A practical guide to using cameras to control perspective—and to understanding how viewpoint, focal length, field of view, focus, movement, cinematography and modern imaging systems transform spatial reality into images.
Cameras are everywhere. They are used for still photography, cinema, television, smartphones, surveillance, scientific imaging, robotic vision, computer vision, autonomous vehicles, drones, aerial survey and satellite remote sensing.
Although these systems differ greatly in scale and purpose, they share a fundamental perspective operation: they receive information from spatial reality from one or more viewpoints and transform it into an image or other spatial representation.
Learning how cameras work therefore means learning how Camera Perspective works.
The most important principle is also frequently misunderstood:
Move the camera and you change the perspective. Change the focal length from the same camera position and you primarily change how much of that perspective is recorded and how large it appears in the image.
This page is not simply a guide to photographic technique. It is a practical manual for understanding and using the camera as a perspective instrument.
A Camera Is a Perspective Instrument
A camera selects, forms and records a perspective image of spatial reality.
In its simplest form:
Object or scene → camera viewpoint → optical system → image plane or sensor → image
Modern imaging usually involves a longer chain:
Physical scene → camera optics → sensor → image processing → stored or transmitted image → display → eye → retinal image → visual interpretation
Every stage can affect the final result.
A camera system can determine or influence:
- viewpoint;
- viewing direction;
- field of view;
- projection geometry;
- image scale;
- focus;
- Depth of Field;
- exposure;
- motion blur;
- resolution;
- colour;
- contrast;
- distortion;
- framing and cropping;
- the spatial information eventually available to the observer or machine.
The resulting image is therefore not simply “the scene”. It is the scene as transformed by a particular perspective imaging system.
Rule One: Choose the Viewpoint Before the Lens
The position of the camera is the primary control of geometrical perspective.
Before thinking about focal length, ask:
- Where should the camera be?
- How close should it be to the subject?
- How high should it be?
- Should it be directly in front, to one side, above or below?
- Which surfaces should be visible?
- Which objects should overlap?
- How large should foreground objects appear relative to background objects?
Moving the camera changes the spatial relationships represented in the image.
It can change:
- relative apparent size;
- occlusion;
- overlap;
- Aspect of Form;
- foreshortening;
- line direction;
- vanishing relationships;
- foreground-to-background scale;
- which surfaces are visible or concealed.
These changes cannot normally be reproduced merely by changing focal length from the original camera position.
Distance Changes Perspective
Camera-to-subject distance has a powerful effect on perspective.
Move close to a subject and differences between nearer and farther parts become exaggerated. Near parts subtend much larger visual angles than parts only slightly farther away.
Move farther away and these relative differences decrease.
A human face provides an obvious example. Photograph it from very close range and the nose can appear disproportionately large relative to the ears and sides of the head. Move farther away and the relative proportions become much less exaggerated.
The principal cause is viewpoint distance, not merely the use of a wide-angle lens.
Choose camera distance for perspective. Choose focal length for framing.
Camera Height Matters
Raising or lowering the camera produces a different perspective.
A high camera may reveal more upper surfaces. A low camera may reveal more undersides and can make nearby forms visually dominant.
Camera height also determines the relationship between the image and eye level or horizon level.
In an ordinary level view across a horizontal environment, the horizon corresponds to the level of the camera viewpoint. Raise the camera and the horizon rises relative to the scene. Lower the camera and it falls.
Camera height is consequently fundamental in:
- architecture;
- interiors;
- portraits;
- landscapes;
- product photography;
- cinematography;
- aerial and drone imaging;
- survey and measurement.
Move Sideways and the View Changes
Move the camera sideways and a different set of spatial relationships becomes visible.
Objects that previously overlapped may separate. Hidden surfaces may appear. Apparent alignments change.
This changing relationship is the basis of parallax.
When the camera moves, nearby objects generally change image position more strongly than distant objects. Sequential images taken from changing viewpoints therefore contain important information about spatial depth.
This principle is fundamental not only to photography but also to stereoscopic imaging, photogrammetry, computer vision, robotic navigation and three-dimensional reconstruction.
Direction and Orientation
The position of the camera and the direction in which it points are separate variables.
