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.