Digitally Processing Images

Once you have your raw scanned image, you are going to need to edit it so it looks nice and crispy. Here are the steps you should take, in this order (in Photoshop):

Crop– Crop out the borders of the negative; unless you’re into that kind of thing

Dust and scratches– You can use the ‘spot healing brush’ tool to remove dust and scratches from a wide area, like a sky or water. It just takes one click. For more complicated, detailed areas, you will need to use the ‘clone stamp’ tool. Basically you will pick out a place you want to clone, usually a small area next to the dust or scratch (by using the option key), and then stamping that sample over the blemish.

At this point you should save the file (as a TIF) because it is clean and ready to be edited. This will be your original copy, and you can save other edited versions without messing with this file.

Mode– Your scan will probably be in 16-bit color and in the Adobe1998 color space. This is great for printing, but when you are saving an image for web, you will need to change these. The most common image compression used is JPEG, and it doesn’t support 16-bit color. Change the mode to 8-bit (Image > Mode). The color space, Adobe1998, has a tad bit wider gamut than sRGB, but is not designed to be viewed on a computer, only in prints. Change the color space to sRGB IEC61966-2.1 (Edit > Convert to Profile)

Levels– Levels means the balance of shadows, highlights and midtones in an image. Basically it is the contrast. (Image > Adjustments > Levels) Grab the white arrow and drag it towards the middle. You will see your image getting brighter. If you want to be really precise, hold down ‘option’ and drag the arrow. You can see exactly where the first pure white pixel appears. Do the same for the shadows, which is the left arrow. Then use the middle arrow to control the midtones, the overall brightness to the image. I prefer a darker image and sometimes drag the midtones down to .75 to achieve the desired effect.

Curves– This is the balance of colors in an image. (Image > Adjustments > Curves) For the most precise, almost automatic color correction, use the eyedroppers at the bottom to assign the white and black points. If you are having trouble finding those points, click the box ‘show clipping’ and you can see the brightest and darkest pixels. If you want more manual control over the colors, use the drop menu at top and adjust each color itself. First is red; the opposite of red is cyan, so if you were to grab the middle of that bar in the graph and drag it down (to the lower right) your image will take on a cyan tint. This works inversely if you drag it up for more red. Do the same adjustments with green and blue until you are happy with the results.

Size– Resize your image depending on where you will be using it. (Image > Image Size) Change your resolution to 100 (the highest resolution a computer monitor can see). This will shrink the image a bit, but you will probably still need to downsize it further. A good rule of thumb (for the web) is to never have it smaller than 600px (on the longer edge) but no larger than 1000px. The ‘resample’ method should be ‘bicubic’.

Sharpen– There are a couple different ways to sharpen an image, my preferred method is using the unsharp mask. Read up here about sharpening.

Save– I don’t use the ‘save for web’ option, even though I am saving for web; it usually degrades the quality of the image and sometimes gives it a weird color shift. I simply go to File > Save As… and save the image as a JPEG. Check the box for ‘Embed Color Profile’ (which should be sRGB). Click save and a popup will ask you the quality you’d like to use. I always use 12, the highest. This just makes sense if you want the image to look good. Select ‘Baseline’ under Format Options.