Lewis Downey: Zone System

Sat, 25 Dec 2004 12:00:37 GMT

Merry Christmas!

Thu, 09 Dec 2004 19:40:20 GMT

David Kachel's new web site.

Tue, 02 Nov 2004 09:38:58 GMT

please vote

Sat, 02 Oct 2004 04:48:12 GMT

BBC US photography icon Avedon dies

Wed, 29 Sep 2004 21:13:49 GMT

art.transindex.ro

Reuters Famed Photographer Avedon Suffers Stroke in Texas. LOS ANGELES (Reuters) - Famed portrait photographer Richard Avedon has suffered a brain hemorrhage while on assignment for the New Yorker magazine in Texas and has been hospitalized, a spokeswoman for the magazine said on Wednesday. [Reuters: US Domestic News]

Thu, 05 Aug 2004 14:27:06 GMT

http://www.henricartierbresson.org/

Thu, 05 Aug 2004 13:45:19 GMT

Yahoo! Tributes pour in for Cartier-Bresson (AFP). AFP - Lavish tributes were paid to Henri Cartier-Bresson, the pioneering French photographer and co-founder of the Magnum picture agency who died earlier this week at the age of 95. [Yahoo! News - Reader Ratings]

Wed, 04 Aug 2004 20:12:53 GMT

Boing Boing Photographer Henri Cartier-Bresson dies. Legendary French photographer Henri Cartier-Bresson died Monday at the age of 95 in the South of France. Some web resources: his Foundation, Photology, Tete a Tete. Among his great works were portraits of Matisse, Bonnard, Braque, Rouault, Claudel (at the end of the Second World War). (Merci, Jean-Luc, who adds "His foundation website is so slow at the moment because just about everyone online in France is hitting it right now.") [Boing Boing]

more...

Reuters French Photo Legend Cartier-Bresson Dead. PARIS (Reuters) - Frenchman Henri Cartier-Bresson, one of the great photographers of the 20th century and a founding father of modern photojournalism, has died aged 95, family friends said Wednesday. [Reuters: Top News]

BBC Cartier-Bresson dies at 95. Legendary French photographer Henri Cartier-Bresson dies, weeks short of his 96th birthday. [BBC News | News Front Page | World Edition]

Yahoo! Photographer Henri Cartier-Bresson Dies (AP). AP - Legendary French photographer Henri Cartier-Bresson, who traveled the world for more than a half century capturing human drama with his camera, has died at age 95. [Yahoo! News - Reader Ratings]

Wed, 23 Jun 2004 16:58:29 GMT

Boing Boing Photographers' bust card. Here's a great printable one-pager that describes what you're legally allowed to take pictures of, and what to do if someone tries to bust you for it.

148K PDF Link (Thanks, Tom!) [Boing Boing]

Tue, 02 Sep 2003 23:55:06 GMT

more on textures Hi Marc, I will do a little thinking out loud here. As usual I understand you will mine these thoughts for anything useful -- but feel free to discard any or all of it. I think we should speculate that a texture of lower contrast becomes visible (light tones) discernibly slower than a texture of higher contrast. Likewise we should speculate that a texture of lower contrast becomes invisible (dark tones) discernibly faster than a texture of higher contrast. That could be a wrong guess, but it seems probable and easy to check. If it is true, then which luminance range do we chose for a standard texture? Seems arbitrary. Maybe emergence times for maybe three textures with a range of luminance ranges. that sort of information would also begin to describe a toe and/shoulder on the paper. Following this a little further, I can imagine a scenario where a texture with more contrast is discernible as a texture but a texture of less contrast is no quite a perceptible texture -- more like a gray tone. Perhaps, conceptually, with the proper selection of luminance range(s), that scenario is the one that is noted as the definitive emergence time:

light gray (dense negative) -- print values around zone 8:

condition 1) texture a is only a gray (solid looking)
condition 2) texture b is a perceptibly a texture.
(I am less clear about the practicality of dark tones) dark gray (thin negative) -- print values around zone 2

condition 3) texture c is a perceptibly a texture.
condition 4) texture d has become solid looking

Not sure here whether texture b needs to be the same for both, but am taking that as a starting point. I am also not sure if I am thinking about three textures in this wanderment or two textures or four textures. My guess is four or more. That then begs the question of how much the paper contrast goofs the anticipated behavior -- will keep noodling this. Obviously the paper contrast is going to cause the overall (i.e.. average) dark tone texture densities to vary. In other words, a paper of high contrast will need sets of average densities (forgetting the luminance range within each texture for the moment only) that are closer together than low contrast papers. An approximate rewording of the forgoing is that high contrast papers will be need textures of lower *average* density than lower contrast papers. Amongst many characteristics of a texture, two are:

contrast, i.e. luminance range, within a given texture (film)
average density within a given texture (film)

