Went out shooting with the multimedia class today, which is now in photography mode for the first 3-4 weeks, before moving on to blogging/podcast production. We were out in Myeongdong taking pictures, with the assignment being that of taking pictures of people we didn't know. It's actually easier than one would think – and I always say this – to take such pictures, as most people are so wrapped up in what they're doing that they don't notice, anyway. But that's not the point I'm trying to make right now.
I'm want to talk about how pictures are made these days, as well as some of the decisions that go into making even the simplest of shots, since they usually involve many steps before even getting to the page, even if it looks like the simplest of snaps. The differences are mostly not monumental, but they are important.
Below, I have the original shot from my new digital Canon EOS 350, which makes some pretty spectacular images that satisfactorily surpass 35mm film. I took this shot in JPEG mode, since I was practicing and not wanting to take up a lot of space on the CF card. Usually, when I'm out shooting on the streets, I am shooting in RAW mode, which preserves all of the information in the shot, as opposed to JPEG, which compresses the information and eliminates parts of the image humans don't really need.
It's actually similar to an MP3 versus a CD in that the song you hear on an uncompressed CD track is actually reproduced in a format called AIFF, versus the compressed MP3, which eliminates parts of the aural spectrum that humans tend not to hear or notice, while emphasizing the parts that are the most important to conveying the music. Most of the time, it's "good enough", and to the untrained ear, there's no real difference, especially if you don't turn the compression up too high.
The JPEG images that most digital cameras save are like the visual equivalent of an MP3. What's more, for the resolution you end up using to display things on a computer, you don't need that much resolution at all. Most of the resolution power over 1megapixel is pretty much overkill to anyone who generally sends pictures by email; megapixels only matter for printing, and since most of us don't print over 5x7 or even 8x10, anything over 2 megapixels is really overkill and just marketing hype. What they should be concentrating on is making better lenses and better forms of compression at the resolutions offered, but they mostly just keep rolling out higher and higher resolutions that most people don't need. It's like Mhz and chip speeds – most people just use the web and email, so the speed of the chip doesn't really matter. But who's gonna tell consumers that and just make a simple, cheap, upgradeable box that gives you what you need, rather than sell you something that you'll chuck out $2000 for every three years?
Same goes for digital cameras – we mostly buy them because of the new features bundled into the cameras with extra megapixels, rather than the megapixels themselves. Wouldn't it be cool if they still sold a 2-megapixel, entry-level camera that had all the bells and whistles, but for cheaper, rather than be forced to buy the 4-megapixel one at the same price as the 2-megapixel one two years before? It's just an excuse to keep the price points the same.
Anyway, I shot this in JPEG mode; the photo below is as I shot it, except I had to reduce it to 500 pixels across in order for it to fit on this page properly. Otherwise, it's not cropped, processed (except for the automatic processing and compression that was already done in the camera, actually), or altered in any way. So this is what I basically got straight from the camera:
Not, the image below represents what I did in Photoshop, and why it's a powerful, necessary program. Here, I adjusted the Levels, which in normal English means I set the white, black, and midpoint values for the information in the frame. Actually, to save time, since there is in theory a "right" point to come to, I just had PS set the levels automatically; when my visual common sense indicates that PS made a mistake, I just do it manually.
Anyway, I set the Levels to make the image accurate, then boosted the overall Saturation (richness and depth of color) a tad, then gave the contrast a little nudge. Would-be "purists" these days tend to yell, "But you're cheating, since it's on a computer!" But people who actually know Photoshop as well as the general ways pictures used to be developed know that these were things that were possible to do in the darkroom, which is the standard rule-of-thumb for defining the limits of what you do in PS.
This means dodging (exposing a section of the photographic paper for less, so it comes out lighter or less contrasty), burning (exposing a section for longer so it comes out darker and with more color saturation), cropping (exposing a select part of the frame that you want on the photographic paper, or actually slicing it off), as well as other myriad overall exposure tricks that master printers knew I and I don't profess to have know.
Of course, the advantage is that you can do those things more exactly on the computer, as well as some of the ones that were technically harder to do. But in the end, most photographers boost a bit here and there, adjust this and that, but they did in the old days as well. Now, it's just power that's available to the rest of us and a skill/artform in itself, just as old-fashioned developing and printing was.
And what most people don't realize is that digital was a part of the process for much more than people thought, anyway. I shot on film since forever until this year, but in order to get it onto the computer, I had to scan it. Now, it may sound all romantic that I "shot on film" and people think that's all cool and whatnot, but actually, it was all ending up the same place. When I had my beautiful medium-format slide negatives scanned, you can't see why slide film was so superior to negative film unless you see it blown up and projected onto a white screen; in the computer, everything can be fixed up once it's in digital. I'm not saying this to down digital, but rather to make the point that film was marked by limitations that have become meaningless in the age of digital. And once you have to spend money to buy 36 shots, then develop them, and then have them scanned – versus just going straight there with a digital camera – film doesn't seem so "romantic." For me, it was just irritating and expensive after a certain point, because – hey, it's a digital world.
What you might not notice in the same shot below is that I used a Brush tool to select the girls and again boost the color, brightness, and contrast on them so they pop out a bit more from the frame. You don't notice it, but compare the two pictures again, the one below to the original above. See the difference?
If you have a sharp eye, you might also notice that I cropped the picture a bit from the original. My little rule is that I don't like to crop at all, but when I do, I maintain the same aspect ratio as the original frame, in order to preserve the feeling of presenting a shot that I would have actually taken in camera, but didn't. So I use Constrain Proportions in the Select tool in PS to keep the horizontal length "1", but the vertical height ".667." Then, when I adjust in the crop tool, not matter what the size of the box, it still maintains the same aspect ratio.
I did the same for the picture here, which is a vertical version crop with the same aspect ration constraints; I force myself to choose a crop that would have been defined by me having just taken the picture this way originally. It resulted in a nice, tight crop in the vertical shot that really focuses in on the girls and the first girl's great expression and pose.
You might notice that the girls look a little "sharper" than in the original. That's because I used the Unsharp Mask function to tighten up lines and patterns. I purposely do it a bit on the heavier side when I print on the web, since that degrades the image by making it look sharper. That may sound like a contradiction, but combined with the fact that I shrink all my pics down to 500 pixels across, there's not enough resolution and clarity to make a good physical print any larger than a couple of inches across. I could upload a high-resolution version of the image, but then I'd be giving it away for effectively free, allowing people to make huge prints from my work – it'd be a like putting my negatives up for anyone to use, if we were talking in "old days" talk. Pictures look great on a screen – because the pixels are represented exactly as-is on a computer screen – which is why you don't need much resolution to do anything short of printing.
The irritating thing about JPEGs and saving them at a reasonable resolution on the web is that when you compress them to a level that will keep file sizes down, you start losing a lot of things from the original PS version. It's a shame, too, since certain images look so spectacular in PS, but when they're exported, compressed, and crunched into web-and-server-friendly images, they can end up looking washed out and boring. Such was a problem above, since the girls looked so colorful in the frame. The JPEG version is a shadow of the original.
But this is why people still print and go to art shows, right?