The camera can remain at approximately one station point while changing its orientation.
Pan
A pan rotates the camera horizontally.
Tilt
A tilt rotates it upwards or downwards.
Roll
A roll rotates the camera around its forward viewing direction.
These operations change viewing direction or image orientation without necessarily translating the camera centre through space.
This distinction becomes particularly important when comparing:
- pan with tracking;
- tilt with pedestal movement;
- zoom with dolly movement.
Focal Length Does Not by Itself Determine Perspective
For a given camera or sensor format:
- a shorter focal length records a wider field of view;
- a longer focal length records a narrower field of view.
But if the camera remains in the same position, changing focal length does not fundamentally change the relative geometrical perspective of the scene.
The same surfaces remain visible. The same objects overlap. The same near–far relationships continue to originate from the same camera position.
A longer-focal-length photograph taken from the same position is geometrically comparable to taking a smaller central region of the wider image and enlarging it, subject to differences of optical quality, resolution, focus and Depth of Field.
Focal length nevertheless remains important because it determines the field selected from that viewpoint and its scale on the sensor.
Field of View
Field of view is the angular extent of the spatial scene recorded by the camera.
It depends upon both:
- focal length;
- image or sensor format.
A shorter focal length on a given format normally records a wider field. A longer focal length records a narrower field.
The same focal length used on a smaller sensor also records a narrower field because less of the lens image is captured.
This is why focal length cannot always be compared meaningfully without also knowing the image format or sensor size.
Wide-Angle Photography: Two Different Effects
The expression “wide-angle distortion” is frequently used for several quite different perspective phenomena.
Close-Viewpoint Perspective
A wide-angle lens often encourages the photographer to move closer to the subject while retaining a large field.
The closer viewpoint exaggerates near–far size relationships.
This can make:
- foreground objects exceptionally large;
- background objects comparatively small;
- faces exaggerated at close range;
- interiors appear unusually deep.
Rectilinear Edge Stretching
A separate effect occurs when a wide angular field is projected onto one flat rectilinear image plane.
Straight spatial lines remain straight, but planar scale per degree of visual angle increases towards the image margins.
Rounded and volumetric Forms near the edges may therefore appear:
- stretched;
- elongated;
- enlarged;
- asymmetrical.
This is a property of wide-field rectilinear projection and should not automatically be interpreted as a defective lens.
Telephoto Compression
It is common to say that telephoto lenses “compress perspective”, but this can conceal the real cause.
If the camera remains fixed, increasing focal length primarily changes framing and magnification.
The familiar compressed appearance normally arises because the photographer moves farther away and then uses a longer focal length to retain the required subject size.
The greater camera distance reduces the relative difference between camera-to-foreground and camera-to-background distances.
The changed camera position therefore produces the changed perspective. The long lens permits that more distant view to fill the frame.
Camera Tilt and Converging Verticals
Photograph a tall building while keeping the camera image plane vertical and its vertical edges may remain parallel in an ordinary rectilinear photograph.
Tilt the camera upwards and the image plane is no longer parallel to those spatial verticals. They then converge towards a finite vanishing position.
This familiar effect is often incorrectly attributed to a wide-angle lens.
The principal cause is camera orientation.
Where parallel architectural verticals are required, possible methods include:
- keeping the camera image plane vertical;
- raising the camera;
- using lens or sensor shift;
- using a perspective-control lens;
- applying controlled projective correction after capture.
Lens Distortion Is Not Perspective Distortion
Real lenses may also depart from ideal projection.
Barrel Distortion
Straight lines bow outwards from the image centre.
Pincushion Distortion
Straight lines bend inwards.
Moustache or Wave Distortion
The radial distortion changes character across the image.
Decentring or Tangential Distortion
Misalignment within the optical system produces asymmetric displacement.
These are optical lens distortions.
They should be distinguished from:
- close-viewpoint exaggeration;
- foreshortening;
- rectilinear edge stretching;
- converging verticals;
- ordinary perspective diminution.
Rectilinear, Fisheye and Other Camera Projections
Not every camera maps surrounding space onto the image in the same way.
