Telescope Angular Resolution: Rayleigh Limit, Airy Disks and Camera Sampling

How much angular detail can a telescope separate?

For an ideal clear circular aperture, the Rayleigh separation is approximately θ=1.22λ/D radians, where λ is the wavelength and D is the aperture diameter in the same length unit. Multiply by 206264.806247 to convert radians to arcseconds. At 550 nm, a 200 mm aperture gives approximately 0.692018 arcseconds.

This is a diffraction-based criterion for separating point-source images. It is not a guarantee that any two objects at that separation will be visibly distinct through an actual telescope. Atmospheric seeing, optical alignment, brightness contrast, noise and camera sampling affect the observation.

The examples below are calculated instrument models, not reviews of commercial equipment. They show what changes when aperture, wavelength or focal length changes, while keeping the assumptions visible.

What the Rayleigh criterion actually compares

A monochromatic point source does not form a mathematical point in a perfect telescope. A clear circular pupil produces a bright central diffraction pattern surrounded by rings. The central region out to the first dark ring is commonly called the Airy disk.

The Rayleigh criterion places the centre of one point-source pattern at the first minimum of the other. For the ideal circular-aperture model, that angular radius is approximately 1.22λ/D. The central disk's first-minimum diameter is twice that angle, approximately 2.44λ/D.

Do not confuse this radius with the full width at half maximum, which uses a different definition of image width. A value labelled “resolution” is incomplete until its criterion is stated. Rayleigh separation, an empirical double-star rule, and a measured stellar-image width are not automatically interchangeable quantities.

Convert the units before using the formula

For D=200 mm, convert to 0.2 m. For λ=550 nm, convert to 5.50×10⁻⁷ m. Then θ=1.22×5.50×10⁻⁷/0.2=3.355×10⁻⁶ radians. Multiplying by 180/π and then by 3600 gives 0.692018 arcseconds after rounding.

One arcminute contains 60 arcseconds, and one degree contains 3600 arcseconds. The quoted telescope diameter is the entrance aperture diameter, not its radius. Substituting a 100 mm radius for a 200 mm diameter doubles the predicted separation and makes the instrument look two times worse in this model.

If wavelength is entered in nanometres and diameter in millimetres, include a factor of 10⁻⁶ in λ/D. Forgetting it produces a millionfold unit error. Keep the intermediate answer in radians so that a mistaken conversion can be isolated.

Compare eight apertures at one wavelength

At a fixed wavelength, doubling D halves the Rayleigh separation. Going from 100 to 200 mm changes the model from about 1.384 to 0.692 arcseconds. Going from 200 to 400 mm halves it again. The aperture dependence is inverse, not inverse square.

Light-collecting area follows a different relation: area is proportional to D² for an unobstructed aperture. Doubling diameter gives four times the geometric area, before considering transmission or detector efficiency. Collecting more light and resolving a smaller angle are related advantages with different scaling laws.

The table's many decimal places make the arithmetic reproducible. They do not mean a real telescope's performance has been established to a millionth of an arcsecond. The 1.22 coefficient itself is a conventional rounding of the circular-aperture first-zero result.

Read the aperture–wavelength curves

Each curve holds wavelength fixed while aperture varies from 50 to 400 mm. The 800 nm curve lies above the 550 nm curve because longer wavelengths give larger diffraction angles at the same aperture. A point lower on this graph means a smaller angular separation, not less light.

The curves contain 351 samples each. They do not show a sequence of observations or imply that a telescope is equally well corrected at every wavelength. A broadband image combines light over a range of wavelengths rather than producing one unique monochromatic Airy pattern.

Check wavelength separately from atmosphere

At D=200 mm, changing from 550 to 1100 nm would double the diffraction angle. Changing to a shorter wavelength reduces the ideal angle. This comparison holds optical quality and all other conditions fixed; it is a statement about diffraction scaling.

An actual observation does not hold everything else fixed. Atmospheric behaviour, source brightness, filter transmission and detector response can change across wavelength. A shorter-wavelength theoretical number is therefore not sufficient to predict which real image will reveal more useful detail.

State whether a quoted wavelength is a monochromatic model, a filter's representative value or part of a full spectral calculation. Treating a broad filter as one wavelength can be a useful estimate, provided that approximation is identified.

Work backward from a target separation

Rearrange the equation as D=1.22λ/θ, converting the target angle to radians before division. For θ=1 arcsecond and λ=550 nm, the ideal diameter is about 138.404 mm. For half that angle, twice the diameter is required under the same assumptions.

This is a mathematical lower-scale estimate for the specified criterion, not a purchase recommendation or a promise of resolving a faint companion. A bright nearby primary, a turbulent atmosphere or poor focus can make the practical task substantially harder.

