Each vane holding the secondary mirror in front of the primary causes a diffraction spike. Couldn't the secondary mirror be held with only one vane?

If stability is the issue, couldn't it be substituted be steering the vane to hold the secondary in place with precision, similar to active optics, regardless of winds? And why does the James Webb Space Telescope use three vanes when it will operate in undisturbed microgravity and shadow?

(Maybe the star spikes are desired 'cause they make stars look nice?)

  • 1
    $\begingroup$ I can't speak specifically to James Web, but i know telescopes like Hubble actually vibrate a lot. Mostly due to temperature gradients. If that vibration were to knock the seconds mirror out of whack seriously, you may never get it back in alignment just right and would forever have poorer image quality. Besides, the spikes don't prevent science being done. Astronomers are well equipped to deal with them $\endgroup$
    – zephyr
    Commented Sep 23, 2017 at 12:55

4 Answers 4


I can't speak for the JWST, but I do work at a company that designs telescopes for spacecraft and I can tell you that a lot of the structure of the telescope is to protect the instruments during the initial ascent to space where the launch vibrations and g-loading will be the worst. Plus mirrors for telescopes that big have really precise alignment tolerances and if the secondary mirror isn't very rigidly attached to the spacecraft even simple microstrains from heat loads or spacecraft pointing can move the mirror out of alignment. Even the fact that the telescope is going from a 1g environment to a 0g environment can move the mirror because it was aligned on Earth when the beam was slightly bent and then it goes back to being straight in microgravity and the mirror is misaligned! It stinks that there is stuff in the way of some beautiful science that you're trying to take a picture of, but it's the only way to make sure the telescope survives up there!

  • $\begingroup$ "but I do work at a company that designs telescopes for spacecraft " Wow, not many do that! This is a pretty impressive community. What about using a movable secondary mirror that could be aligned after launch and deployment? Why does it all have to be fixed on the ground and survive the launch G-forces? $\endgroup$
    – LocalFluff
    Commented Jul 15, 2019 at 16:56
  • $\begingroup$ Haha there's more than you think! The issue is that the displacements like fractions of a human hair so it would be hard to design any precision actuation systems that can achieve that resolution. And if you did have that, then you would need a motor and cables and such that would also weigh down the mirror and make it deflect more! Also it's nice to align the telescope on the ground because then you can use really precise, custom equipment to make sure everything looks good before shipping whereas doing it in orbit would have to be autonomous and leaves more room for failure :( $\endgroup$ Commented Jul 15, 2019 at 17:09
  • $\begingroup$ What do you think about the Star Shade concept? A specially shaped shadow caster thousands of kilometers away that blocks out the light of a certain star. Isn't that MUCH harder than aligning two mirrors within tens of meters of each other where one can say a bit to the right, a bit to the left, and down and... And what about my favorite mission, the telescope launched to 500 AU using the Sun as a gravity lens to finally discover what's behind the Sun? Can you align that? (Let's see now, one meter versus one AU, that's a baffling triangle! That won't fly). $\endgroup$
    – LocalFluff
    Commented Jul 15, 2019 at 21:06
  • $\begingroup$ @LocalFluff Such a telescope doesn't need to be sent to 500AU away, any probe left the Earth-Moon system can see, what is behind the Sun. The sad truth is: nothing. (What is not surprising, anything on the other side would have an unstable orbit for us.) $\endgroup$
    – peterh
    Commented Jul 15, 2019 at 21:27
  • $\begingroup$ @LocalFluff Never worked with star shades personally, but they are completely different systems. Tare so far away from the telescope so any error in their position or rotation is minimized because the angle between where it's supposed to be and where it is is still very small. They are separate spacecraft with their own propulsion and attitude systems so they already have stuff to adjust their position. With telescope optics, it's not just that the optic is rotated a bit off, if there are microstrains, they can actually change the optics surface which introduces coma and abberation errors. $\endgroup$ Commented Jul 16, 2019 at 15:06

I'm an amateur telescope maker.

