most of the time. Without HDR that is very often what we DO have. The average camera or scanner is usually capable of producing an image with a 12 or 14 bit range of brightnesses. The usual problem presents itself as a mapping of THAT range onto a target space of 7, 8, or 9 bits. The average mismatch is on the order of 4 or 5 bits and that represents about two and a half stops worth of brightness which would be lost through clipping unless we did something about it, such as a remapping of the brightness range.

        Now, in one sense, it’s easy to do this remapping. We take the highest and lowest values in the source image and reduce them (and, proportionally, all the intermediate values) to fit into the target range. Easy. No clipping. The average engineer could write this algorithm over his morning coffee and no one would be the wiser. Except for one thing.

        The result would be catastrophically low contrast. From a high-contrast picture it would instantly become ‘sfumato’; it would look as though fog had drifted in front of the camera. Smoke indeed!

        Now I’m going to digress and talk about the origin of the word ‘sfumato’. In Italian it means ‘smoked’ or ‘smoky’ and, indeed, such paintings or photographs do look as though a fog had drifted in front of the subject and the camera had captured it. Let’s experiment with this idea. Let’s say we did set up a model, positioned the camera, hooked it up to the computer, set up the lights, etc. etc. and then someone in the studio started smoking a powerful cigar. Now there’s smoke between the camera and the model. What effect does that have? Why should the resulting histogram have no extremes? Because the smoke would partly block the highlights in the scene lowering them (pushing them to the left) in the histogram. But that same smoke would add reflected energy to the darks and so lighten them and pushing their values to the right. Instant low-contrast. This is exactly what would happen in our naïve HDR remapping.

        In a 14-bit source image the middle tones occupy, let’s say, 8 bits. If we just squeezed the high and low tones together those same 8 bits would wind up occupying only 5 bits in the target output. They would just be pushed-together mush. And, believe me, you would instantly perceive it and you really wouldn’t like it. With multiple-exposure HDR theproblem would be even worse. I’m just going to pick numbers. With multiple images you might have a total brightness range of 20 bits, let’s say. But the middle tones that you’re interested in still occupy only 8 bits of that 20 bit range. Now we squeeze 20 bits, say, into 9 and the middle tones go from 8 to 3. All numbers are notional and approximations only. But in our example there wouldn’t be much left of the mid-tones. And, in scientific language, the result would suck. The wider the bracketed range that you produce the worse the middle tones would become.

        Having said all this, then, it’s clear that the real challenge that the HDR engineer sets himself to address is to come up with a way of remapping the brights and darks to suit the target medium, but not squeeze the middle tones together into a contrastless mush. HDR tone-mapping is the answer; tone-mapping is the art of automatically moving some values and not others for the same target representation.

        There are different ways to do this and, as a result, many different software packages and approaches. But the basic goal is always the same: to avoid crushing low-contrast in the mid-tones.

        Here’s your letter with my notes and reactions:

        There are two classical ways to create gradual or imperceptible transitions in paintings. Sfumato, as I understand it, is a way of creating those transitions between different color tones
        (This is fair and correct) and unione refers to a way of creating those transitions between light
        and shadow. Do you agree?

        (I’m sorry. I don’t know what this last clause means. I know what unione is but I’m not following your assertion.)

        HDR is the process of combining multiple exposures of standard 16-bit digital images into a single file that represents the aggregate dynamic range of all the source files.

        (Fair enough but there’s also single-image HDR which I referred to above.)

        In order to express that expanded tonal range, HDR processing expands the bit depth of the new file to 32-bits.

        (Yes. My understanding is that the standard approach is to adopt a 32-bit space and map all the individual exposures onto that one space.)

        This leads to an infinite theoretical maximum dynamic range, (well, 4 billion values, anyway)
        and corrects the limitations of the image sensor in trying to capture high contrast scenes.

        (Yes)

        The problem is that it is impossible for monitors or printers to display 32-bit images,

        (That’s right.)

        so the HDR file must be tone-mapped to display the expanded range of 32-bit luminance values in a 16-bit display environment.

        (More like 8 or 9 bit space, but yes)

        Tone mapping algorithms generate the maximum contrast in the displayable range while preserving the appearance of the color and the details of the HDR image.

        (Ye..ss. More or less. All tone crunching (HDR) is inherently low-contrast despite what I tried to say above about the challenge to the HDR engineer. Most algorithms do their best to preserve contrast in the middle tones but low-contrast really is the HDR look. Restoring that contrast (almost always in the mid-tones) to a desirable level is often just up to the individual processing the image after they’ve applied the HDR software.)

        What this means is simply that what we call an HDR image is actually a standard 16-bit image representing more tonal range and detail than was captured in the original camera file.

        (Yes)

        So fundamentally, the way I see it, the tone-mapping process allows us to make choices in creating transitions between light and shadow (chiaroscuro).

