Rastering Explained: What Actually Happens to Your Image on the Way to Print

You hit "Print" and a photograph comes out the other end. Between those two events, your image undergoes a complete transformation — from continuous-tone pixels to discrete dots on paper. Here's every stage of that pipeline.

Stage 1: Your image file

Your photograph exists as a grid of pixels. Each pixel contains color values — typically three channels (RGB) with 8 bits per channel (256 levels per channel, ~16.7 million possible colors) or 16 bits per channel (~281 trillion possible colors). The image has a resolution measured in pixels per inch (PPI) at a given physical size. A 6000×4000 pixel image printed at 20×13.3 inches is 300 PPI. That same image printed at 40×26.6 inches is 150 PPI.

At this stage, the image is continuous-tone — every pixel can be any of millions of colors at any brightness level. Printing, however, cannot reproduce continuous tone directly. A printer can either deposit ink/toner at a specific location or not. The challenge is converting your continuous-tone image into a pattern of discrete marks that, viewed at normal distance, creates the illusion of continuous tone. That conversion is rasterization.

Stage 2: The application sends a page description

When you print from Photoshop, Lightroom, InDesign, or any application, the software doesn't send raw pixel data to the printer. It generates a page description — a structured file in a Page Description Language (PDL) like PostScript or PDF. This page description contains: the image data (either embedded or referenced), its position and size on the page, color space information, any vector elements (text, shapes, paths), transparency instructions, and color management directives (which ICC profiles to use, which rendering intent to apply).

In a desktop photo printing workflow, this page description is relatively simple: one image, placed on one page, with color management instructions. In a production environment, a single PDF might contain hundreds of pages with mixed vector and raster content, spot colors, overprints, transparency, and variable data — all of which the next stage must interpret correctly.

Stage 3: The RIP interprets the page

The Raster Image Processor (RIP) is the engine that converts the page description into the actual data stream the printer uses. As described by the industry literature, the RIP performs several operations in sequence:

Interpretation: The RIP parses the incoming PDF/PostScript file and builds an internal display list — a database of every graphical element that must appear on the output: images, text characters (with font, size, and color), vector fills and strokes, and transparency relationships. Each element's color is identified and mapped to the output color space.

Color conversion: The RIP applies ICC profile transformations to convert all colors from their source spaces (sRGB, Adobe RGB, CMYK with specific profiles, spot colors) into the printer's device color space. This is where rendering intent decisions are applied. In production, the RIP may also apply G7 calibration curves at this stage — modifying the CMYK values to maintain the target grayscale tonality and gray balance.

Rendering: The display list is rasterized — converted from vector/object descriptions into a grid of pixels at the printer's native resolution. A printer running at 1200 DPI creates a grid where each inch contains 1200 rows and 1200 columns of addressable dots. This grid is far denser than your original image's pixel grid: your 300 PPI image is mapped onto a 1200 DPI output grid, with each image pixel occupying approximately a 4×4 block of printer dots.

Stage 4: Screening — the key transformation

This is where the magic happens. Your continuous-tone raster data — where each pixel position has a tone value from 0% to 100% for each color channel — must be converted into a binary pattern: ink/toner ON or OFF at each dot position. The algorithm that performs this conversion is called screening.

AM screening (Amplitude Modulation), also called conventional or halftone screening, is the traditional method used since the 1880s. The output area is divided into a grid of cells. Each cell contains a cluster of addressable dots. The tone value at that location determines how many dots in the cell are turned on — a 50% tone turns on half the dots, creating a half-sized "dot" in that cell. Lighter tones produce smaller dots; darker tones produce larger dots. The spacing of these cells is measured in lines per inch (LPI). Typical commercial print runs at 150-175 LPI. Newspaper runs at 85-100 LPI. Fine art inkjet may run at 200+ LPI equivalents.

Each CMYK separation is screened at a different angle (typically C=15°, M=75°, Y=0°, K=45°) to minimize visible patterning (moiré) when the four separations are printed on top of each other. The overlapping pattern creates what's called a rosette — the characteristic dot pattern visible under magnification on any offset-printed or laser-printed photograph.

FM screening (Frequency Modulation), also called stochastic screening, takes a fundamentally different approach. Instead of varying dot size in a fixed grid, FM screening uses dots of a fixed (very small) size and varies their placement density. Darker tones have more dots packed closely together; lighter tones have fewer dots spread farther apart. Because there's no regular grid pattern, there's no moiré, no rosette, and no visible screening angle. FM screening produces output that looks closer to continuous tone, particularly in highlight areas and smooth gradients. Most desktop inkjet photo printers use a form of FM or hybrid screening.

Hybrid screening combines AM and FM approaches — using AM patterns in midtones where it excels and FM patterns in highlights and shadows where AM dots become too small to reproduce reliably or too large to show detail. Many modern production RIPs offer hybrid screening modes.

Stage 5: Output

The screened data — now a binary bitmap for each color separation (CMYK plus any additional channels like light cyan, light magenta, orange, green, or violet) — is streamed to the print engine. In an inkjet printer, the data controls the firing of individual nozzles in the printhead as the head traverses the media. In a laser/toner system, the data controls the laser writing the electrostatic charge pattern on the photoconductor drum.

The mechanical precision of this stage determines final output quality. Printhead alignment, media advance accuracy, drop placement consistency (in inkjet), and toner transfer/fusing uniformity (in laser) all affect whether the carefully calculated screening pattern is reproduced faithfully on paper.

Why resolution rules matter

The relationship between your image resolution (PPI), the screening method (LPI for AM), and the printer's device resolution (DPI) determines final quality. The standard rule for AM screening: required PPI = 2 × LPI at final output size. A 175 LPI print requires 350 PPI. Below that threshold, the RIP doesn't have enough image data to fill each halftone cell with the correct tone value, producing softness or visible pixelation. Above that threshold, extra data is discarded — no harm done, but no benefit either.

For FM screening and inkjet photo printing, the 2× rule doesn't directly apply because there's no fixed LPI grid. A general guideline of 300-400 PPI at output size provides sufficient data for desktop photo printers running at 1440-2880 DPI. Sending 720 PPI to a 1440 DPI inkjet printer provides no visible improvement over 360 PPI for photographic content viewed at normal distance.

Understanding this pipeline explains a question photographers often ask: "Why does my printer need to run at 2880 DPI if my image is only 300 PPI?" The answer is that the printer's DPI is its dot-placement resolution — the grid on which it creates screening patterns. Your image's PPI is the tonal information density. The printer needs many small dots to simulate the continuous tones your 300 PPI image describes. They're measuring different things.