The Raster Canvas

Chapter 1 defined a raster image as a fixed grid of pixels. This chapter covers the technical properties of that grid directly: how it relates to physical size, what color mode it stores its values in, and how much precision each of those values actually holds.

Resolution: PPI vs. DPI

PPI (pixels per inch) describes an image file's own resolution — how densely its pixel data is packed. DPI (dots per inch) is technically a printing term, describing how densely a printer lays down physical ink dots. The two are often used interchangeably in casual conversation, but they aren't strictly identical: PPI describes the file itself, DPI describes a printer's own output. Web images generally only care about absolute pixel dimensions; print work genuinely cares about both.

Image Dimensions vs. Print Size

A raster image's total pixel dimensions (say, 3000×2000) combined with its PPI setting together determine its actual physical print size. The exact same pixel dimensions can print small-and-sharp at a high PPI value, or large-and-blurry at a low PPI value — the pixel data itself hasn't changed, only how many inches it's being spread across. This relationship is a genuinely common source of confusion worth understanding directly rather than treating PPI as a mysterious quality dial.

Color Modes: RGB vs. CMYK

RGB is additive color — red, green, and blue light combining, used for anything viewed on a screen. CMYK is subtractive color — cyan, magenta, yellow, and black ink absorbing light, used for professional printing. Converting between the two isn't always lossless: some RGB colors, especially vibrant, saturated ones, simply can't be reproduced exactly within CMYK's own smaller color gamut. This is a real, practical reason to work in the correct mode from the very start of a print project, rather than converting only at the very end and discovering colors have shifted.

Bit Depth

Bit depth describes how many distinct tonal values each color channel can represent — 8-bit per channel is the common default; 16-bit per channel offers dramatically more tonal precision. This matters specifically during extensive editing or color grading: 8-bit images can show visible banding (abrupt tonal steps instead of a smooth gradient) after aggressive adjustments, where 16-bit tolerates the same adjustments far better. Both Photoshop and GIMP support both bit depths today, though GIMP's own full 16-bit (and higher) support was a more recent addition, arriving in GIMP 2.10.

Hands-On Exercises