A conventional rectilinear lens attempts to preserve straight scene lines as straight image lines.
A fisheye lens employs a different wide-field mapping and allows many spatial straight lines to become curved.
Panoramic, cylindrical, spherical and computational systems employ still other mappings.
There is no single flat projection that can simultaneously preserve every desirable property across an unlimited field.
Different imaging systems therefore make different compromises between:
- straightness;
- angular scale;
- local shape;
- area;
- magnification;
- field coverage.
Still Camera Shots
A camera shot is not defined only by the equipment used. It also describes the relationship between the camera, subject, field and resulting composition.
Still photography uses many of the same viewpoint and framing categories later developed extensively in cinematography.
Extreme Close-Up
Records a very small part of the subject at large image scale. Fine texture and local Form dominate while much surrounding spatial context is removed.
Close-Up
Concentrates upon a subject or part of a subject while retaining slightly more surrounding information.
Medium Shot
Shows the principal subject together with a meaningful portion of its immediate spatial surroundings.
Long Shot
Places the subject within a larger environment, making its relation to surrounding spatial structure more evident.
Extreme Long or Wide Shot
Emphasises the wider environment, landscape, architecture or spatial setting rather than the fine structure of an individual subject.
Establishing Shot
Provides an overall view that establishes where subsequent closer views are located within the larger scene.
High-Angle Shot
The camera is positioned above the subject and looks downwards.
Low-Angle Shot
The camera is positioned below the subject and looks upwards.
Bird’s-Eye or Overhead View
The camera observes the scene from a very high or nearly vertical downward direction.
Worm’s-Eye View
The camera adopts an exceptionally low viewpoint, often looking substantially upwards.
Point-of-View Shot
The camera is positioned so that the image approximates the viewpoint attributed to a particular observer or participant.
Panoramic or Wide-Field Shot
A broad horizontal or multi-directional field is recorded to reveal a larger spatial environment.
These are not merely compositional labels. Each shot establishes a particular relationship between viewpoint, viewing direction, field of view, image scale and target space.
Moving Cameras and Cinematography
Cinematography extends Camera Perspective through time.
Instead of presenting only one fixed view, a moving-image camera can record a succession of related viewpoints, viewing directions and fields.
This allows perspective itself to change continuously during the shot.
Pan Shot
The camera rotates horizontally from approximately one station point, revealing successive viewing directions across the scene.
Tilt Shot
The camera rotates upwards or downwards while remaining at approximately the same location.
Dolly Shot
The entire camera moves forwards or backwards through space.
Unlike zooming, a dolly movement changes the camera position and therefore changes perspective, parallax and near–far relationships.
Tracking or Trucking Shot
The camera moves laterally or follows a path alongside the subject. Foreground and background objects exhibit changing parallax.
Following Shot
The camera moves with a subject, maintaining a changing spatial relationship as both travel through the environment.
Pedestal Shot
The camera moves vertically upwards or downwards without merely tilting.
Crane or Jib Shot
The camera travels through a larger vertical and horizontal path, producing continuously changing viewpoint, elevation and visible spatial relationships.
Orbit Shot
The camera moves around a subject while continuing to direct its view towards it. This is closely related to the Sphere of Revolution.
Hand-Held Shot
The camera moves with the operator and may combine continual translation, rotation and vibration.
Steadicam or Gimbal Shot
A stabilised camera moves through spatial reality while reducing unwanted rotational or translational irregularity.
Drone Shot
A remotely controlled or autonomous camera can move freely through three-dimensional space, combining changes of height, lateral position, distance and viewing direction.
Zoom Shot
The focal length changes while the camera position may remain fixed. Image scale and field of view change, but a new camera viewpoint is not produced.
Dolly Zoom
The camera moves while focal length changes in the opposite direction, keeping a principal subject approximately constant in image size while the background perspective expands or contracts.
The Dolly Zoom dramatically demonstrates that framing and perspective are not the same thing.
Camera Movement, Subject Movement and Motion Perspective
Moving-image perspective can arise because:
- the camera moves;
- the object moves;
- both camera and object move;
- the camera rotates;
- focal length changes;
- focus changes;
- the scene itself transforms.
