Laughing Knees

The clarity of my photos when I originally take them is probably not much better than anybody else’s. It’s just the nature of digital photography that post-production is probably just as important as when you actually take the photos. I use PhotoShop CS to work on my photos and try to bring out what I hope will be their best in their nature. I find that sometimes they work better in black and white, or else with colors strongly desaturated. Almost always sharpening is necessary, though I am still learning a lot about this. I’ve taken the liberty of using my method of retouching my photos to rework one of of the Alps (I hope you don’t mind… I just want to show you just how much you can punch out your own photos by retouching them):

Before retouching:

After retouching:

Pretty dramatic, no?

These instructions are for PhotoShop. I think you can use this method with whatever software is out there, but the controls may be a little different. Many different people have many different ways of working on their sharpening. Here’s what I do:

1) Open the original JPEG photo(image file type… for photos it’s better not to use GIF, unless there is a lot of uniform color).

2) Go to the menu bar and choose “File”/”Save As”. Go down to where it says “Format” and choose the software’s proprietary file type. In PhotoShop it is “PSD” or “PhotoShop”. The reason for saving the file in a different format is because JPEG is a “lossy” format in that with each subsequent save it loses information, until the file eventually degrades very badly. Use JPEG for image files that won’t be saved again, and for uploading files with small sizes onto the Web. Use PSD for reworking images, and TIFF or EPS for archiving your original photos in a more secure format. TIFF and EPS have huge file sizes though, so they are better left just for archiving.

3) Save sharpening till last. Do all the sizing, color adjustments and such beforehand.

4) After saving to PSD determine whether you want to change the image size or crop the image. For chainging image size go to “Menu Bar”/”Image”/”Image Size” and put in your coordinates. It’s better not to resize the image larger than its original size, because the larger size will cause the image to lose pixels, thus leaving an image with less resolution. For cropping, go to the Tool Bar on the screen, choose the “Crop” Tool and crop according to what you need.

5) From here I always start with “Menu Bar”/”Image”/”Adjustments”/”Levels”. Sometimes I just use “Auto Levels” instead, because it often does a good job with balancing natural colors, but at other times it changes everything so badly that it is better not to use it then. In “Levels” a histogram will appear, with a mountain range-like diagram. Beneath the diagram are three triangular adjusters. Move the leftmost triangle inward to the right until it sits at the leftmost point of the “mountain range”. Move the rightmost triangle inward to the left until it sits at the rightmost point of the “mountain range. This will adjust the brightness and contrast of the image into a more natural range. The center triangle is used to refine the image. Often I adjust this because my original images out of the camera are often too dark. You don’t need to adjust the histogram if the “mountain range” sits neatly within the box.

6) From here it is a matter of taste. I often go to “Menu Bar”/”Image”/”Adjustments”/”Hue/Saturation” to work on the colors and saturation. Very often I find that what I get out of the camera contains too much of one color and not enough of another. My camera tends to saturate reds too strongly, so I tone them down. Here it is that I often will pull out all the color and leave the image with a black and white tone, or a monochrome tone. I also desaturate all the colors but one in some images, to make that one color stand out.

7) When all the image adjustment is done and I am satisfied with the overall look, I finally go to the sharpening. Sharpening can be quite harsh on the pixels so care needs to be taken when doing it. That is why it is better to wait until the end to do it.

8) Go to the “Channels” palette on the side (if it is not open, got to “Menu Bar”/”Window”/”Channels”). This is usually sitting side-by-side with the “Layers” and “Paths” palettes. Select the “Green” channel (other colors are okay, too, but green tends to be the most neutral). Right click this and select “Duplicate Channel”. A new window will open named “Green Copy”. Click OK. A new channel, “Green Copy”, will appear in the Channel palette. Select this channel. Your image should should turn black and white now.

9) Go to “Menu Bar”/”Filter”/”Stylize”/”Find Edges”. Your image will now appear as a heavily stylized group of black edge lines.