A target's angular separation and the instrument's resolving criterion are different inputs. The great-circle separation guide calculates the angle between sky coordinates. The present guide asks how that angle compares with an image-forming system.

Focal length controls image scale, not the aperture limit

For pixel pitch p in micrometres and effective focal length f in millimetres, the small-angle pixel scale is s≈206.264806×p/f arcseconds per pixel. With p=3.76 μm and f=1000 mm, s≈0.775556 arcseconds per pixel. This is the angle covered by one pixel, not the optical Rayleigh limit.

The physical radius to the first Airy minimum is r≈1.22λ(f/D). At 550 nm and f/5, r=3.355 μm, so the first-minimum diameter is 6.710 μm. The radius is about 0.8923 of a 3.76 μm pixel, and the diameter about 1.7846 pixels.

Increasing effective focal length spreads the same angular pattern over more pixels. It can improve sampling of detail already delivered by the aperture, while leaving 1.22λ/D unchanged. Record effective focal length after any amplifying optics; the native focal length alone may not describe the camera configuration.

Use the diffraction sampling condition carefully

For an ideal incoherent diffraction-limited image, the angular pixel pitch at the Nyquist boundary is λ/(2D) radians. In the focal plane this becomes p≤λ(f/D)/2, with matching length units. The ST-ECF report's equation 9 gives the corresponding angular sampling relation.

At 550 nm and f/5, the maximum pitch from this condition is 1.375 μm. A 3.76 μm pixel exceeds it. At f/15, the boundary is 4.125 μm, so the same 3.76 μm pitch meets this ideal geometric sampling condition. Noise, pixel response and the actual optical transfer still matter.

Because the Rayleigh angle is 1.22λ/D, it spans 2.44 Nyquist-boundary pixels. “Two pixels across the Rayleigh separation” is therefore a different, coarser rule. Avoid swapping a casual resolution-element rule for a precisely stated sampling criterion.

Rebinning an existing image into smaller displayed pixels does not change the original sampling. Multiple shifted exposures and reconstruction methods introduce additional assumptions; this simple worksheet makes no claim that enlarging a single image recovers missing information.

Atmospheric seeing is a separate limitation

ESO's optics explanation distinguishes telescope errors from changing atmospheric wavefront distortions. Focus and alignment address some instrumental errors; adaptive optics addresses atmospheric distortions within its operating limits. A diffraction-only curve does not include either class of imperfection.

Imagine an observation with a measured 2 arcsecond stellar-image width while a 200 mm aperture has the 0.692 arcsecond Rayleigh value above. Those two numbers use different image-width concepts. They show that atmospheric or other blur can dominate, but they should not simply be added or combined by a square-root rule without specifying a compatible image model.

The airmass guide describes path length through the atmosphere. Airmass alone is not a complete seeing model. Likewise, tracking errors and field rotation can smear a long exposure even when the optical aperture is capable of finer detail.

Brightness contrast and magnification change the practical task

Two equally bright model point sources are easier to discuss than a faint source beside a much brighter one. Diffraction rings, scattered light and noise can hide the faint companion. The magnitude and flux guide quantifies the brightness ratio, which is a separate dimension of the problem.

Extended objects introduce additional contrast questions. Detecting a dark line, recognising a crater and separating two stars are not identical tests. A low-contrast feature can be difficult even when its nominal angular scale exceeds the Rayleigh number.

Magnification enlarges the image presented to an eye or detector. For an illustrative 1000 mm telescope focal length, a 20 mm eyepiece gives 50× and a 10 mm eyepiece gives 100×. The clear aperture and wavelength have not changed, so the diffraction angle remains the same. The larger display may help perception, but it does not create new optical spatial frequencies.

Convert an angle to a physical separation only with distance

For a small angular separation θ in radians at distance L, the corresponding transverse separation is approximately s=Lθ. At an illustrative 384400 km, one arcsecond corresponds to about 1.864 km. This chosen distance resembles a familiar lunar scale but is not a dated Moon-distance prediction.

The calculation translates an angle into a length; it does not guarantee that a lunar feature of that size is recognisable. Illumination, projection, contrast and the instrument's actual image quality remain relevant. Also use the object's distance, not the telescope's focal length, when converting sky angle to physical size.

Make an auditable resolution estimate

1. State the clear-aperture model and the chosen resolution criterion.
2. Record aperture diameter and wavelength in compatible units.
3. Calculate radians, then convert to arcseconds.
4. Separate the Rayleigh radius, Airy diameter and other image widths.
5. Record effective focal length and pixel pitch if using a camera.
6. Check sampling and identify atmosphere, tracking and contrast limitations.
7. Report a model estimate separately from a measured observing result.

This sequence prevents a familiar mix-up: an impressive magnification, small pixel pitch or long focal length can describe an image scale while saying little about the angular information delivered by the aperture under the night's actual conditions.

Frequently asked questions