The viewer can therefore receive changing information about:
- relative depth;
- distance;
- parallax;
- occlusion;
- Aspect of Form;
- spatial continuity;
- movement through the environment.
Cinema is consequently not simply a succession of still photographs. It can present a continuously changing perspective relationship through time.
Focus and Depth of Field
Focus determines which object distance is imaged most sharply.
Depth of Field is the object-space range over which subjects appear acceptably sharp.
It is influenced by factors including:
- focus distance;
- aperture;
- focal length;
- image format;
- final image size and viewing conditions.
A shallow Depth of Field can visually isolate one spatial region while leaving nearer or farther objects defocussed.
A greater Depth of Field can preserve recognisable structure across a larger range of object distances.
Defocus itself can also provide information about relative spatial depth.
Depth of Focus Is Different
Depth of Field and Depth of Focus are related but different.
Depth of Field concerns acceptable sharpness within object space in front of the camera.
Depth of Focus concerns the image-space tolerance behind the lens around the receiving sensor or film plane.
The two terms should therefore not be used interchangeably.
Aperture
The aperture controls the effective opening through which light enters the camera.
It affects exposure but also influences Depth of Field.
In general:
- a wider aperture tends to produce a shallower Depth of Field;
- a smaller aperture tends to increase Depth of Field, within practical optical limits.
Aperture changes optical image character without itself relocating the camera viewpoint.
This is another reason to distinguish perspective geometry from other optical characteristics of an image.
Exposure and Time
A camera can integrate light over a selected interval of time.
A short exposure can freeze rapid movement.
A longer exposure can transform movement into:
- motion blur;
- light trails;
- softened water;
- blurred people or vehicles;
- star trails;
- continuous traces of changing position.
The resulting image may therefore contain spatial information gathered across multiple moments rather than representing one perfectly instantaneous state.
Resolution and Optical Vanishing
An object can remain geometrically projected by the camera even when its finer details can no longer be resolved.
As objects recede, their projected images become smaller. Fine structure may eventually fall below the combined resolving ability of:
- the lens;
- the sensor;
- focus;
- atmospheric transmission;
- processing;
- compression;
- display or print;
- the final observer.
Features can therefore disappear without reaching a geometrical vanishing point.
This belongs to the wider problems of Optical Vanishing, Diminution of Form, resolution and visible detail.
Atmospheric Perspective Still Applies to Cameras
A camera does not remove the atmosphere between itself and the target.
At large distances, scattering, absorption, haze, moisture and airborne particles can reduce:
- contrast;
- colour saturation;
- clarity;
- visible detail;
- separation between distant Forms.
This is Atmospheric Perspective.
A distant mountain can therefore be accurately focussed yet still appear pale, low in contrast or blue-grey because the atmosphere has altered the light reaching the camera.
Atmospheric loss of clarity should therefore not be confused with optical defocus.
Zooming Is Not Moving
A zoom changes focal length while the camera centre can remain approximately fixed.
A dolly or track physically moves the camera through space.
Moving the camera changes:
- parallax;
- occlusion;
- near–far scale relationships;
- Aspect of Form;
- visible surfaces;
- spatial perspective.
Thus:
Zoom = change focal length, image scale and field of view.
Dolly or track = change viewpoint and therefore perspective.
Looking Around: Sphere of Vision
A camera can remain at approximately one location while photographing different directions around itself.
Conceptually, these changing directions can be organised around a Sphere of Vision.
This principle underlies:
- panoramic photography;
- 360° imaging;
- spherical panoramas;
- environment mapping;
- Virtual Reality capture.
Looking At: Sphere of Revolution
A camera can alternatively move around a subject while continuing to look towards it.
This is closely related to the Sphere of Revolution.
Successive viewpoints reveal different:
- surfaces;
- profiles;
- occlusions;
- Aspects of Form;
- foreshortenings;
- spatial relationships.
This principle is fundamental to:
- product photography;
- turntable imaging;
- photogrammetry;
- 3-D reconstruction;
- visual-effects capture;
- multi-view imaging.
Stereo and Multi-Camera Perspective
Most conventional photographs are formed from one camera centre.
