Publication-Quality Figures: DPI, TIFF, CMYK & Resolution Explained

Few things are as frustrating as a beautiful figure rejected at upload for the wrong resolution or file type. The terminology — DPI, raster, vector, TIFF, CMYK — is rarely explained clearly to researchers, so this guide demystifies all of it in plain English and gives you a simple checklist to export figures that pass every time.

What DPI actually means

DPI (dots per inch) describes how many pixels are packed into each printed inch. A higher DPI means finer detail in print. Crucially, DPI only has meaning alongside a physical size: an image is not "300 DPI" in the abstract — it is 300 DPI at a given width in inches. To find the pixels you need, multiply the printed width by the DPI. A figure printed 4 inches wide at 300 DPI needs 1200 pixels across. This is why you should decide the final print size first, then make sure you have enough pixels — you cannot add real detail by simply increasing the DPI number on a small image.

The resolution journals expect

Requirements vary by publisher, but the common standards are: 300 DPI for colour and greyscale photographic images, 600 DPI for combination figures that mix images and line art, and roughly 1000–1200 DPI for pure black-and-white line art, where crisp edges matter most. Covers are printed large and often demand 300 DPI at full cover size. Always confirm the exact figures in your journal's author guidelines, since individual titles differ.

Raster vs vector — the key distinction

A raster image is a grid of pixels (TIFF, PNG, JPG). It is necessary for photographs, microscopy, and 3D renders, but it loses quality when enlarged beyond its native resolution. A vector image (EPS, PDF, SVG) is built from mathematical paths, so it stays perfectly sharp at any size and is ideal for charts, diagrams, line art, and text. The practical rule: use vector for anything drawn or plotted, and high-resolution raster for anything photographic or rendered. Many submissions combine both.

Choosing a file format

TIFF is the workhorse for high-resolution raster figures because it is lossless and widely accepted. PNG is also lossless and fine for many figures and web use. Avoid JPG for final figures: its compression introduces artefacts around edges and text that look unprofessional and can't be undone. For vector artwork, EPS and PDF are the standard submission formats, with SVG common as a working format. When in doubt, the journal's guidelines state exactly which formats they accept.

RGB vs CMYK colour

RGB is the colour model of screens and is what most software exports by default; it offers a wide, vivid range. CMYK is the model of physical printing and has a smaller gamut, so bright RGB colours — intense blues, greens, and oranges — can look duller once converted. Many online-first journals now accept or prefer RGB, while some print workflows still request CMYK. If your journal wants CMYK, convert and preview the result before submitting so you aren't surprised by colour shifts, and adjust your palette if key colours fade.

A pre-submission checklist

Before you upload, run through five quick checks. First, is the figure at the journal's required DPI at its intended print size? Second, is it in an accepted file format (usually TIFF for images, EPS or PDF for vector)? Third, is the colour space what the journal asks for, RGB or CMYK? Fourth, is all text legible at final size, with embedded or outlined fonts so nothing reflows? Fifth, view it at 100% and at thumbnail size to confirm it stays sharp and readable. Two minutes here saves a frustrating rejection later.

Related reading: Best Colour Palettes for Scientific Figures and What is a TOC Graphic?