HDR, LucisArt and Dynamic Range for Photos and Art

One complaint which I sometimes hear from painters who base their paintings on photos is that photos typically have a limited dynamic range in comparison with the human eye. In other words, if you expose for the majority of the image, you typically lose out in terms of details, both in the shadow areas and the highlight areas.

Some painters have compensated by creating several different prints, some of which were overexposed and some of which were underexposed, and then made conscious choices based on the information found in multiple photos. Others have just compensated by making informed decisions based on personal experience, or on memories (sometimes enhanced with quick sketches and/or notes) of the way things really looked in the field, or on the way that things look in front of their eyes when actually creating paintings from life in the studio.

One reason for the success of certain film photographers such as Ansel Adams is that they've traditionally been experts at the use of dodging and burning in the darkroom in order to compensate for such defects. But most people didn't have their own darkrooms back in the days when such techniques offered the only way to effectively address the issue of dynamic range, and the few labs which offered such custom services charged very high prices.

Now, thanks to digital techniques, it's a whole new ballgame. The use of Photoshop's Levels, Curves, Layer Masks and other tools began to improve things quite some time ago. Then they introduced the Shadow/Highlight control, and that improved things even more. Finally, they introduced features (also offered in different variations by various third-party companies) pertaining to a new HDR (High Dynamic Range) technique which merged multiple bracketed exposures in order to create composite images which combined the best of all of the exposures. (When using studio lighting setups, one can do this with the use of layer masks --- the old fashioned way --- or with the HDR feature, which is more automated but with less hands-on control.)

Here's a link to one web page containing an article which will give you a better idea of the effects of HDR. I've seen better examples of the amazing results one can achieve with the HDR technique (mostly because I think he overdid the effect slightly, to the point that some of the image's drama was lost), but this page will at least help readers of this post from a conceptual point of view. If it had been me, I would have merged the HDR version with a little bit of the original "0 exposure" image, using an additional layer with significantly reduced opacity.

If you have an older version of Photoshop which doesn't yet have the HDR feature, there's a way to simulate that feature (without the need for multiple exposures, making it a great choice for subjects which are likely to move in-between exposures) using the High Pass Filter. Also, it has the effect of sharpening the image. Here's a link to an article which shows how to do this, as well as side-by-side before & after photos.

(NOTE: Some HDR software has features designed to eliminate the problems caused by trying to merge images in which certain elements move substantially. Sometimes those features work well, sometimes they don't, depending on the images.)

Another very cool option is to use LucisArt 3. When used intelligently, it can create amazing results (including some very cool special artistic effects). Like the High Pass Filter option, it can be used with single photos, with no need for bracketing. And like the other techniques, it can be overdone, so it's sometimes a good idea to blend it with portions of the original image. Here's a link to some good examples (in the form of online slide shows) of what it can do.

Lucis Pro 6.0 uses the same technology as LucisArt, but its primary application seems to be for technical (e.g., scientific) purposes, not for art. LucisArt costs less, and actually seems to be better for what I'd want to use it for.

If you view the online slide shows, you can see that some of the best artists are able to use LucisArt to produce images which look more like photorealistic paintings. In other cases, the images just look like extremely good photos. The choice is up to you.

LucisArt would seem to be a particularly good partner for another excellent program, EngraveSoft from Pixation.com, due to the maximization of details available from LucisArt. In other words, process the photo with LucisArt first, and then create excellent line art (which could be used as a basis for a variety of fine art images such as Solarplate photopolymer intaglio prints) using EngraveSoft. Consider, for example, the EngraveSoft image created from a photo of a man named Dan. Pretty nice, right? But notice that the original photo has some blown out highlights (i.e., Dan's forehead, on the left side of his face), which would almost certainly have been improved with LucisArt. Of course, when it comes to line art, a certain amount of contrast is what gives the image its drama, so one doesn't want to overdo it. But I think that the digital engraving of Dan would have been enhanced if the photo had first been processed with LucisArt.

