---
title: "How Ryle built a large radio telescope from smaller ones"
description: "How can a few antennas spread over time reveal the shape of a radio galaxy?"
canonical_url: "https://sandboxphysics.com/stories/astronomy-gravity/ryle-aperture-synthesis/"
language: "en"
page_type: "story"
alternate_languages:
  zh-CN: "https://sandboxphysics.com/zh/stories/astronomy-gravity/ryle-aperture-synthesis/"
last_modified: "2026-10-24"
---

[Learning to read radio signals](https://sandboxphysics.com/stories/astronomy-gravity/#radio)

Astronomy · Cosmology · Gravity

# How Ryle built a large radio telescope from smaller ones

Sandbox Physics · 2026-10-24

How can a few antennas spread over time reveal the shape of a radio galaxy?

1. Observe the same source with an antenna pair
2. Change the projected antenna separation
3. Combine calibrated wave correlations
4. Recover structure across the sky

Antenna separation selects an angular pattern. Many separations and directions supply the pieces needed for an image.

## A source can be bright and still look like a blur

In 1960, Martin Ryle and Antony Hewish described a way to obtain the resolving power of a large radio telescope without filling its whole area with metal. Two antennas would occupy different positions in succession. A calculation would then combine what they recorded. Their proposal grew out of a practical frustration: the radio sky contained bright objects whose shapes an ordinary antenna could not separate.

A radio galaxy can send us abundant radiation and still be difficult to image. Radio wavelengths are much longer than visible wavelengths. For an opening of the same size, they spread into a much wider blur. Making a receiver more sensitive helps it detect fainter objects, but does little to distinguish two bright objects that fall inside the same blur.

The usual answer is a larger opening. Ryle’s question was how much of that opening actually needed to exist at once. The answer depended on measuring the arriving waves carefully enough that observations from separate days could later be combined.

Sources: [1] [4]

## An antenna pair measures one pattern on the sky

Imagine a distant source sending nearly parallel wavefronts across a field. One antenna receives a crest slightly before the other. The difference depends on the direction of the source and the separation of the antennas. Correcting that delay lets the two receivers compare the same arriving signal.

The separation is called a baseline. The electronics multiply the two received voltages and average the result. This correlation keeps the part of the wave that the antennas share. By retaining both the strength and relative timing of that shared signal, the instrument measures a quantity called visibility.

Two sources in different directions give different delays. Their contributions can reinforce or partly cancel in the correlation. A long baseline is sensitive to finer angular structure than a short one. Rotate the baseline and it responds to structure along a different direction. Each measurement therefore supplies one piece of information about how brightness is distributed.

For two nearby sources, a baseline along their separation can distinguish them much more strongly than one viewed across that separation. A second direction supplies information the first measurement missed. At a fixed wavelength, doubling the projected separation roughly halves the angular scale to which the pair is most sensitive.

One pair, held at one separation, gives too little information to draw a detailed map. It can still answer a limited question, such as whether a source has a particular angular width. Imaging asks for many such pieces: broad structure, fine structure and structure pointing in different directions.

Sources: [4] [1]

## The turning Earth changes the view between the antennas

The antennas need not physically circle the source. As Earth turns, their fixed separation appears to change direction and length when seen from the source. Following the same object for hours samples a curved sequence of baselines. Moving an antenna to another position on a later day adds another sequence.

The One-Mile Telescope used this combination of Earth rotation and adjustable separation. The 1966 paper by Bruce Elsmore, Sidney Kenderdine and Ryle describes a synthesized aperture one mile across. It reported beam widths of about 80 arcseconds at 408 megahertz and 23 arcseconds at 1407 megahertz. The finer result corresponds to about one hundred and fiftieth of a degree.

The synthesized diameter describes resolving power. The amount of radiation actually collected still depends on the antennas that are present. An array with a large separation can therefore have fine resolution and still need long observations to detect faint extended emission.

This approach depended on a source staying sufficiently steady while its different views were collected. It also needed stable electronics and accurate relative timing. A timing error can look like a shift in the source direction, so calibration becomes part of forming the image.

The method was already producing surveys before the One-Mile Telescope. In 1961, Peter Scott, Ryle and Hewish published results from four sky strips observed at a wavelength of 1.7 metres. They described the observations, calibration and numerical analysis together. The calculation was an essential part of the telescope’s operation.