A stereoscopic system uses two spatially separated cameras or viewpoints to form two related images.
The differences between these images provide binocular disparity or stereoscopic parallax.
Multi-camera systems extend this principle further and are widely used for:
- photogrammetry;
- motion capture;
- volumetric imaging;
- light-field imaging;
- 3-D reconstruction;
- scientific measurement;
- machine vision;
- autonomous navigation.
Additional viewpoints can reveal spatial information that is concealed or ambiguous within a single image.
Cameras Beyond Photography
The camera is now one of the fundamental sensing instruments of modern technology.
Many machines do not use cameras primarily to make pictures for human viewing. They use them to measure, recognise, locate, classify and navigate through spatial reality.
Computer Vision
Computer vision uses digital images and image sequences to extract information about objects, scenes, movement and spatial relationships.
Camera images can be analysed to identify:
- objects;
- people;
- faces;
- edges and contours;
- surfaces;
- depth;
- movement;
- orientation;
- scale;
- position;
- three-dimensional structure.
Computer vision therefore faces many of the same fundamental perspective problems encountered by human vision: a two-dimensional image must be interpreted in relation to an underlying three-dimensional reality.
Robotic Vision
Robots commonly use one or more cameras as part of their visual sensing systems.
Camera-based robotic vision can help a machine:
- locate itself;
- recognise objects;
- estimate depth;
- avoid obstacles;
- follow routes;
- inspect components;
- grasp objects;
- navigate through buildings or outdoor environments.
Robotic vision may combine conventional cameras with stereo cameras, depth cameras, infrared sensors, LiDAR and other instruments.
The underlying task remains perspectival: the system must relate image information to spatial structure and position.
Autonomous Vehicles
Self-driving vehicles and driver-assistance systems use multiple cameras to observe roads, traffic, pedestrians, markings, signs and surrounding objects.
Images from different cameras and successive moments can be combined with other sensing systems to estimate:
- distance;
- direction;
- relative movement;
- road geometry;
- object position;
- free navigable space.
The vehicle is therefore continually interpreting changing Camera Perspective while itself moving through spatial reality.
Drones and Aerial Imaging
Drones carry cameras through three-dimensional space and can obtain viewpoints that were previously difficult or expensive to achieve.
They are used for:
- photography and filmmaking;
- mapping;
- surveying;
- inspection;
- agriculture;
- construction monitoring;
- archaeology;
- emergency response;
- photogrammetric 3-D reconstruction.
The camera viewpoint may change continually in height, direction and horizontal position, producing a particularly rich form of moving Camera Perspective.
Satellite Cameras and Remote Sensing
Cameras and related imaging instruments are also carried far above the Earth.
Earth-observation satellites use optical and multispectral imaging systems to record large areas of the Earth’s surface from orbital viewpoints.
Satellite imaging is therefore an extreme form of Camera and Instrument Perspective: the camera may be hundreds or thousands of kilometres from the target while recording selected angular fields at very high spatial resolution.
Remote-sensing imagery is used to study:
- land use;
- weather;
- vegetation;
- agriculture;
- oceans;
- ice;
- urban development;
- geology;
- environmental change;
- natural disasters;
- mapping and surveying.
Some satellite systems record ordinary visible wavelengths, while others use infrared, thermal or multispectral imaging.
Remote sensing also includes non-camera systems such as radar and LiDAR, but camera-like optical imaging systems remain one of its principal forms of spatial observation.
The resulting image must often be corrected for:
- viewing angle;
- Earth curvature;
- satellite movement;
- terrain relief;
- atmospheric effects;
- sensor geometry;
- map projection.
This makes satellite imagery a powerful demonstration that Camera Perspective extends far beyond conventional photography.
The Camera as a Measuring Instrument
Once the geometry of the camera system is known, images can be used not merely for representation but also for measurement.
Camera-based systems can estimate:
- angles;
- position;
- distance;
- depth;
- size;
- shape;
- surface orientation;
- movement;
- three-dimensional coordinates.
This principle underlies much of:
- photogrammetry;
- machine vision;
- robotics;
- surveying;
- remote sensing;
- 3-D scanning;
- medical and scientific imaging.