        (If I understand you, yes, and see my first remarks.)

        What I am not sure of is whether tone mapping allows us to create theimperceptible changes associated with unione or sfumato.

        (Let’s leave ‘unione’ out of it for the nonce. I don’t understand quite why you have introduced the idea of imperceptible changes in tone but I suspect that I may have answered this question above. I’ve cast some aspersions on sfumato above by implying that low-contrast is bad. It isn’t, if that’s your intention. Sfumato, under the hand of Da Vinci, became very influential because of Da Vinci’s ability to represent subtle changes in shading caused by the delicate muscles inthe face and, in particular, around the mouth. That’s what we see in the Mona Lisa. It appeared that with sfumato a delicacy of range was now available that hadn’t been possible before. Remember that I’ve identified sfumato with ‘low-contrast’ and if that subtle and smooth
        transition (you may also want to stop sharpening in cases like this) is what you want (this might be what you mean when you say ‘imperceptible changes’) then HDR and low-contrast is the way you want to go. But, if you do, bid spiritual drama goodbye.)

        On this matter, I need help. Sorry to get geeky, but my previous career was as as founder/CEO of a software company that worked with fractals in digital image space.

        (Sounds pretty high-powered to me!)

        (Remember, I’m not speaking ex cathedra but in notional terms. For example, I wrongly implied in my previous post that there was no tenebrism before Caravaggio. That’s wrong; tenebrism has a long and honorable history (in the form of chiaroscuro experiments, just to name one)).

        Final Remarks: When we go from a partly articulated space to the fully-articulated space that HDR makes possible then what happens to the traditional rules of composition? THAT is a crucial question and deserves a post of its own.

        Can hardly believe that all this helps but anyway…

        Best,

        Bob

  3. iI very much enjoyed this discussion, like the kindergarten pianist listening to Beethoven, I am in awe of all the knowledge displayed. In painting, as in music, and my other arts, patchwork and stained glass, contrast, tonal values are paramount. My husband also does wonderful “paintings with the needle” as well as in pastels and we consistently fall out when I suggest that he works from monochrome photographs, simply to get the half-tones correct. Sorry, my two-pennworth is pretty well irrelevant to your discussion… I plan to put my amateurish photos of some of our stained glass work on my blog, this weekend, which will show what I mean clearer than words.

    • Viv, looking forward to seeing the stained glass. The decision process in stained glass must be very complicated because you not only have the relationship of colors in the glass itself to consider, but the effect of the light “projecting” through. I’ve often wondered how much the mixing of the colors in this manner was considered by the medieval craftsmen who created the windows of Chartres and the other great cathedrals. Thanks for joining in.

    • Your remarks are not irrelevant at all. Monochrome ideas pervade all image work. The standard method of the old masters was to paint an entire painting in monochrome (‘grisaille’) and, once the tones (values) were right, to glaze the color over afterwards. One way of post-processing photos which I use in Lightroom is to set the photo to ‘black and white’ and then work on the tones until the photo looks something like Ansel Adams would have produced. Then you can convert back into color and correct color casts at that stage. One problem with Lightroom is that it doesn’t allow (as far as I know) you to modify value without changing the saturation. You have to do that in a separate operation. I’ll go and look at your blog.

  4. Gentlemen:
    Your continued discussions are fascinating explorations of aesthetic theory and most helpful in evolving an imaging style. We cannot eliminate shadows: imagine Casablanca without the deep black backgrounds, or Ansel’s forrest scenes minus the foreboding darks. A cathedral needs both light and darkness….it was drawn (literally) from the archetypal recesses of our minds as a space wherein we can encounter the eternal: ethereal light from primordial darkness, like stars on a canvas of black night. One needs the other. Yet hidden details of architecture demand illumination for their complexity to stimulate the senses.
    Please carry on with your dialogue; there could easily be a very long article or even a book come from this.
    Stephan Thomas Vitas PhD
    Washington DC

    • Stephan, thanks for your contribution to this ongoing discussion. As Bob was saying, it really depends on the use of the images. PJ doesn’t care that much for documentation, she is the quintessential artist, looking to respond to what she sees and then uses the camera to record her reactions. I am more concerned about the documentation and if, anything, probably lean more towards the idea that we need to see the hidden details. But fundamentally, PJ and I have such strong emotional reactions to the churches that those feelings guide our photography and certainly the writing.

      • Dennis: then there’s your book. Cathedral photos by all three of you with theoretical commentary. From what I’ve seen/read thus far there’s very worthwhile material.
        Stephan

      • Stephan, thanks. PJ and I are working on a book called “Light & Stone” – the photography for it will be complete in October. The text is almost complete now. We appreciate your support on the project and welcome you as a fellow traveler in the Romanesque world that we inhabit.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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