Presets
Playback
Lens, focus and light
Quick actions
Camera
Layers
Quality
Model check
The science
Thin lens. A lens of focal length f images an object at distance s to a point at distance v behind the lens with 1/s + 1/v = 1/f, so v = f s/(s - f). Focusing a camera means moving the sensor (here on the bellows) to v(sF). Magnification at focus is m = v/s = f/(sF - f).
Aperture. The f-number is N = f/D, so the opening has diameter D = f/N and area proportional to 1/N2. Each stop (a factor of √2 in N) halves the light, so the shutter time for the same exposure follows t = N2 / 2EV (EV for ISO 100, shifted by log2(ISO/100)).
Blur circle. A point at distance s whose image lies at v(s) ≠ v(sF) is cut by the sensor plane in a cone of rays: a disc of diameter b = D |v(sF) - v(s)| / v(s) = f2 |s - sF| / (N (sF - f) s). In the 3D view the cones are drawn from the subjects through the iris to the sensor, where the blur discs appear. The disc has the shape of the iris (the bokeh polygon).
Depth of field. A point still looks sharp when b ≤ c, the acceptable blur circle (here the sensor width / 1500). That gives the hyperfocal distance H = f2/(N c) + f and the limits near = sF (H - f)/(H + sF - 2f) and far = sF (H - f)/(H - sF), infinite once sF ≥ H. Smaller aperture (larger N), shorter f or a more distant focus all deepen the field.
Honest simplifications. The lens is ideal and thin: no aberrations, vignetting or glass thickness (real principal planes shift the numbers slightly). The picture is a teaching-scale view camera (f = 0.3 to 0.9 m, a 360 mm sensor), drawn in real metres so the rays are readable; the same equations apply to a small camera, but its sensor, c and f must be scaled together. Diffraction (Airy disc diameter 2.44 λ N v/f) is shown as a number only because on this large sensor it is far below c. The acceptable blur c is a convention, not a law of physics. Subject sizes in the sensor picture use the true projection x' = X v/s; blur is drawn with a Gaussian filter of similar width to the disc.