The camera is therefore simultaneously an image-making instrument and a spatial information instrument.
Cropping Does Not Move the Camera
Cropping changes which part of the recorded image is finally presented.
It can strongly change composition, apparent image scale and field presented to the viewer.
But cropping does not retrospectively change the original station point.
If two images are made from exactly the same camera position, a crop from a wider image can reproduce much of the geometrical perspective of a tighter photograph made with a longer focal length from the same position, provided sufficient resolution is available.
This again demonstrates the distinction between field of view and viewpoint perspective.
Digital Perspective Correction
Software can transform camera images after capture.
It can:
- straighten converging architectural lines;
- correct barrel or pincushion distortion;
- alter aspect;
- warp local areas;
- stitch multiple photographs;
- reproject panoramic views;
- alter cropping and framing.
These operations change the representation. They do not physically move the original camera viewpoint.
A corrected photograph can imitate selected geometrical characteristics of another image-plane orientation, but it cannot automatically recover every part of the three-dimensional scene that would have been visible from another station point.
The Photograph Is Only One Stage of the Image Chain
The camera image may subsequently be enlarged, reduced, cropped, printed, transmitted, compressed, projected or displayed on screens of many different sizes.
The complete sequence may therefore become:
Spatial reality → Camera Perspective → recorded image → processed image → displayed image → Visual Perspective Type 2
A photograph viewed on a mobile phone is not optically identical to the same image projected across a cinema screen.
The stored image may contain the same underlying geometry, but it now subtends a very different visual angle at the final observer.
The capture stage and viewing stage should therefore be distinguished.
A Camera Does Not See Exactly Like an Eye
The eye and camera share important optical principles, but they are not equivalent systems.
A conventional camera commonly has:
- one principal optical centre;
- a selected bounded field;
- a comparatively flat sensor;
- a chosen exposure interval;
- relatively uniform digital image sampling;
- a stored image independent of the original scene.
Human vision normally involves:
- two moving eyes;
- curved retinas;
- high central and lower peripheral acuity;
- changing fixation;
- accommodation and vergence;
- eye, head and body movement;
- binocular integration;
- perceptual interpretation across time.
A photograph may reproduce selected geometrical aspects of a view remarkably well, but it is not literally identical to what an observer sees.
A Camera Image Does Not Tell You Everything
A camera image is a projection or imaging outcome derived from spatial reality, not the original three-dimensional reality itself.
A single photograph may leave unknown:
- absolute distance;
- absolute scale;
- hidden surfaces;
- the exact Physical Form of ambiguous objects;
- the original camera position;
- the exact focal length and sensor format;
- whether digital transformation has occurred;
- whether the photographed scene itself was manipulated.
Different three-dimensional arrangements may sometimes produce identical or very similar images from one viewpoint.
Camera Perspective therefore participates in the wider Correspondence / Equivalence Problem.
A camera image can be extraordinarily informative, but it remains viewpoint-dependent and system-dependent information about reality.
A Practical Method for Controlling Camera Perspective
1. Decide What Spatial Relationship You Want
Strong foreground depth? A compressed distant view? Symmetrical architecture? A high viewpoint? A low viewpoint? An immersive wide field?
2. Move the Camera
Choose the station point that produces the required relative sizes, visible surfaces, overlaps and Aspect of Form.
3. Set Camera Height
Establish the required eye level, horizon relationship and visibility of upper or lower surfaces.
4. Set Viewing Direction
Pan or tilt deliberately rather than allowing camera orientation to change accidentally.
5. Choose the Required Field
Select focal length and image format to include the required portion of the scene from the viewpoint already chosen.
6. Check the Image Margins
With wide rectilinear fields, inspect people and rounded Forms near the edges for strong lateral stretching.
7. Set Focus and Depth of Field
Decide which range of spatial depth must remain acceptably sharp.
8. Set Exposure in Relation to Movement
Decide whether movement should be frozen, blurred or represented as a temporal trace.
9. Distinguish Optical Distortion from Perspective
Check separately for genuine barrel, pincushion or other lens aberrations.
10. Consider the Final Use
The optimal image may differ depending upon whether it will be viewed by a person, projected in a cinema, analysed by a computer, used for measurement or processed as remote-sensing data.