NOTE: While intaglio printing with a process such as the SolarPlate process offers one way to create fine art prints from such digitally generated line art, another interesting option is to use what's known as polyester plate lithography. The most commonly used method is to use Pronto Plates, or a similar product offered by ZAcryl.com. Both products are designed to be run directly through toner-based laser printers, and actually used as printing plates for fine art lithography! (No need for separate exposures through digital negatives onto photosensitive materials.) The main drawback seems to be that one is somewhat limited in size. (For some strange reason, the "plates" sold for polyester plate lithography seem to exceed the maximum printable size for most toner-based copiers and laser printers, which is usually 11x17-inches or 12x18-inches at most. But maybe there are commercial toner-based imagesetters which can go larger. I'll have to ask about that.)

Anyway, I've read that Sharpie ink, applied to such polyester plates, holds the lithographic printing ink just like toner holds it (which, in turn, works roughly the same way that lithographic crayon, lithographic pencil and lithographic tusche work). I'm thinking that solvent-based markers from other manufacturers ought to also work, and that it might even be possible to make really large plates, for polyester plate lithography, by using a wide format solvent-based printer such as the Epson GS6000. That would be very cool! (LexJet.com offers a polyester film specifically designed to be printed on with dye-based and pigment-based printers such as the Epson 9900. If it also works with the GS6000, and if the inks from the GS6000 hold litho ink the same as toner, that could be a really good way to make enormous polyester "plates" for polyester plate lithography. Something worth checking out, at any rate.)

Of course, one might ask why bother, when the digital printer itself produces such great prints directly. Well, the answer, dear friends, is that there's still a ridiculous bias against digital prints in certain fine art circles (such as art contests, certain art fairs and galleries, etc.), so using digital techniques as a means of preparing to make prints with older processes such as intaglio printing, relief printing, planographic printing (lithography) or screen printing (serigraphy) offers a potential end run around such anti-digital biases. Just tell such people (truthfully) that it's an etching, an engraving, a lithographic print, a serigraphic print, etc. No need to volunteer the fact that your utilization of digital techniques enabled you to produce the print in a fraction of the time it would have taken by using older techniques involving laborious hand work. Of course, if you still want to use such techniques (possibly in conjunction with digital techniques), you're free to do that, too.


---------

Incidentally, on a separate but somewhat related note: For those who think that infinitely scalable digital vector art can never equal the appearance of photos, here's a link to a page which will blow your mind.

How did Mr. Miyamoto achieve such realism in Adobe Illustrator? Well, I'm guessing that he used the new Live Trace feature (at least in part), which does a much better job of autotracing photos than Adobe Streamline and other similar solutions did in the past. It would also help, undoubtedly, to have the best possible photo on which to base such vector art, which is where the aforementioned HDR/High Pass Filter/LucisArt/Lucis Pro techniques come in. But it's also clear that Mr. Miyamoto's expertise in the use of the Gradient Mesh filter played a big role as well. (Even the best Live Trace images can occasionally err when it comes to translating photos to vector perfectly. Sometimes they look a bit posterized. Based on what I've seen, I think that the Live Trace feature in CS4 is better than in earlier versions, though.) These images remind me of some of the best realistic airbrush art being created by Japanese illustrators (and a few Americans and others) back in the 80's. To say that there's a huge chasm between such images and the simplistic clip art images more familiar to most of us is an understatement.

In Praise of the Four Square Collage

More often than not, when it comes to picture framing, people purchase single frames for single pictures. But there are other options, particularly when it comes to smaller photographic prints.

For example, one can purchase "collage frames" which are basically frames in which multiple window openings have been cut into the "overmat", thereby enabling one to display multiple photos or prints within a single frame.

Collage frames can be particularly nice as a means of presenting small prints (such as the 4x6 prints typically made for very low prices at numerous photo labs) in a unified manner which makes a much nicer impression than a scattered collection of framed individual small prints. It's a lot easier to hang a single collage frame featuring sixteen small prints than it would be to hang sixteen small prints in their own individual frames!