The contrast I think we can standardize, even if arbitrarily. The contrast range of the paper is largely what we are after. We will need many sets of textures. Those sets may be grouped into calibration groups useful for marking the conditions sited earlier. There is a special texture within a calibration group (the one that either just appeared or is about to disappear depending on whether the tone in question is light or dark). The special texture has an average density. The paper contrast is somewhat defined by the difference in contrast between the average density of the emerging density and the average density of the almost disappearing density. We still have to work out how much exposure time though. I might have just worked myself back toward min_exp_time_to_max_black. Apologies, I probably see this more clearly than I am wording it, but have to go. All of this will vary in part with the paper developer and development time too of course. Lewis ps. The luminance range of the reproduction (print) will vary dependant on the paper contrast range Textural luminance should be tuned to the negative. I do not currently have a clear insight into average contrasts of textures on papers of varying contrast range. Perhaps that is a notable characteristic of a texture -- for instance: how does it look when it is perceptibly a texture and has a certain average density (paper). I cannot quite see through that relationship right now.

Fri, 29 Aug 2003 17:58:55 GMT

More about textures from Marc Torzynski:

There is a proverb (?) in France: << La nuit porte conseil >> (which means "let pass some time and things will go clearer by itself"). Another one states that if you are in trouble, better ask some friends about the subject. To get better help, I put them in practice together and have some discussion with a friend of mine last night. She's not a scientist but works at a contemporary art museum and she had several ideas on what makes a texture and what makes an image. So we summarize/formalize the question in this way: 1) Shapes, objects, forms (the whole world in fact), no matter of being bidimensional like for example a picture or tridimensional like for example a sculpture, are seen by the eye because they exhibit a spatial luminance distribution (which may be approximated by saying "a light and shadows distribution"). 2) We will define an _scene_ as a lights and shadows distribution which leads to something "intellegible" (a face where lights and shadows helps to identify the nose, mouth, eyes etc., a landscape where we can identify grass, forests, lakes areas, etc.). In a simplified, but operative model, a scene can be decomposed in a set of several _different_ objects, each of them having its own texture. 3) We will define a _texture_ as a lights and shadows distribution which leads to an "unintellegible" image where we cannot really identify what is the object (a close-up of grass may, for example, by as well seen as hairs, a detail of a cloud as a piece of cotton ball, etc.). A texture may be decomposed as a [random] distribution of bright and dark _identical_ elements (this again is a simplified model). 4a) A scene may be "contrasty" or not (a black cadillac drived by Abby vs. a grey car in a misty day). It is the same for a texture (sand is a flat texture but salt and pepper is a contrasty texture). 4b) The contrast of a scene may be quantified by its luminance range, by looking for its darkest and brightest areas. The same for a texture, except that black/white areas are now "micro-areas". Practically, measuring the contrast range of a scene may be done with a spotmeter. Measuring the contrast range of a texture could be achieved in the same way. Just is it more difficult, because the spotmeter has to focus to these micro-areas (difficult because common spotmeters are not designed for this purpose, but it is just a question of technology). 5a) We may easily also define a "scene brighntess" which would be the luminance measured over the whole area. It is the luminance measured not by a spot meter, but by a wide-angle photometer. Brightness is in fact the mean value of the luminances of all the several different areas (so brightness will be low if there are many dark areas with just a few bright ones, etc.) Usually zone system users are not at all concerned by this "integral" measurement. 5b) Texture brightness may be defined in exactly the same way. Texture brightness measurment is in fact exactly what every zone system user does when it measures the brightness in the different areas of interest....

Thu, 28 Aug 2003 14:59:06 GMT

Here is another excerpt in the conversation with Marc Torzinski about a laboratory definition of a texture and using the emergence of textures in the calibration of film tests. Interesting stuff.

