Wavelength is the main limiting factor in resolution because light bends when it passes through an opening, and shorter wavelengths bend less. This bending, called diffraction, creates a blur around every point in an image. No matter how perfect your lens or mirror is, you cannot resolve any detail smaller than roughly half the wavelength of light you’re using. For visible light, that puts a hard floor at about 200 nanometers.
How Diffraction Creates the Limit
When light passes through any opening, whether it’s a microscope lens, a telescope mirror, or the pupil of your eye, it doesn’t travel in perfectly straight lines. It spreads out and creates a circular pattern of bright and dark rings called an Airy disk. Every single point of light in your image becomes one of these tiny bullseyes instead of a crisp dot.
If two objects are close together, their Airy disks overlap. Once the bright center of one pattern lands on top of the other, the two objects blur into what looks like a single blob. The Rayleigh criterion defines the exact threshold: two points are “just resolved” when the central bright spot of one falls on the first dark ring of the other. Below that separation, they merge.
The angular size of each Airy disk is proportional to the wavelength divided by the diameter of the opening. The formula is simple: the minimum resolvable angle equals 1.22 times the wavelength divided by the aperture diameter. This is why wavelength sits at the core of the problem. You can build a bigger lens or mirror to shrink the blur, but wavelength sets the scale of the diffraction pattern itself. A bigger aperture helps, but it’s working against a constraint that wavelength defines.
Why Aperture Alone Can’t Fix It
It’s natural to wonder why you can’t just make the lens bigger and solve the problem entirely. Increasing the aperture does improve resolution, but the improvement is always relative to the wavelength. Ernst Abbe formalized this in the 1870s with his diffraction limit for microscopes: the smallest resolvable feature equals the wavelength divided by twice the numerical aperture. Numerical aperture combines the lens’s opening angle with the refractive index of the medium between the lens and the specimen (air, water, or oil).
In practice, numerical aperture has a ceiling. The best oil-immersion microscope objectives reach a numerical aperture of about 1.4 to 1.5. Plug in the shortest visible wavelength (around 400 nm, which is violet light) and the Abbe formula gives you roughly 200 nm. That’s it. You could engineer the most flawless glass optics ever made, and you’d still hit that 200 nm wall because diffraction is not an engineering flaw. It’s a property of waves.
Telescopes face the same physics from the opposite direction. Instead of tiny specimens, they’re resolving distant objects. A telescope’s resolving power in arc seconds is proportional to the wavelength divided by the mirror diameter. For a wavelength of 500 nm, the resolving power in arc seconds works out to roughly 12 divided by the mirror diameter in centimeters. A 100 cm telescope resolves about 0.12 arc seconds. Doubling the mirror helps, but switching to a shorter wavelength would accomplish the same improvement with half the hardware.
The Limit in Everyday Vision
Your eyes obey the same rules. With a pupil diameter of about 5 mm and a mid-spectrum wavelength of 500 nm, the Rayleigh criterion predicts a diffraction-limited angular resolution of roughly 1 arc minute. That aligns closely with the 20/20 vision standard established by Snellen, which defines normal acuity as resolving detail at 1 arc minute.
Recent research published in Nature Communications found that foveal resolution can actually reach about 94 pixels per degree for black-and-white patterns, somewhat higher than the long-assumed 60 pixels per degree. But the study also confirmed that resolution drops substantially for colored patterns: 89 pixels per degree for red-green and just 53 for yellow-violet. The resolution of the eye depends on a combination of diffraction at the pupil, optical aberrations, and the spacing of photoreceptors and retinal cells. Diffraction sets the upper bound, and everything else can only make it worse.
How Chipmakers Fight the Wavelength Wall
Semiconductor manufacturing is perhaps the most dramatic example of wavelength as a limiting factor, and of the extraordinary lengths industries will go to work around it. To print transistor features on a chip, manufacturers project light through a stencil-like mask onto a silicon wafer. The smallest features they can print are governed by the same diffraction physics.
For decades, the industry used deep ultraviolet light at 193 nm. When that wavelength couldn’t keep up with shrinking transistor sizes, engineers switched to extreme ultraviolet (EUV) lithography, which uses light at just 13.5 nm. That jump to a much shorter wavelength was essential for printing features at modern scales. The latest generation, called High-NA EUV lithography, combines the 13.5 nm wavelength with a larger numerical aperture to achieve optical resolution below 10 nm, packing roughly three times more structures onto the same chip area compared to earlier EUV systems.
The cost and complexity of this transition illustrate the point perfectly. EUV machines are among the most expensive and intricate devices ever built. The industry didn’t switch to them by choice. It switched because wavelength made it physically impossible to go smaller with the old light source.
Techniques That Get Around the Limit
Starting in the late 1990s and early 2000s, physicists developed microscopy methods that sidestep the classical diffraction limit entirely. These super-resolution techniques don’t violate the physics of diffraction. Instead, they use clever tricks to extract information that diffraction would normally hide.
One approach, stimulated emission depletion (STED) microscopy, uses a second laser beam shaped like a donut to switch off fluorescent molecules around the edges of the focal spot, leaving only a tiny central point glowing. This shrinks the effective spot size well below the diffraction limit. Another family of methods, called single-molecule localization microscopy, works by activating only a few fluorescent molecules at a time, pinpointing each one’s position with high precision, and building up an image from thousands of these individual localizations over time. The most popular variant, dSTORM, works with conventional fluorescent dyes rather than requiring specialized labels.
Structured illumination microscopy takes a different path, using patterned light to extract fine details through computational processing. It roughly doubles the resolution compared to conventional microscopy. Newer hybrid methods like MINFLUX combine single-molecule localization with other illumination strategies to push localization precision even further.
These techniques earned the 2014 Nobel Prize in Chemistry. They confirmed something important: the diffraction limit is real and fundamental, but it can be circumvented by reframing the problem. Instead of trying to image finer details with shorter wavelengths, these methods use the photophysical properties of fluorescent molecules to encode spatial information in other ways. The wavelength limit remains the default boundary for any straightforward imaging system. Getting past it requires abandoning straightforward imaging.
Why Wavelength Wins Over Other Factors
Optical systems have many imperfections: lens aberrations, misalignment, vibration, atmospheric turbulence for telescopes, and noise in detectors. All of these degrade image quality. But they are, in principle, fixable. You can grind a better lens, stabilize a platform, or use adaptive optics to correct for atmospheric distortion. Once you’ve eliminated every correctable flaw, you’re left with diffraction, and diffraction is set by wavelength.
This is what makes wavelength the “main” limiting factor rather than just one factor among many. It represents the theoretical best case. Every other source of blur sits on top of it. A system where all other errors have been minimized is called “diffraction-limited,” and that term itself reveals the hierarchy: wavelength is the floor, and everything else is overhead.