The standard FOV formula in microscopy, machine vision, and projection is FOV = sensor size / magnification [S2][S4]. Magnification here is the image-to-object ratio, so a 1:1 macro setup with a 10 mm sensor maps a 10 mm object field directly onto the sensor [S2].
For viewing-angle work (TVs, projection, head-mounted displays, surveillance), the FOV is the angle subtended by the image width at the viewer: FOV = 2 × atan(W / 2d), where W is the active image width derived from the diagonal via Pythagoras, and d is the eye-to-screen distance in the same unit [S5].
Core formulas and what each variable actually is
The thin-lens relationship ties focal length f, subject distance s1, and image distance s2 via 1/f = 1/s1 + 1/s2; at infinity focus, s2 collapses to f, which is why lens markings are nominal-infinity values [S2]. The marked focal length is a rounded nominal number, for example 50 mm, and the actual value can deviate by a few percent, with the true focal length roughly doubling at 1:1 macro [S2].
In the FOV = sensor / magnification identity, magnification m is the image height on the sensor divided by the object height in the scene, so a 5 megapixel 2/3" sensor (about 8.45 mm × 7.07 mm active area) behind a 0.1× lens covers roughly an 84.5 mm × 70.7 mm object field [S2][S4]. Schneider's online calculator (published 2026-04-14) implements the same identity and also outputs horizontal and vertical angle of view directly when working distance and image circle are entered [S3].
Three FOV cases that engineers actually compute
Case 1, microscope: FOV diameter on the specimen equals the eyepiece field number (FN, in mm) divided by the objective magnification; a 10×/20 FN objective pair therefore yields a 2.0 mm specimen FOV, while a 40×/20 FN pair drops it to 0.5 mm [S6]. Higher magnification always shrinks the FOV linearly: 4× → 5.0 mm, 10× → 2.0 mm, 40× → 0.5 mm, 100× → 0.2 mm on the same 20 FN eyepiece [S8].
Case 2, machine vision: pick the smallest sensor that fits the required object field at the working distance, then size the focal length to hit the magnification m = sensor / object [S2][S3]. Arducam's FOV calculator uses the same identity, taking working distance and target object size as inputs and returning the minimum imaging FOV [S7].
Case 3, display viewing angle: a 65-inch 16:9 screen (active image roughly 143.9 cm × 80.94 cm) viewed at 2.43 m fills 33° horizontally, the same angle a 50-inch screen would fill at 1.87 m, because horizontal FOV depends on the seat-to-image ratio in picture heights, not on the diagonal alone [S5]. The same 16:9 image at 1.5 picture heights (about 1.21 m for 80.94 cm tall) gives 61.3°, while at 5 picture heights (about 4.05 m) it gives 20.2° [S5].
Common pitfalls and where the formulas break
FOV calculators do not work for macro distances: at 1:1 and beyond, the simple thin-lens infinity-focus approximation loses several percent of accuracy, so reproduction-ratio arithmetic replaces the focal-length method [S2]. Internal-focusing lenses (some zooms and macros) can also shorten at close range instead of lengthening, so the standard "focal length grows at close focus" rule does not hold universally [S2].
Fisheye lenses are a separate family: their projection is non-rectilinear and they cover a wider image than the rectilinear formula predicts, with straight scene lines drawn as curves [S2]. For rectilinear lenses, sensor size must be the true active dimension in mm, not the package size, because EXIF crop-factor lookups give only an approximation and any guess on sensor size directly multiplies into the FOV error [S2].
Selection criteria: which approach fits which job
For a fixed working distance and known part size (pick-and-place, AOI, metrology), drive the calculation from FOV: divide the largest object dimension by the sensor's long edge to get the required magnification, then back-solve the focal length from the lens equation at that working distance [S2][S3]. The output of Schneider's calculator in this mode lists focal length, magnification, image circle, working distance, and object-to-image distance side by side, so mechanical clearance can be checked against housing depth before the lens is ordered [S3].
For a given lens and a known sensor, drive the calculation from the lens: read the marked focal length, divide it by the working distance, and apply m = f / (d - f) for distant subjects, or use the sensor/magnification identity directly when magnification is the spec input [S2][S4]. For viewing-angle work (control rooms, simulators, conference rooms), fix the seat distance and pick a diagonal that delivers the target horizontal angle, using the 2 × atan(W / 2d) identity rather than the sensor/magnification identity [S5].