Marc: "Hi again! Regarding your first mail, I am not sure that I am really willing to model a texture creating it trough CAD program (Photoshop, 3D studio...) because I am convinced that we will be fooled by such artificial modelling. The idea of making a "laboratory textured object" by use of pyramidal shapes seems more interesting; but if so, we may as well use the numerous textures of the real life, and this is maybe the best way to go: after all, the goal is to get texture in our shots when we plan to do so... On your second mail, you point an intersting fact that I was for some time aware of: you say you prefer a zone 1 density slightly higher that the common value of 0.1. Maybe is it just your personal preference. But maybe also have discover by trial and errors a point that may be explained in sensitometric terms. For example, let's suppose your paper has a long shoulder (a perfect, however extreme case, is the grades 0 & 00 of Ilford Warmtone - have at look at the characteristic curves at http://www.ilford.com/html/us_english/pdf/warm_tone.pdf). A negative having density of 0 above base+fog will of course be printed as pure black, but because the shoulder is long, even with a density of 0.1 you are still in the nearly pure black domain. In other terms, for such a paper, zone 0 is very wide. With this paper, if you plan to get a dark but discernable gray at zone 1 you have to calibrate your film with zone 1 target at more than 0.1, maybe up to 0.24 ! (this was pointed to me by Thierry Fumey http://www.cepv.ch/annuaire/profs/fumey/zonesyst.htm - article in french, sorry) From the above two paragraphs, you may understand how interesting it will be to develop a method taking into account the primary definition of zones when matching paper to film: instead of - no mattter how - measuring the extreme points of pure white/pure black for a given paper, I am trying to measure the extreme points of where textures are still discernable. As you point out, you may exercise yourself to perceive very slight gray differences nearly pure white and pure black. This is just the delicate point: such an ability is certainly perfect when looking at pictures (like a music lover able to discern subtle tone differences), but may also be a handicap when, for instance, you examine a strip test in order to determine what you call "min_exposure_for_maximum_density" in your matching procedure: if you look too carefully, or with a brighter than usual light, you will end with an exposure time and a Max black _far inside_ zone 0 instead of determining just the beginning of this zone... In my opinion, calibrating against textures should be less dependant from the viewer abilities and is more adequate for photography because we prefer of course scrutinize on detailed areas rather than on wide pure black or pure white areas (which in a photograph are usually here specifically to help/lead the wiever to focus on the other -detailed- areas...) Of course, this is, as you may say, just a matter of taste and someone may prefer to work using the common practice (determining Dmax, Dmin, and letting the rest - long shoulder/toe of paper/film be the "intrinsic nature of the things"). My goal when starting this reflexion was to avoid ending up with deceptive nearly pure white/pure black when I have planned to have suggestion of texture...:"

Wed, 27 Aug 2003 22:57:44 GMT

I am enjoying an excellent discussion with Marc Torzynski about defining a definition for textures. Marc writes:

You hit exactly the point! Let me first fully agree with your interpretation of texture as a result of the interaction between light and a three dimensional surface. As you point out, the appearance of a surface will depend on the light properties (very directionnal light from one side -> highly textured surface, diffuse light coming from above the surface -> nearly no texture at all). Your interpretation of the "strongness" of a texture in terms of lighting ratio between bright areas and shadow areas is also very useful. Back to the beginning: when I read your Zone system tutorial for the first time, I mean, the paragraph saying something like "zone 0 is full black... zone 2 is the first suggestion of texture, zone 3-7 are fully textured, etc.", all this was perfectly clear and this description seems to me a very rigorous one. But when I really start to thing about practical cases, I fell in trouble and I quickly understand that the definition of (for example) zone 1 is partly arbitrary: for example, if a highly textured object (a concrete wall illuminated from one side by the sun) just suggests its texture when placed at zone VIII, the same object, once poorly textured by a more diffuse light will show no texture at all when placed, once again, at zone VIII. I hope this example shows clearly why I am trying to define a "standard texture". According to your description, we have to define this "standard texture" by specifying the ratio between shadowed and illuminated areas. But, as you mention it, the strongness of the shadows areas depends on the multiple reflections between the bumps. It is nearly impossible to do any computer simulation because there are too much parameters. So I have absolutely no idea about the "stop range" of a textured object. Maybe should I modify a spot meter to make close-up measurments on several textured objects (wall, stones, etc.). Or use your idea about the "scalability" of the textures...

Marc is responding to a quick note with a drawing

I will post more if the conversation seems to go anywhere.

Wed, 27 Aug 2003 22:53:29 GMT

I do not know that we must define textures by the luminance range, but it might be a useful element in a definition. It seems to me there ought to be some way to model an extremely simplified texture. Maybe Photoshop or a 3D rendering program would be helpful in figuring this out. To get a handle on the basic properties, try reading a surface with shadows on it -- both the shadowed and illuminated parts. This subject could be a driveway and your shadow. Watch out for those specular reflections you alluded too; asphalt is full of them -- concrete much less so. I got slammed by the specular reflections in asphalt while calculating reciprocity corrections for a nighttime exposure. It definitely seems better to work at human scale to me; although I do not know how to model the thing. Maybe we could start with a bunch of little pyramids evenly distributed over a surface -- either touching or close to touching. Then consider the light -- first using directional light later diffused light. Even if the first model is grossly simplified, there must be some way of doing it, or do you think that the effort can never achieve practical results.

Wed, 12 Mar 2003 11:09:05 GMT

"zslist" There is a little discussion getting started in the ZoneSystem listserv about graded papers. (Permalink)