(Click here to see some particularly nice collage frames from The Pottery Barn.)

As nice as collage frames can be, I can attest to the fact that it can be time-consuming and difficult to attach the individual prints to the undermat in such a way that they align perfectly with the window openings cut out of the overmat. A lot of precise measuring is necessary.

These days, digital imaging offers a superior alternative. That alternative is to digitally place multiple images in a precise arrangement on a single large background, and then send the file to a company which is capable of making a very large print from that file.

One advantage is that cutting the overmat is a lot easier. The overmat need not feature multiple window openings. A single window opening capable of showing the entire digital collage will suffice.

It's also much easier to mount a single print onto the backing board or undermat.

A third advantage is that one is not limited to printing on paper. One can just as easily send the digital file to one of the numerous companies which will now print one's image file onto canvas.

A fourth advantage, when it comes to a digital collage arrangement, is that the amount of space between each individual image doesn't need to be as large as it would need to be if one were cutting multiple window openings out of a single overmat. A thin line between each image would offer adequate separation, and such separation between images is not absolutely essential.

A fifth advantage is that the physical overmat does not actually cover the edges of any of the individual images in a digital collage. The images are not "cropped" by the overmat. This can be particularly important in cases where even the slightest image loss compromises a particular image.

The ideal, when creating a digital collage arrangement, is to create a reusable digital template which can be used again and again for similar projects, in order to aid in positioning the individual elements in the digital collage.

In Photoshop, each of the individual "placeholders" can be placed on a separate layer, and one can then use the alignment features in Photoshop in order to align each of the individual photos or artistic images with the placeholders to which they correspond. What do I mean by "placeholders"? Simply this: A "placeholder" would be a shape, with a solid color such as black, which was the exact same dimensions as the image one intended to put into its place. (Once the placeholder from the template has been replaced or covered by the actual image, the placeholder can be deleted.)

The question then arises as to what the best possible arrangement might be. There are infinite possibilities, including arrangements in which the individual images overlap one another or even blend with one another. But I prefer an arrangement which meets the following criteria:

1. The collage arrangement features an equal number of horizontal (landscape mode) images and vertical (portrait mode) images, so that the arrangement itself is independent of the particular images displayed in that arrangement.
2. The collage arrangement is well balanced in terms of the size of the exterior borders between the images and the edge of the paper or canvas, and in terms of the amount of spaces which separate the individual images from each other.
3. The arrangement is not perfectly symmetrical. Perfect symmetry tends to lend a static feel to any composition, so a collage arrangement which looks like a perfect grid tends to be a bit boring, even if the images themselves are quite nice.

With those criteria in mind, I came up with what I like to describe as a "four square collage arrangement". Essentially, it amounts to four images, all of the same size, which are placed together in order to form a perfect square. In order to achieve that goal, two of the images must be horizontal and two must be vertical. Placement in this manner always leaves a small square in the center, and that square can be used for a considerably smaller square image if desired. The aspect ratio of the four images doesn't much matter, as long as the aspect ratio of all four images is the same. The narrower the four images are, the larger the inner square will be. With four perfectly square images, there will be no inner square at all; and of course, that kind of arrangement will be perfectly symmetrical, unlike an arrangement featuring four rectangular images.

Now, it logically follows that a four square collage looks best when printed on a square sheet of paper or canvas. But that isn't always an option, since not all photo labs and printing companies offer square prints.

One option in such a situation is to have top, left and right borders which are all the same size, and then have a "weighted bottom" which has a considerable amount of open space.

Another option (which works well for posters) is to fill out the remaining part of the print with text. Or one can place simple abstract shapes or symbols in the remaining space.

