Photos

The vocabulary of the submission side

Foreshortening, field of view, dynamic range, exposure, white balance, crop, baseline, control shot, slot: the photo-side terms, defined in one place.

By 5 min readPhotos

Guides on Photos: A complete protocol for comparable submissions, Everything about submitting more than one image, How to find out how much of your score is you

This glossary defines every photo-side term that can move a rating score, from angle and baseline to white balance, alphabetised in a sentence or two each. The score side has its own vocabulary, covered in the score-side glossary; most terms here get fuller treatment elsewhere on this site.

The terms

Angle. The tilt of the camera relative to the subject. Level is the one angle you can reliably reproduce; tilting down foreshortens by an amount you cannot judge by feel.

Background. Whatever fills the frame besides the subject. Left uncontrolled, it adds variance to the subjective axes without touching anything physical.

Baseline. A reference submission taken under recorded conditions, kept specifically so later results have something to compare against. Without one, no result has anything to be compared with.

Compression, JPEG. Lossy file compression that discards detail to shrink a file, more aggressively at lower quality settings. A screenshot or a re-saved copy has usually been compressed more than once.

Control shot. A fixed reference image resubmitted alongside new ones. If the control's score moves, the tool moved, and the new result inherits that shift.

Crop. Pixels removed from the edges of a file after it was captured. Unlike distance or angle, cropping does not change the lens's own geometry - see framing vs cropping for the full distinction.

Crop factor. How a sensor's field of view compares to a reference sensor at the same focal length. It is a fixed property of the camera, not something you set per shot.

Distance. How far the camera sits from the subject. The largest single source of score movement that has nothing to do with the subject, because phone lenses are wide and enlarge whatever is nearest them.

Dynamic range. The span of brightness a sensor can hold in one exposure before highlights blow white or shadows go black. Exceed it and the missing detail cannot be recovered later.

Exposure. How much light reaches the sensor during a shot, before any digital amplification is applied to it.

Field of view. How much of a scene a lens takes in at a given distance. A wider field of view means more background and more distortion near the edges of the frame.

Focal length. The property of a lens that sets its field of view. Phones switch between several lenses of different focal length automatically, which is why an app's "zoom" is sometimes a different lens rather than a crop.

Foreshortening. The apparent compression of length along the line of sight, produced by tilting toward or away from the camera. It is the mechanism behind why angle moves a score, not a separate effect from it.

Framing. What you choose to include in the shot at the moment of capture - distance, position, orientation. Distinct from cropping, which happens to the file afterward.

ISO. The amplification applied to whatever exposure the sensor captured. Raising it to compensate for a dark scene brings noise along with the extra brightness.

Log (the conditions). A written record of distance, angle, light, crop and device for a given submission. Without it, a comparison months later is a comparison between two unknown setups.

Mirror shot. A photograph of a reflection rather than of the subject directly, which adds a reflection layer and an extra distance to the geometry. Not interchangeable with a direct or front-camera shot.

Noise. Random speckle in an image, worst in shadows and worst at high ISO, that degrades exactly the texture and edge detail a rating axis is trying to read. Dodge and Karam (2016) found deep image classifiers particularly susceptible to noise and blur.

Orientation. Portrait or landscape. Changes the fill ratio of the subject in the frame and how a tool crops the image on input.

Protocol. The full written set of fixed conditions for a submission series - device, distance, angle, light, crop, background, state - recorded once and reused every time.

Resolution. The pixel dimensions of a file. Below some threshold it costs real detail; above it, more pixels mainly cost upload time.

Series. A set of submissions taken under the same protocol over time, intended to be compared against each other rather than read one at a time.

Slot. A defined position in a multi-photo set - front, side, close - that a future set can fill again in the same way, which is what makes a set comparable across sessions.

Standardised submission. A submission taken under recorded, repeatable conditions and chosen for comparability rather than for the highest score. The full definition covers what it excludes.

State. The condition of the subject at the time of the shot. A series only holds together if the state is the same across it; state is a variable like any other.

White balance. The camera's correction for the colour of the ambient light. Auto white balance can shift between otherwise identical shots, so it belongs in the written protocol rather than being left on automatic.

Reading this list

Photography terms show up on this site because a submission photo is an input to a measurement of sorts, and inputs have properties worth naming precisely rather than waving at. None of it maps onto the older, better-known vocabulary of an actual tape measurement - that method has its own terms and its own glossary, and this site does not attempt to cover them, only to name where a photo term ends and a measurement term begins. The overlap with how a human judge reads a submitted photo is real but partial: a person also benefits from good light and a level angle, for reasons of legibility rather than model input, which is a different justification arriving at a similar list. A tool built to show its breakdown rather than one blended figure makes it easier to notice which of these variables moved a given axis, which is the practical reason this vocabulary is worth having rather than trivia for its own sake.

Read next

Full archive