Six Simple Camera Experiments
1. Viewpoint versus Focal Length
Keep the camera fixed and photograph the same scene with several focal lengths. Compare equivalent central portions of the images.
The field changes greatly; the underlying viewpoint perspective does not.
2. Change the Viewpoint
Photograph a subject from close range and then from farther away. Change focal length so that the principal subject remains approximately the same image size.
The background relationship changes because the viewpoint changed.
3. Tilt a Building
Photograph a tall building with the camera image plane vertical, then tilt upwards.
Compare the vertical convergence.
4. Test Edge Stretching
With a wide rectilinear lens, photograph a person first near the image centre and then near the edge from the same camera position.
Compare the Apparent Form.
5. Zoom versus Dolly
Zoom towards a subject without moving. Then return to the original focal length and physically move towards the subject.
The two operations produce different perspective results.
6. Move Around the Subject
Photograph the same object from several positions around it while keeping it approximately centred.
Compare how visible surfaces, occlusions, Aspect of Form and apparent depth change.
The Essential Rules
- The camera viewpoint is the primary control of perspective geometry.
- Choose camera distance for perspective and focal length for framing.
- Changing focal length from a fixed position chiefly changes field of view and image scale.
- Camera height changes horizon relationships and visible surfaces.
- Moving the camera produces new perspective and parallax information.
- Zooming is not the same as dollying or tracking.
- Close-viewpoint exaggeration should not simply be called wide-angle lens distortion.
- Rectilinear edge stretching is distinct from close-viewpoint perspective.
- Camera tilt is the principal cause of converging architectural verticals in an ordinary rectilinear image.
- Barrel and pincushion distortion are optical lens distortions.
- Depth of Field concerns object space; Depth of Focus concerns image space.
- Cinematography extends Camera Perspective through time.
- Rotating the camera changes viewing direction; translating it changes viewpoint.
- Still, moving, stereo and multi-camera systems produce different kinds of perspective information.
- Robotic and computer vision use camera images to infer spatial structure.
- Drones and satellites extend Camera Perspective to aerial and orbital viewpoints.
- Remote sensing uses optical cameras and related imaging systems alongside other sensing technologies.
- A camera can be both an image-making instrument and a measuring instrument.
- Cropping changes the presented field but does not change the original viewpoint.
- A camera image is one stage within a larger perspective image chain.
- A photograph is not identical to human visual experience.
Learn to Use the Camera as a Perspective Instrument
The most effective way to control Camera Perspective is to stop treating the lens as the first decision.
First choose where the camera should be.
Move until the relative apparent sizes, visible surfaces, overlaps, aspect and spatial depth are appropriate. Establish camera height and viewing direction. Only then choose focal length and format to frame the required field.
Next control focus, Depth of Field, exposure, movement, resolution and optical correction.
The same principles extend far beyond conventional photography. Cinema cameras, smartphones, stereo rigs, drones, robots, autonomous vehicles and Earth-observation satellites all use perspective imaging to obtain information about spatial reality.
The camera is therefore not merely a device for taking pictures. It is one of the principal modern instruments for seeing, recording, measuring, interpreting and navigating space.
Related Pages
Continue with these closely related Perspective Research Centre pages:
- Camera Perspective → — the wider theory of viewpoint, optical projection and photographic image formation.
- Visual Perspective → — how camera images and other visible appearances are finally received and interpreted through vision.
- Natural Perspective → — perspective arising from spatial reality before it is transformed by the camera.
- Perspective Phenomena → — diminution, foreshortening, convergence, occlusion, parallax, Atmospheric Perspective and related effects.
- Optical & Natural Perspective Views → — natural and optically formed views of spatial reality.
- Perspective Defined → — the wider framework within which Camera Perspective belongs.
- Dictionary of Perspective → — search approximately 1,200 named perspective types and forms and the wider terminology of camera, optical and visual perspective.
For the fuller theoretical treatment of these subjects, see The Art and Science of Perspective, Volume 1: The Past, Present and Future of Visual and Optical Perspective and the Dictionary of Perspective, Second Edition.