Another option is to fill out the remaining space at the bottom of the presentation with additional images, so that the overall presentation looks perfectly balanced even though it's not perfectly symmetrical. That's what I've done with the arrangement shown at the top of this blog post. It's essentially a four square collage enhanced with a small fifth image in the middle of the four square collage, plus two large images of equal size at the bottom. The two images at the bottom are not perfect squares, but they're close enough to a 1:1 aspect ratio that they don't strongly favor the "portrait mode" the way they would if they had the 3:2 aspect ratios typical of 4x6 prints.

The above layout was designed to enable me to present multiple images on 20x30 paper or 24x36 paper, since those are standard print sizes often offered by digital photo labs.

20x30 is a good print size if one plans to mount and mat one's print on a standard piece of mat board measuring 30x40-inches, because it means that one's borders will be 5 inches on all four sides. (20 + 5 + 5 = 30. 30 + 5 + 5 =40.) Hence, the need to do any cutting on such a piece of mat board will be limited to the need to cut out the single window opening in the overmat, making it very easy for any halfway competent frame shop to do the job. (I tend to prefer the look of double mats myself, but they are not absolutely necessary.)

In general, any type of collage arrangement of images tends to look best when all of the images share some coherent theme. For example, one could create a collage arrangement featuring images of one's family, or a collage arrangement featuring images of sailboats, or a collage of images which all share a predominant color scheme.

In my case, I plan to create a number of collage arrangements featuring multiple photos of flowers and gardens.

One nice aspect of that approach is that I need not know the names of all of the plants in my photos in order to give the collages appropriate titles. If I present a photo which only features tulips, the title probably ought to contain some reference to the fact that they're tulips. That's no problem when it comes to tulips, since I recognize tulips when I see them, but I'm not such a botanical expert that I can easily identify all of the flowers in my photos. With a collage approach, that's unimportant. I can just give my pieces pieces names such as "Floral Collage #1", "Floral Collage #2", etc.

Another nice thing about collage arrangements is that they enable one to create really large framed pieces even though the resolution of the individual images may not be sufficient for pieces that large. For instance, a six megapixel image, when printed at 200 pixels per inch, would measure 10 inches by 15 inches, since it would typically have 2000 pixels by 3000 pixels (or something very close to those measurements). That's only 1/4 the number of pixels needed for a print measuring 20 inches by 30 inches. But it's more than large enough for one of the images in a digital collage which would be printed on a piece of paper or canvas measuring 20 inches by 30 inches.

If one assumes that the image at the top of this blog post represents a 20x30 print, then each of the four images in the "four square collage" at the top of the design would measure 7 inches by 10.5 inches, with a 1/2-inch space between the images, thereby creating a four square collage with overall dimensions of 18 inches by 18 inches. Add the 1-inch borders on each side, and you have a space measuring 20 inches by 20 inches.

In such a scenario, the two almost-square images at the bottom of the layout measure 9.5 inches high by 8.75 inches wide. I could have made both of those images perfectly square, but then there would have been a little bit of extra space between those two images and the four square collage above them, or else there would have been a little bit of extra space at the bottom of the print (or divided evenly between the top of the print and the bottom of the print).

One nice thing about collage arrangements such as the one shown here is that one can present far more images while still offering a fairly limited number of products. If each collage features 7 images, then one could present 49 images using just seven collage arrangements. (Of course, seven of those 49 images would consist of small images measuring just 3 inches by 3 inches.) This is particularly advantageous in terms of making the most of a limited amount of space on the walls of an art gallery, and equally advantageous in terms of making the most of a low-cost e-commerce solution which limits the amount of products one can sell at any given time. (For example, the Economy Edition of the Quick Shopping Cart e-commerce package offered by GoDaddy.com limits one to 20 products. But if each one of those products featured six or seven photos, one could present quite a few photos without exceeding that limitation.)

Conversely, one could create multiple distinctive products using a limited number of photos, simply by varying the placement of photos within each collage. In other words, there's no reason why any given image can only be used in one collage arrangement.

Changing the background is yet another way to create additional variations on a design. For instance, the background color could be changed from black in one design to white in another design which featured the exact same layout and the exact same images.