Comparison of the three calculation paths
Path A, sensor / magnification, is best when magnification is fixed by the optics and the unknown is the object field; it works down to macro distances only if m is the true image-to-object ratio and not an infinity-focus projection [S2][S4]. Path B, horizontal angle 2 × atan(W / 2d), is best for displays and projection where the "sensor" is a screen diagonal and the "magnification" concept does not apply; it requires the aspect ratio to derive W from the diagonal [S5]. Path C, focal-length from working distance via m = f / (d - f), is best when a catalog lens is the starting point and the goal is to verify the resulting FOV before mounting; it ties directly into Schneider's calculator outputs of focal length, image circle, and total object-to-image distance [S2][S3].
For machine vision stacks that share chassis space with pressure transmitter signal lines and similar 4-wire instrumentation, the object-to-image distance from path C is the figure that has to clear the housing, not the working distance alone [S3]. For projection stacks that share a control room with flow meter HMIs, the seat-to-image distance in picture heights is the figure that drives viewing comfort, not the diagonal alone [S5].
Worked numeric examples grounded in the references
Microscope: a 20 FN eyepiece with a 10× objective gives FOV = 20 / 10 = 2.0 mm, while the same eyepiece with a 40× objective gives 0.5 mm and with a 100× oil objective gives 0.2 mm [S6][S8]. The 4× scan-table reading of 5.0 mm is the upper limit for low-magnification survey work on the same eyepiece [S8].
Machine vision: a 1/2.3" sensor with an active area around 6.17 mm × 4.55 mm behind a 0.05× lens covers 123.4 mm × 91.0 mm of object field, with required working distance solved by the thin-lens equation for the chosen focal length [S2][S3]. The 8.45 mm × 7.07 mm active area of a typical 2/3" sensor at 0.1× maps 84.5 mm × 70.7 mm, and the same sensor at 0.5× maps 16.9 mm × 14.1 mm [S2].
Display: a 65-inch 16:9 panel (active image about 143.9 cm × 80.94 cm) at 2.43 m gives 33° horizontal; the same panel at 1.21 m gives 61.3°, and at 4.05 m gives 20.2°, with the ratio of distance to picture height H (80.94 cm here) fixed at 3 H, 1.5 H, and 5 H respectively [S5].
Limitations, standards, and sourcing notes
None of the FOV formulas are ISO- or IEC-numbered; they are derivations from the thin-lens equation, which the references treat as a standard geometric identity rather than a regulated document [S2][S3]. The Schneider calculator (2026-04-14) is the most recently dated reference tool and outputs FOV, angle of view, depth of field, image circle, and total object-to-image distance from the same set of inputs, so it functions as a single cross-check for all three paths above [S3].
The two failure modes worth flagging in a spec sheet are: a 1:1 macro assumption slipped into a 1 m+ distance calculation (use the rectilinear formula, not reproduction ratio), and a fisheye lens specified where a rectilinear one was needed (the FOV will be wider than calculated, and straight lines will curve) [S2]. Edmund Optics' focal-length and FOV application note remains the canonical text reference for tying focal length, working distance, and FOV together in an imaging-lens spec [S1].
For a sanity check on the sensor side, request the active pixel area in mm from the camera vendor rather than reading the package diagonal, because the FOV error scales linearly with any sensor-size guess [S2]. For a sanity check on the display side, convert the screen diagonal in inches (1 inch = 2.54 cm exactly) to width via Pythagoras with the chosen aspect ratio, then feed width and seat distance in metres into 2 × atan(W / 2d) to recover the same 33° figure at 2.43 m for 16:9 [S5].
Trackable signals for the next review: the Schneider calculator's published date (2026-04-14) is the freshest reference in this set [S3], and the LibreTexts microscopy page (2025-12-03) is the most recent FOV table for microscope optics [S8]. Edmund's application note on focal length and FOV, Meetoptics' academy entry on the FOV = H / m identity, and the ff.de viewing-angle calculator all remain live as cross-checks for imaging-lens, sensor-sizing, and display-angle work respectively [S1][S4][S5].
For component-level specifications, see construction machinery and equipment.
Related analysis: Power-and-free vs monorail for paint line accumulation: how to choose.