Reference
Image glossary
24 terms you meet while working with images — in a tool's options, in a print specification, or in an error message.
Every entry says what the thing is, where it matters, and where it stops mattering. Where a term has more than one common name, all of them are listed — “colour profile” and “ICC” are the same page.
- Alpha channelalso: alpha, transparency channel
- An alpha channel is a per-pixel transparency map stored alongside the colour channels, recording how much of each pixel is opaque.
- Aspect ratioalso: ratio, image proportions
- Aspect ratio is the proportional relationship between an image's width and height, expressed as two numbers such as 16:9, and it is what determines whether a platform crops your image.
- AVIFalso: AV1 Image File Format, .avif
- AVIF is an image format derived from the AV1 video codec that compresses roughly 40-55% smaller than JPEG at matched visual quality, at the cost of much slower encoding.
- Bandingalso: posterization, colour banding
- Banding is the appearance of visible steps in what should be a smooth gradient, caused by too few distinct values being available to represent a gradual change.
- Bicubic interpolationalso: bicubic, image interpolation, resampling
- Bicubic interpolation is a resampling method that estimates each new pixel from a weighted cubic function of the sixteen surrounding source pixels, producing smoother results than simpler methods without adding any real detail.
- Bit depthalso: colour depth, bits per channel
- Bit depth is the number of bits used to store each colour channel of a pixel, determining how many distinct levels are available per channel.
- Bleedalso: print bleed, full bleed
- Bleed is extra artwork extended beyond the trim line of a printed page so that colour still reaches the edge after the paper is cut.
- Chroma subsamplingalso: 4:2:0, 4:4:4, chroma subsampling ratio
- Chroma subsampling stores colour information at lower resolution than brightness information, exploiting the eye's poor sensitivity to fine colour detail to remove data before compression even begins.
- Display P3also: P3, wide gamut
- Display P3 is a wide-gamut colour space used by modern Apple devices and many recent screens, covering roughly 25% more visible colour than sRGB, particularly in reds and greens.
- DPIalso: dots per inch
- DPI is the number of ink dots a printer places per inch, and in everyday use it is the ratio that converts an image's pixel dimensions into a physical printed size.
- EXIFalso: Exchangeable image file format, image metadata
- EXIF is metadata embedded in an image file by cameras and phones, recording capture settings, device information, orientation and frequently GPS coordinates.
- ICC profilealso: colour profile, color profile, ICC
- An ICC profile is a standardised description of how a device or file's colour numbers map to actual perceived colours, allowing the same image to look consistent across screens and printers.
- JPEGalso: JPG, .jpg, .jpeg
- JPEG is a lossy raster image format from 1992 that remains the most universally readable image format in existence, with no support for transparency.
- Lossless compressionalso: lossless
- Lossless compression makes a file smaller by encoding redundancy more efficiently, and decompression returns data bit-for-bit identical to the input.
- Lossy compressionalso: lossy
- Lossy compression permanently discards image information judged imperceptible in order to make files dramatically smaller, and the discarded data cannot be recovered.
- OCRalso: optical character recognition, image to text
- OCR is the process of detecting and transcribing text from an image into machine-readable characters.
- PDF/Aalso: PDF/A-1, PDF/A-2, archival PDF
- PDF/A is an ISO-standardised subset of PDF designed for long-term archiving, requiring a document to be fully self-contained and forbidding anything whose rendering depends on external resources.
- PNGalso: Portable Network Graphics, .png
- PNG is a lossless raster image format with full alpha transparency, excellent on screenshots and line art and very inefficient on photographs.
- PPIalso: pixels per inch
- PPI is the number of pixels per inch in an image or on a display, and it is the measurement people usually mean when they say DPI.
- Quantizationalso: quantisation, quantization table
- Quantization is the step in lossy image compression that rounds transformed image data to coarser values, and it is the point at which information is permanently lost.
- Raster vs vectoralso: bitmap vs vector, raster graphics, vector graphics
- A raster image stores a fixed grid of pixels and degrades when enlarged; a vector image stores shapes as mathematical descriptions and renders sharply at any size.
- sRGBalso: standard RGB
- sRGB is the default colour space of the web, defining a specific set of red, green and blue primaries and a transfer curve that browsers assume when an image carries no colour profile.
- Super-resolutionalso: AI upscaling, neural upscaling
- Super-resolution is the use of a trained model to increase an image's resolution by generating plausible detail consistent with its training data, rather than by interpolating between existing pixels.
- WebPalso: WebP format, .webp
- WebP is an image format developed by Google that supports both lossy and lossless compression with transparency, and produces files roughly 25-35% smaller than JPEG at matched visual quality.