I find, when looking at some of my photos, that there are individual images which are only marginally interesting when viewed by themselves. In a collage arrangement featuring multiple images, the whole is sometimes greater than the sum of its parts. That's another reason why presenting one's images in the form of a digital collage can sometimes be a great idea.

Yellow Coneflowers --- A Cropped Digital Watercolor

If you have any experience with Photoshop, you know that it comes with a number of digital "plug-in" filters designed to transform digital photos and scanned images in a variety of ways. Some less expensive programs come with similar filters. Some of those filters work better than others, but it's possible to get good results even when using filters which leave a bit to be desired, if you know what you're doing.

In the above image, I cropped a high resolution photo of a large bed of yellow coneflowers, in order to focus on a smaller grouping of the flowers, so as to create the sense that the photo was a closeup. Then I decided to see what it would look like if I processed it with the Watercolor filter in Microsoft PictureIt!

I have learned through experience that photos often look pretty bad when processed with a digital Watercolor filter, unless one knows what the weaknesses of such filters are and how to overcome those weaknesses.

Specifically, the Watercolor filter seems to work better with images which are lighter and less contrasty than one would normally prefer. The effect of running the Watercolor filter is to darken the shadows and increase the contrast, in comparison with the way the regular unprocessed photo looked. If the photo looks just right to you before running it through the Watercolor filter, it probably won't look very good after being run through that filter, unless you like murky looking images with no real tonal range.

Sometimes it takes a bit of experimentation to get things just right. In some cases, it may be necessary to copy the image a couple of times and process each copy in a slightly different way, and then blend the multiple copies together in order to get the best of all worlds. That's a lot easier in Photoshop than it is in Microsoft PictureIt!, because Photoshop offers a huge assortment of blend modes, whereas PictureIt! only lets you adjust the amount of transparency on the top layer. Nevertheless, I was fairly pleased with this little "digital watercolor" which I created with the Microsoft program. The final image size of the original version had adequate resolution for a 4x6-inch print. Printed onto hot press watercolor paper with a high quality printer such as an Epson 7800 or 9800, I think it would look very nice.

Later, I might try using colored pencils atop such a print in order to add details, and then scanning the enhanced print so that I could make additional prints of that enhanced version.

NOTE: The effect which digital filters have on images can vary greatly, depending on the resolution of the unprocessed images. Sometimes, such filters work better with high resolution images. In other cases, strangely enough, they work better with low resolution images. The latter situation is a bit of a drag, since it prevents one from making really big prints, unless one has a special software program (such as Alien Skin BlowUp or Genuine Fractals) which is able to enlarge an image more effectively than the standard interpolation options in programs such as Photoshop.

I Give You Handala

Sometimes in art, the challenge is not to create whole images from scratch, but rather, to modify existing images in new and exciting ways. Or at least that's the case for collage artists and digital artists.

Here's an example. I recently downloaded a reference image from WetCanvas.com. For those of you who haven't yet discovered that website, it offers numerous cool features for artists, including access to a library of reference photos which you can download for free in order to create your own art based on those photos.

As one might expect, given the fact that the images are basically donated by anyone and everyone, some of the images are much better than others. But if you have a good eye, you can find some real treasures.

For example, I recently found a full color photo of a person's hand. The photo stood out on account of its clarity and good composition. My original intention, when downloading it, was to create a pen & ink drawing based on the photo. I still intend to do that, but tonight, I converted it to grayscale mode (known to you film afficionados as "black and white"). And then the fun really began. I started playing with mirrored copies and with features which enabled me to tint the image various colors.

The first image shown here is the grayscale version of the color photo I downloaded. The second version ... well, the second version is hardly recognizable in terms of being able to tell what source image it was originally based on. I call it "Handala" because it's a mandala constructed entirely of modified copies of a single photo of a man's hand.

And to think that I did it all with a simple program called Microsoft PictureIt! Imagine what I could have done with Photoshop.

I have additional variations on the hand photo which I will probably present in this blog at some point in the future.