Sources: [4] [3] [2]

## A reconstructed image carries the marks of the measurements

The mathematical operation connecting these correlations to a small sky image is a Fourier transform. In this setting it combines wave-like brightness patterns of different spacing and direction. Their sum can describe two compact sources, an extended radio lobe or a more intricate arrangement.

Real arrays sample only some of the possible baselines. A preliminary image therefore surrounds a compact source with a characteristic pattern of sidelobes, faint responses caused by the sampling. Astronomers use the known response of the array when estimating the underlying sky. Very extended emission also needs sufficiently short baselines or complementary single-dish measurements.

These details explain why a long maximum baseline is only one property of an imaging instrument. The distribution of separations matters too. A pair of antennas placed far apart can distinguish fine structure along one direction, while a well-sampled array can recover a much richer picture.

The 1974 Nobel Prize recognized Ryle especially for aperture synthesis, alongside Hewish’s role in discovering pulsars. The scientific change was visible in the maps: radio sources became objects with structure that could be compared with optical observations and physical models of galaxies.

Sources: [4] [6] [5]

## Explore the related models

Choose a model to explore the physical idea.

### [See how a baseline responds to a star’s size](https://sandboxphysics.com/labs/stellar-interferometer/)

Change the projected baseline and compare the visibility of a disk and a binary.

The optical model measures a uniform disk or a two-point separation from recorded fringe powers. Radio sky-image reconstruction is the next step described here.

## Sources and reading notes

The 1960 Ryle–Hewish paper supplies the synthesis proposal; the 1961 Scott–Ryle–Hewish survey and the 1966 One-Mile Telescope paper show its observational use. NRAO explains the modern correlation and imaging language.

1. [M. Ryle and A. Hewish, The Synthesis of Large Radio Telescopes (1960), MNRAS 120, 220–230](https://doi.org/10.1093/mnras/120.3.220)

   The publisher abstract and the ADS-hosted original-paper text describe movable antenna pairs and mathematical synthesis. Publication was March 1960; receipt was August 1959. [Read the hosted original](https://adsabs.harvard.edu/pdf/1960MNRAS.120..220R)

2. [P. F. Scott, M. Ryle and A. Hewish, First Results of Radio Star Observations Using the Method of Aperture Synthesis (1961)](https://doi.org/10.1093/mnras/122.1.95)

   The publisher abstract specifies four surveys at 1.7-m wavelength and identifies calibration, reduction and source analysis as parts of the work. [Read the hosted original](https://adsabs.harvard.edu/pdf/1961MNRAS.122...95S)

3. [B. Elsmore, S. Kenderdine and M. Ryle, The Operation of the Cambridge One-Mile Diameter Radio Telescope (1966)](https://doi.org/10.1093/mnras/134.1.87)

   The original-paper abstract reports 80- and 23-arcsecond beams at 408 and 1407 MHz; its introduction describes Earth-rotation synthesis over a twelve-hour observation. [Read the hosted original](https://adsabs.harvard.edu/pdf/1966MNRAS.134...87E)

4. [J. J. Condon and S. M. Ransom, Essential Radio Astronomy, chapter 3, §§3.7.1–3.7.6](https://www.cv.nrao.edu/~sransom/web/Ch3.html)

   Sections on voltage correlation, complex visibility, projected baselines, Earth rotation and sensitivity supply the modern explanation of what antenna pairs measure.

5. [Cavendish Laboratory, Mullard Radio Astronomy Observatory at Lord’s Bridge](https://www.phy.cam.ac.uk/about/our-history/mullard-radio-astronomy-observatory-at-lords-bridge/)

   The observatory’s own history identifies the arrays, the move from detection to imaging, Bell Burnell’s 1967 observation and the two parts of the 1974 citation.

6. [Nobel Prize in Physics 1974, official citation for Martin Ryle and Antony Hewish](https://www.nobelprize.org/prizes/physics/1974/summary/)

   The indexed official citation distinguishes aperture synthesis and the decisive role in discovering pulsars. Full official pages and the year-filtered API returned HTTP 403 in this pass. Cambridge’s institutional history independently confirms the awardees.

[How we learned to read the universe](https://sandboxphysics.com/stories/astronomy-gravity/) [The regular pulses Jocelyn Bell refused to discard](https://sandboxphysics.com/stories/astronomy-gravity/hewish-bell-pulsar/)