10) Go to “Menu Bar”/”Filter” /”Blur”/”Gaussian Blur”. Set the value to “2.8”. This can be changed, of course, but this value has always worked well for me. You will get a blurred B/W image. Click “OK”. This effect is used to later create a softer overlayer that allows sharpening with a more feathered edge.

11) Go to “Menu Bar”/”Image”/”Adjustments”/”Invert”. Your image will now appear negative. This is done to make it easier to view the levels contrast hereafter.

12) Go to “Menu Bar”/”Image”/”Adjustments”/”Levels”. Here you will get the histogram with the “mountain range” again. Adjust the left and right triangles (not the middle one) so that each is more or less a centimeter (1/2 inch) inwards from the histogram “mountain range’s” tips. You want to make the blacks in the image quite dark and the whites almost shimmering white. Just don’t overdo it or there will be strong halos in the image later. Click OK.

13) Go back to the “Channels”/ “Layers” palette (if the “Layers” palette is not open go to “Menu Bar”/”Window”/”Layers” to open it). Select the “Layers” palette, click on your image. The original color image will re-appear.

14) Go to “Menu Bar”/”Select”/”Load Slection…”. Your image will now have some scintillating selection lines. These will determine which part of your image will be selectively sharpened. On your keyboard type (Mac) Command-H (Windows? something similar I guess) to hide the selection lines.

15) Go to “Menu Bar”/”Filter”/”Sharpen”/”Unsharp Mask”. A new window will open with sharpening adjustments. You can do it two ways from here. Before I discussed this with Pohanginapete I always adjusted the controls this way:
—• Amount: 145~280 %
—• Radius: 0.4 pixels
—• Threshold: 0 levels
—• Click OK.

This worked reasonably well, but I often found that the edges in the images, especially when there was high contrast, came out quite harsh. Pohanginapete suggested these adjustments:
—• Amount: 250 %
—• Radius: 0.2 pixels
—• Threshold: 0 levels
—• Click OK
—• Repeat.

With this method I sometimes had to adjust the Amount the second time around to something lower, like 145 % or so. Don’t adjust the Radius too high or really harsh artifacts appear in the image.

16) On your keyboard type “Command-H” again to bring the selection lines back, then “Command-D” to deselect the image.

17) Go to the “Layers” palette again, select your image layer (should be called “Background”), right click the image name text, and select “Duplicate Layers”. Click “OK”. A new layer called “Background copy” will appear in the Layers palette. In the dropdown menu at the top of the palette select “Soft Light”. Your image will take on a strong contrasty flavor.

18) Go to “Menu Bar”/”Filter”/”Other”/”High Pass…” A new window will appear with a pixel levels control. Here you can adjust the level of details that will either be hidden or allowed through. Using this filter lets you further sharpen your image, but with a softer, more even touch. What it is basically doing is either allowing or preventing the transparency of the earlier sharpening adjustments to filter through to the image. I usually am very sparing with the numbers here, usually between 0.3 to 0.9, though sometimes when there is a lot of contrast and I’m trying to soften the halos on the edge of things, I will go up to 1.5 pixels. You can check the “Preview” box to see how your image looks as you adjust the settings. Click “OK”.

19) Go to “Menu Bar”/”Layer”/”Flatten Image” to merge your two layers in the Layers palette. You cannot save your image unless you do this.

20) Go to “Menu Bar”/”File”/”Save”. Your image is done.

21) To make a JPEG file for uploading to the internet, I first go to “Menu Bar”/”Image”/”Adjustments”/”Brightness/ Contrast” and adjust the Brightness to ”-12” (essentially temporarily darkening the image) so that the later JPEG image does not get washed out on the computer screen. Don’t save the image with this setting, or your original image will be too dark.

22) Go to “Menu Bar”/”File”/”Save For Web…”. Make sure the format is JPEG, quality “very high” (or “high” if the file size is still too big), and check “optimized”. Click “Save”

Your new image is all done. It sounds like a lot of work when you read it here, but once you’ve gotten the hang of it it acutally quite fast. The results are usually very nice.

Trackback URI | Comments RSS