Quick and Simple Modern Art

Last week I dropped in to take a look at the art being shown at a local gallery. On the walls, there were a number of large and colorful but rather simple abstract paintings somewhat reminiscent of the image displayed here. In fact, if anything, I'd say that the image shown here is more complex (and more interesting) than the images I saw at that gallery. The main difference is that the images I saw at the gallery were actually created with real paint (whereas this was created totally within the digital realm), and of course, the scale was much different. (This was originally created at a resolution of 650 x 650 pixels and then scaled down for display in this blog post.)

Oh, yeah, there was one other difference, too. Those paintings were being sold for thousands of dollars. This one hasn't earned me a dime, and probably won't make me any money at all.

Now, I must admit that it was rather fun to create the image shown here, using a combination of Microsoft Paint and Microsoft PictureIt! software. (I used Paint primarily, but the rough painterly effects were added in PictureIt!, since Paint, which is the free software which came with Windows XP, doesn't have any filters.)

But is it great art? If I were to transfer the image to a large piece of canvas and then paint it with real paint, would it become great art? Would the resulting painting be worth thousands of dollars? I must confess that I question that premise. The work involved in creating the image was minimal. I basically just drew in some geometric shapes which intersected with each other and I then filled in various areas with various colors, using the "paint bucket" tool in MS Paint. I experimented with different colors until I got a combination I liked. I had certain loose criteria in term of which colors were applied where, but a lot of the process involved just playing around with the colors.

If people want to pay thousands of dollars for such images, hey, I'll be happy to accommodate them, because I could certainly use the money. It wouldn't be hard at all to turn designs such as this into full-blown paintings.

Even so, I can't help thinking that the amount of money people pay for art ought to be somewhat proportional to the amount of work and skill that went into the creation of the art. There are a lot of gullible people out there who place value on particular works of art, not so much because there's any relationship between the work and skill it took to produce those works and the asking price for those works, but more on account of the fact that those particular works of art are highly praised by the elitists who dominate the world of modern fine art. I wouldn't mind that so much if it weren't for the fact that it sometimes means that people who are only moderately talented receive high praise and big bucks, while people who are far more talented get passed over like yesterday's garbage because their art isn't as trendy.

I also can't help but think that truly great art ought to have meaningful content, and that content ought to exhibit some measure of profundity or insight. When we look at the realistic paintings created by the "great masters" of art, those paintings educate and illuminate our minds. We are transported back in time, and we better understand the worlds in which those painters lived. We see the full gamut of human experiences, from tragedy to comedy. Such paintings tell rich stories.

Modern art pieces such as the one shown here do not do that. If one were to sum up the message of the above image, it would be, "I like bright colors." And that's pretty much it. It's not utterly worthless, since any short-term pleasure which comes from seeing bright and cheerful colors has the capacity to raise one's spirits. But profound it is not.

I think that the world of modern art often mirrors the world of philosophy. Most of our philosophers these days are bereft of any really profound ideas, having abandoned belief in biblical Christianity and the values which are derived from faith in God. Such people have nothing meaningful to say, so the art which they inspire says nothing meaningful.

When it comes to fine art, we need to learn to be more discerning, and to trust our own instincts rather than listening to the self-appointed arbiters of taste, whose alleged qualifications are mainly derived from the fact that they happen to have a lot of money and they happen to live in disproportionately influential places such as New York City.

If a piece of art looks as if it would be easy for a novice to create, it's probably because that is in fact the case. Such art should be priced accordingly. By all means, pay the painter enough to compensate for the time and materials which were necessary in order to create the painting, but reserve the really big bucks for those whose skills and vision deserve to be described as extraordinary.

---

Note: The following is a Flash-based slide show featuring the above image plus 3 others. The timing may be slightly off if you're playing it for the first time and you have a dialup connection, but playing it a second time should fix that problem.



Click here to see the original album.

Note to Self: I had to adjust the size of the player to 400 x 361 pixels in order to fit the entire player into this portion of the web page.