Skip to content

Working With Image Formats

by Richard on December 29, 2016

As Independent Authors, we often have to work directly with our marketing and communications. Much of that involves using digital images, such as book covers. in the interest of helping to clarify what can be a bit confusing to someone not schooled in digital graphics, here are a few notes on image file formats and what they mean for your book cover and online marketing graphics.

Digital images, or bitmaps as they are known, have had a long interesting history. The internet has been a huge influence on how they are formatted. Back in the day of Pac-Man and Compuserve, bitmapped images were severely limited in color fidelity and even resolution by the low bandwidth of the internet as it was then. Most users had a dial-up connection which at best, could handle 56K bits per second. Most of the time it ran at about half of that. Compuserve invented the GIF format as a break-out way to transmit color images in an 8-bit per pixel color model that revealed color into 256 different steps of hue. It was also a faster way of sending an image through a telephone line.

Needless to say, it was pretty rudimentary, but then so were the graphics processors in those days.  My first PC was a 286 IBM clone that could handle some color on a black-screen with scrolling colored type. Graphics were jagged and all flat color sections. No photographic images except greyscale at that time. Full color came later, along with the JPEG format, in 1992 after a long conference of the Joint Photographic Experts Group. It was designed to make color photographic material possible to transmit. Although it used almost full-spectrum color modeling, in order to keep the bandwidth needed to transmit the images, they were compressed by algorithm and slightly blurred. Some detail is generally lost in the compression as is a degree of nuanced color fidelity. Much later, jpeg formats were opened up to include less compression and greater color fidelity.

Along with later systems, including Apple’s Macintosh, came full spectrum images that were not compressed at all. Lossless images. These were called TIFF images and although they took up a great deal more room on a hard drive, were much closer to true color and sharp resolution. It can be confusing trying to keep all the formats straight, but to make it simple, think of TIFF (Also PNG and a few other high-rez formats) as the apex of both color fidelity and sharpness, then comes JPEG in downward-varying degrees of compression and finally, GIF images at the bottom.

Some uses dictate the format preferred. For ebook covers and online catalogs, jpegs are usually chosen. For do-it-yourselfers, I recommend that if your image manipulation software allows, create your jpeg images with no compression at the highest quality settings. They’ll be slightly larger files than the compressed versions, but they will look much better on the screen. I also recommend 96dpi resolution for screen images over the standard 72 dpi. The higher resolution adds a lot of depth and better type contrast. In the case of some service agencies, such as Amazon KDP, you can actually upload your covers in uncompressed TIFF format. This is my own preferred format for online images if they are accepted. From the image they receive, they will re-sample and model your supplied image downwards for size and resolution, but at least they’ll be starting with the best, most faithful version of your image.

The color model (based on the primary components of spectrum that dictate the appearance of the spectrum chosen) most frequently used for both JPEGs and TIFFs is 24-bit RGB (designed for transmitted color, such as from a monitor) and CMYK (designed for reflected color as from a printed page) which should be kept to their respective applications. I’ve been known to use a 32-bit CMYK image, which is a much larger file that a 24-bit JPEG to achieve the contrast and color I want, then converting it downwards to a JPEG and adjusting for brightness to bring it close to the original appearance. It’s a step you really don’t need to take. I just like to tweak things in all kinds of ways. But using the right amount of compressions for the application’s best color model – RGB or CMYK – Monitor or print – you’ll get the best results. The most recent iteration of the JPEG model is the JPG2000 (.jp2 file suffix) which accepts any bit depth for improved scaleability, but outside of professional graphics circles is still not readily accepted. Another format is the PNG format which works best if an image is dominated by a single color or flat colors such as in charts, logos, etc.

One last tip, if you want to use Unsharp Masking (or another name for sharpening, etc.) to make your titling pop a bit more, use it sparingly (lesser percentages) to reduce haloing and jaggies, and then only after all other procedures, such as contrast adjustment, color tweaking and resampling have been accomplished. Jus’ sayin’.

I hope that any Apple users reading this will broaden this post by adding any comments unique to their platform. In addition, feel free to ask questions. If I can answer them I will, or refer them to someone who can.

# # # #

No comments yet

Leave a Reply

Note: XHTML is allowed. Your email address will never be published.

Subscribe to this comment feed via RSS

Turn on pictures to see the captcha *