Last modified: Saturday, January 6, 2001 6:20 PM

A lot of the downloadable graphics in the GFX section aren't exactly ready to go, because in many cases the final print size is guesstimated. If you know the size of the real thing, you can scale those dimensions down using my handy dandy Scale Conversion Calculator. Once you've got the scaled dimensions, it's pretty easy to adjust the graphic to print at the correct 1/6th scale size. For folks who aren't familiar with this stuff, I'll attempt to explain how you can manipulate them to get them to fit your needs. It's the same thing I tried to do in my first article, but this one is more fleshed out. My explanation and terminology isn't from the world of professionals--but maybe that's not such a bad thing?

I've tagged most of the graphics in the GFX section to say something like "at 300 dpi, prints a 1 cm wide sign". This means that the image itself contains information to tell the printer to print the image at a resolution of 300 dots per inch. At that resolution, based on the amount of information (pixels/dots) in the picture, the printed image (but not necessarily your monitor's image) will be 1 centimeter wide. If you double the resolution to 600 dpi, the printed image will be half that size, but all the original dots will still be there (if your printer can handle it).

The resolution information only describes the way the file should be handled by the printer, and doesn't change the file's size or the amount of picture information (detail) it contains. You can edit the resolution information in a program like "Photoshop" (under "Image Size"). I emphasize that this (if set up properly) does not affect the amount of information in a picture, or its pixel size, or the way it appears on your computer monitor. This is not the same thing as resizing the information in a picture, or resampling it (which you can also do in Photoshop's Image Size dialog box). In those cases, you reduce the pixel dimensions and actually lose information. That's not a good thing unless you intentionally want to make a file's size smaller (for faster downloading, for example). Therefore, to change the resolution in Photoshop, you should turn off the "Resample Image" check box before changing the resolution. This locks the "Pixel Dimensions" but lets you see the print output size in whatever units of measure you prefer.

You don't have to do this in a paint program either. If the graphic doesn't need to be retouched in Photoshop, you can skip it entirely. Printing is much more convenient in a drawing program like "Illustrator" or "Corel Draw". You can import or paste in a Photoshop bit-mapped image (like a JPEG format graphic) and resize it onscreen to arrange it in a layout. If you've preset the resolution in Photoshop, they'll plop right in at the correct size. Otherwise, you can just resize the image by dragging the bounding box's "handles". It looks like the image is being resized, but actually only the resolution information (and the screen display) is really being changed. This is very convenient and intuitive, since you can match the displayed size with the onscreen ruler, which corresponds to the printed output size. It eliminates the need to fiddle with the resolution setting numerically. Then you can duplicate the resized image a bunch of times and print a sheet's worth of copies in one printing pass (more or less).

Bear in mind that 20 pasted-in copies of a 100kb file may look small after they've been resolution resized for layout, but each one is still 100kb-- Your computer or printer make choke if you go hog wild with this. If your printer cooperates, you can get around this by printing a row or two, moving the images lower on the page and running the same sheet through again. This is especially useful for doing decals if you don't want to waste a bunch of the expensive paper.


PICTURE INFORMATION: One of the most important concepts is that a digital picture contains a certain amount of information, which is unrelated to the size you see it on screen or printed. If it contains 100 dots of information, the individual dots can be very large or very small, but the fact that there are 100 dots of information doesn't change.

The picture's dimensions, measured in pixels or picture elements (dots) is an indicator of the amount of information contained in a picture. For example, 640 x 480 (horizontal x vertical) pixels is a common setup for a computer screen. Pixels are the count of dots, but it doesn't matter whether they're big or little dots. Unlike inches or centimeters, it's not a measurement which corresponds to an absolute real-world thing. Therefore, a 12" monitor set at 640 x 480 displays the same amount of information as a 19" monitor-- the dots are just smaller on the 12" monitor.

Another good indicator of the quantity of information is the file size, measured in kilobytes (KB). A big file takes longer to download because it contains more information.

Generally, it is a good idea to have as much information as possible because the picture will print with more detail-- logically, because it has more information to work with. However, there's a point of diminishing returns as you approach the limits of a printer's abilities, or your own ability to tell the difference. In other words, it's possible to have a file that takes an hour to download/print that doesn't look noticibly better to you than one that takes 5 minutes to download/print. Similarly, it doesn't make sense to have gazillions of colors in a picture when the printer can't print them.

Color adds another dimension to this concept-- each dot can have color or transparency information too, and the amount of information varies depending on the color mode of the image: Simple black and white obviously takes the least amount of information. But we'll ignore this aspect since most modern computers are set up to display more than black and white, or even 256 colors. Anything above that should produce respectable results, at least good enough for most consumer grade color printers. Bear in mind that the printer is the weak link in the chain: It may not be able to accurately print every shade you're capable of producing on your computer's screen.

Ideally you should use JPEG images, and not GIF images. GIFs have a very limited palette of 256 colors (however they're good for other things like transparency and animation).

JPEG COMPRESSION: Some of the posters I've scanned look really big on your computer screen. Because I set their resolution to 300 dpi, they should print much smaller, close to what I guessed to be a good "Joe size". These are also some fairly hefty files-- they're around 150kb each and will take longer than most graphics at this website to download.

I chose 300 dpi resolution to make the files reasonably sized. In order for them to print with the same dimensions at 600 dpi, the files would be much bigger (kilobyte-wise). In fact, even at 300 dpi, they're pretty big, but they were originally much bigger before I saved them with medium quality JPEG compression.

JPEG compression is one of those compromises between quality and practicality. Basically, some of the less important detail information is generalized and discarded. For example, a blue background might have some dots with a slight color variation which isn't crucial to the image. Instead of storing all the information of that background verbatim, a generalized version is saved in a more efficiently-organized file. That's why it's called a "lossy" format-- you lose information.

For example, say you've got this string of 9 characters: "aaaaaaaba". You could encode it more efficiently like this: "7a1b1a" (seven "a"s, one "b" and one "a") using only 6 characters. With lossy compression, you could discard the "b" and encode it as "9a"-- only 2 characters. It's not the same, but it's pretty close and the data size is much smaller (it's offloaded to the decoding program). Obviously, this isn't how you'd want your bank account information stored, but for photos it's not too bad. I'm not saying that this is specifically how JPEG compression works, but reducing the information storage requirements of repeating patterns is the general idea behind compression.

Practically speaking, the loss of medium quality compression doesn't bother me very much because it seems to keep the detail intact-- and makes much smaller files. But you shouldn't go too far with this: At higher levels of compression, the effect is really noticible and images look awful.