The Rayleigh criterion is a conventional rule in optics for deciding when two nearby point sources are just resolved by an imaging instrument. It states that two equally bright, mutually incoherent point sources are just resolved when the central maximum of one source’s diffraction pattern coincides with the first minimum of the other’s. For an ideal, uniformly illuminated circular aperture, this gives an angular separation of approximately , where is the wavelength and is the aperture diameter. The criterion is widely used to characterize telescopes, microscopes, and other optical instruments, but it is not a universal boundary between distinguishable and indistinguishable objects.
Physical basis
A finite aperture does not form a perfect point image of a point source. Instead, the wave nature of light produces a spatially extended intensity distribution called the point-spread function. For a clear circular aperture in an ideal diffraction-limited system, this distribution consists of a bright central region surrounded by progressively weaker rings. The central region, bounded by the first dark ring, is the Airy disk.
Two nearby point sources therefore produce overlapping patterns. When the sources are mutually incoherent, their intensities add after time averaging; their optical fields do not produce a stable cross-interference term. At large separations, two intensity peaks are readily visible. As their separation decreases, the peaks overlap more strongly and the intensity dip between them becomes shallower.
The Rayleigh criterion chooses one particular overlap geometry as the threshold for “just resolved.” For two equal-brightness sources with ideal circular-aperture patterns, the intensity at the midpoint is about 26.5% below the two peaks at this separation. This provides a recognizable contrast dip, but the choice is conventional rather than a mathematical discontinuity in the image.
The assumption of mutual incoherence matters. For coherent sources, interference depends on their relative phase, and the resulting intensity distribution need not satisfy the same resolution rule.
Mathematical formulation
Circular apertures
For monochromatic light passing through an unobstructed circular aperture of diameter , the far-field intensity pattern is
where is the first-order Bessel function of the first kind, is the angle from the optical axis, and the expression at is understood by continuity.
The first positive zero of occurs at approximately . Consequently, the first dark ring satisfies
For small angles,
The angular separation is expressed in radians. The wavelength is that in the propagation medium; for astronomical observations in air or vacuum, the distinction is usually negligible.
For an image formed at focal length , the corresponding center-to-center image separation is approximately
This is the radius of the first dark ring of either individual Airy pattern, not its diameter. Confusing these quantities introduces a factor-of-two error.
Microscopy
In object space, the conventional lateral Rayleigh separation for a circular objective pupil is
where is the vacuum wavelength and numerical aperture is
Here is the refractive index of the medium between the specimen and objective, and is the objective’s collection half-angle.
For example, with and , the formula gives approximately . This is an idealized lateral two-point separation, not a universal minimum detectable object size. Illumination conditions, aberrations, polarization, and the imaging method can change the relevant resolution behavior. At very high numerical aperture, a full vector treatment of light may be necessary.
Other aperture shapes
The factor 1.22 is specific to a clear circular aperture. For a uniformly illuminated single slit of width , the first diffraction minimum satisfies
so the analogous small-angle Rayleigh separation is approximately .
Rectangular pupils, centrally obstructed telescope apertures, and nonuniform pupil illumination produce different diffraction patterns. Their resolution characteristics cannot always be summarized by the circular-aperture coefficient.
History and related criteria
The criterion is named after Lord Rayleigh, John William Strutt, whose nineteenth-century work helped establish the wave-optical analysis of resolving power. It became a standard practical benchmark because it connects the visible separation of image peaks with the diffraction pattern of a known aperture.
Several other resolution conventions address different questions:
- Abbe’s diffraction limit concerns the transmission of spatial detail, especially periodic structures, through an optical system. A familiar expression is , though its application depends on illumination and imaging conditions.
- The Sparrow criterion places the threshold where the central dip between two equal source images just disappears. It generally permits a smaller separation than the Rayleigh criterion for the same idealized patterns.
- Dawes’ limit is an empirical rule developed for visually resolving approximately equal-brightness double stars, rather than a criterion defined by the first diffraction minimum.
These values need not agree because “resolution” can mean visible separation of two sources, detection of periodic contrast, or estimation of source positions.
Applications
Astronomy
For a telescope with aperture observing at ,
This is an ideal diffraction-based value. Actual ground-based images are often limited by atmospheric turbulence, tracking errors, or optical aberrations. Adaptive optics can reduce atmospheric distortions, allowing performance closer to the telescope’s diffraction limit.
In interferometry, angular resolution often scales with wavelength divided by the separation between collecting elements. However, an interferometer’s image response depends on its array geometry and sampling; the circular-aperture Rayleigh formula cannot simply be transferred unchanged.
Spectroscopy
An analogous criterion applies to separating nearby spectral lines. For an ideal diffraction grating, the Rayleigh criterion gives the resolving power
where is the magnitude of the diffraction order and is the number of illuminated grooves. Here the principal maximum of one spectral line coincides with the first minimum of the neighboring line’s instrumental pattern.
In practical spectroscopy, finite slit width, detector sampling, optical imperfections, and intrinsic line broadening may reduce the achievable resolving power.
Limitations and interpretation
The Rayleigh criterion assumes a particular source model and imaging response. Several distinctions are important:
- Unequal brightness: A faint source can be obscured by the diffraction structure of a brighter neighbor even when their separation exceeds the equal-brightness Rayleigh value.
- Noise and sampling: Visibility depends on detected photon counts, background, detector noise, and pixel spacing. Adequate optical resolution does not ensure that the detector records the necessary information.
- Aberrations and pupil structure: Defocus, optical aberrations, obstructions, and scattering alter the point-spread function and its contrast.
- Localization versus separation: The position of one isolated source can often be estimated much more precisely than the width of its image. This does not imply that two unknown adjacent sources are equally easy to resolve.
- Model-dependent inference: Fitting known source models to sufficiently informative data can distinguish sources below the Rayleigh separation. Such estimates depend on signal quality, assumptions, and measurement uncertainty.
Super-resolution microscopy can also achieve spatial detail beyond conventional diffraction-limited imaging by using additional information, controlled emission, nonlinear responses, or localization of separately emitting sources. These methods do not violate diffraction: they change the measurement process or the information available for reconstruction.
Thus, a claim of “sub-Rayleigh resolution” requires a specified task, imaging model, and performance measure. The Rayleigh criterion remains a useful benchmark for conventional two-point imaging, not an absolute limit on all optical measurements.