Photo: © Remy Levine/Bigstock
By Remy Levine, Bigstock contributor
Thousands of pages could be dedicated to just the basics of processing your digital pictures for use, I’m going to use two or three at most and cover just the core aspects related to stock photography.
The first issue is appropriately the initial quality of the image. Just because you’re shooting digital doesn’t mean you shouldn’t try to achieve the best possible results in camera. Every little bit of digital manipulation potentially adds noise and degrades quality. Make sure there’s enough light, preferably enough to use the slowest ISO rating your camera offers. This would be the lowest number, generally 100 ISO.
Pay attention to the details, which is easier with a tripod since you can lock the camera in place and examine the view to see what is going on. Check the edges for things getting in the way. Check for odd shadows or highlights that confuse the image. Check that your subject is clear and obvious. Stock is all about subjects not just pretty pictures.
There’s an old saying in art (and engineering, just to be fair) that if it doesn’t contribute it shouldn’t be there. If you can, get rid of anything that isn’t adding to your image. Extra details just distract from your subject. Keep it clean and simple. Remember, a lot of these images will be used at fairly small physical sizes, so complex images just won’t be as clear as something simpler and more focused. Technically these aren’t post processing issues, but taking care along the way reduces the need for extra work later.
The next step is saving the initial files on your computer. Transfer the files from your camera/card onto your hard drive, making sure they came through correctly BEFORE wiping the camera or card. Countless images have been lost by a little impatience here. Since the chances are that your images are in JPEG format the next step is to get them into a lossless format. I use PSD, since Photoshop is my software of choice. Open each file (or create a batch process) and have each one saved as a PSD. I do them individually and take the time to name each one and cull the first set of rejects. If it’s all one series a batch process is probably simpler, and can be set to just label them with name and sequential number. If you don’t know how to do this there are plenty of guides to photoshop available that explain it better than I can here. The key here is a lossless compression. JPEG loses quality with each save, even at max quality. This can add up quickly. Limit your use of jpeg to the initial file and then the final save for uploading. If practical, don’t use it at all for the initial file, but some cameras don’t give that option or (like mine) actually look slightly worse when capturing images in a RAW or TIFF format. Maybe someone with more knowledge of the capture systems could explain why that happens, but a little side by side comparison with your own camera can make it clear which is better for you to use. Regardless of what format you choose, confirm that it is a lossless system. I suggest saving the initial file then going through and tossing any clear rejects. Don’t do any adjustments or changes, just throw out images that aren’t worth your time. This can include images that are out of focus, bad exposures, subject moved, something got in the way… After that make a backup. Stuff happens, and it’s really easy to accidentally save over a file during your processing and lose that original. I use a separate drive but even just duplicating to another directory covers you from anything but hardware failure.
Once you’ve got your files ready to actually work on there are a few things that should be done in a particular order. The first is level adjustment because if you find that an image is impossible to adjust to the desired point without it looking like junk then there’s no point wasting more time. Levels are fairly simple in concept, three sliders representing white, middle grey and black. Pull the white and black to the outer edges of data on their ends, this makes sure the image has an actual black and an actual white. You can go further, but doing so reduces details in the highlight or shadow areas and increases noise generally. The middle slider changes the balance of the image toward light or dark by adjusting where middle grey falls. You might find it easier to image this as either being neutral if left alone and centered between white and black or compressing the darks or lights if moved toward that direction. If you move the slider toward the white end you are saying that much more of the image is darker than middle grey. Moving the slider toward the black side biases things lighter. Note that this is not a brightness adjustment, you still have a pure black and a pure white based on where you put the other two points, it just controls the ratio inside those points. Be careful, moving ANY of these points very far creates noise and it adds up fast. If you regularly find your images need significant level’s adjustment you should very carefully look at your exposures and lighting to figure out where the problem is. Check out the article on exposure for some ideas and check out books on exposure and lighting. The principle hasn’t changed significantly since photography started so don’t worry about finding digital specific material.
For those technical folks who want hard numbers, generally you can get a ten stop range. With digital, if you go beyond that you simply lose the extra, it’s just not there. Within reason it’s not a problem except on very high contrast days where you have intense light and shadows. The other direction, not having a very wide range, is easier to deal with but must be handled delicately. On a histogram (the chart showing in the level’s window or on the lcd of some cameras with that option) an image with a narrow range will not have information all the way to the edge of the histogram. Preferably the data is central and you would just adjust the black and white sliders to where the data begins for each of them. If it is not central you still do that, but quality likely suffers more for whichever end had to be moved the most. An image like this will look flat or dull until adjusted, it lacks contrast. Once the levels are adjusted it should look normal.
I’ve always had a pet peeve about this issue. There are people who say that every image should cover the whole range and look roughly like a mountain centered in the middle and tapering to the sides. I say BULL. That is just fine if every picture was a normal scene that averaged to neutral grey, but our world is not like that and neither are our images. Pictures can be mostly light or dark, even have huge sections of pure white or black. They can have inverted mountains on the histogram or look like a triangle. The key is that for a “normal” range you need to have both black and white, so you put the appropriate sliders to where data begins on their side and then tweak the middle bar as needed. For Over White images I often crank the white slider further in, removing minor variations in the background, it’s not a hard and fast rule, fiddle around and learn what happens. The big thing to remember is that too big an adjustment will create lots of noise.
Once levels are adjusted you can tweak the color. There are a number of ways to do this, starting with the obvious option of “Color Balance” and ending with the “ Hue / Saturation” control. I use both and pick based on the image. Work carefully and whenever possible use the numeric values in the info window to judge color balance rather than going by eye. Less is more here, and notice that it will let you adjust the highlights, mid tones and shadows separately to isolate problems. Often only one of these areas needs adjustment or they each need a different adjustment. Try to keep the total changes to a minimum and for the real purists, fiddle around to discover the proper settings then undo all the incremental changes and input those changes in one step. Technically this results in higher quality since each successive change generated slightly more noise. In practice it probably won’t matter unless you were making dozens of adjustments but some folks want perfection. I wait to adjust the saturation until later.
After getting the color and levels correct you can go through and clean up the image. Remove any dust spots, logos, blemishes… Again, less is more, try to make your changes invisible. I use the clone stamp, bandaid and the selection tools to do this work, depending on what I’m fixing. If you have an Over White background, now is the time to make sure it’s clean. The exact shade is less important than it being even and clean. Obviously an Isolated image must have a completely pure background surrounding the subject. I make all these changes before adjusting saturation and make as small a saturation change as possible since it is a prime culprit in noise generation.
Filters are nifty and cool, and generally not a good idea for stock images. Stock is the building block, let the designer build the final image with whatever filters and effects they want. Sharpening is generally a bad idea, it adds noise and noise is one of the top reasons for stock photo rejections. A focused image is fine even if not absolutely crisp, a noisy image is always noisy. Some use of the sharpening tools are ok but be very careful. The same goes for noise reduction, oddly enough. Median and Despeckle are handy but not as good as some other options. My choice is Neat Image, a third party software package that does an almost magical job removing noise without harming details. There is a free version and I highly recommend it. If your volume is high enough it’s worth paying for the version that allows batching and offers a Photoshop plug-in. There are other packages, I simply haven’t tried them myself since I’m satisfied with Neat Image. A quick check of the various forums on photography sites will yield numerous opinions and options.
The final step is the noise reduction, in my case Neat Image. Because it won’t operate on PSD files I make a “final” save as a PSD into my archive directory then another save as a TIFF into a temporary directory. When I’m done with all the images I run each TIFF file through Neat Image and it outputs a high quality JPEG which I put in my stock photo jpeg directory for uploading and any other use.