Single vane designs do exist. Their main problem is lack of stability. You would have to use a very thick vane to hold the secondary in place in a stable way. That would block off a significant amount of light and may impact the performance of the instrument.

Active systems, while theoretically possible, would be very expensive. Many vibrations are quite fast, you'd need the system to track alignment, detect deviations, make decisions very quickly and apply corrections. Meanwhile, for large telescopes, the mass of the secondary mirror is not insignificant. The actuators themselves would flex and tilt the vane while they're steering the mirror around. This is not a trivial job at all.

It is much simpler to just use multiple vanes.

One lesser-known fact: each straight vane actually makes 2 spikes, diametrically opposed. So even the single vane system would still make 2 spikes.

A 3 vane system makes 6 spikes, and a 4 vane system makes 8 spikes.

So why the 4 vane systems appear to only make 4 spikes? Because they coincide two by two. Within each pair of opposed vanes, each vane is carefully aligned with its opposed partner. This will merge the spikes two by two and will make the overall figure simpler.

You can actually make the spikes disappear. Each spike is perpendicular to the vane causing it. It's light being diffracted at the edge, going sideways from the edge of the vane. Each point on the edge diffracts light - the spike is the result of all points on the edge each diffracting a little bit of light.

So what happens if you curve the vane? The spike "spreads out", like when you open a fan. It's not a sharp line anymore, it's a whole area.

If the vane is curved into exactly half a circle, then the spike is maximally washed out - it becomes a full 360 degree figure. In practice, when it's spread out that much, the spike becomes invisible to the eye of the observer. Telescopes with curved vanes are said to have no spikes.

But there's no free lunch. Now that you've spread out the spikes, that diffracted light is still there. What happens is that the field of view becomes overall slightly more "foggy", as in a fog of light. The effect is very slight, but it matters for critical observations.

Finally, you can always install a flat glass plate at the top of the instrument and glue the secondary mirror support to it. This will definitely eliminate all diffraction.

But it's hard to manufacture a large, optically flat plate like that. It will be thick and heavy. This will not scale beyond the size of small reflector telescopes. There are also issues with cooling, which becomes slower and more complex, because the telescope is now closed at the top.

SCT telescopes that use a corrector plate (not flat) do this anyway. Same with Maksutov-Cassegrain systems (MCT). The plate is there to correct the instrument, you might as well glue the secondary to it.

(Actually, I lied. The SCT and MCT designs are done on purpose so that the location of the corrector plate, and the location of the secondary mirror, coincide. It's just easier to build it that way in practice. But in theory the plate and the mirror do not have to coincide.)

So the main choices are:

  1. Three or four straight vanes. Maximum stability. You get 6 or 4 (nominally 6 or 8) spikes, but the image in between spikes is clean.

  2. One curved, half-circle vane. Sometimes two opposing curved vanes, like an X with arms curved up/down into two half-circles, or like the number 8 with the top and bottom caps lopped off (this is more stable but generates more diffraction). No visible spikes. Somewhat reduced contrast because the spikes have been smeared all over the image.

  3. One single, thick vane. Either not very stable, or reduces the amount of light captured by the instrument.

  4. No vanes, just a glass plate. Expensive, heavy, doesn't scale. But if the plate is not flat you could use it to apply corrections to the instrument, so then its existence is justified.

Choice #1 is the most popular for professional and amateur reflector telescopes. #2 is used sometimes by people who object to spikes on esthetic grounds. #3 is rarely seen in practice. #4 is used with corrected SCT systems, MCT, and the like, because the plate is already there by design.


I cannot answer about the JWST. Regarding amateur telescopes, there is the option of using curved supports, which are intended to eliminate diffraction spikes and reduce overall diffraction. http://www.fpi-protostar.com/crvmnts.htm

Athough diffraction spikes are eliminated by curved supports, diffraction occurs symmetrically around bright objects.

  • 1
    $\begingroup$ Good find (and by “#2” I believe you mean the second entry on their FAQ page). However, your answer as it stands is a “link-only answer”, and those are frowned upon – in particular because the linked page might go down at any time, and because the reader needs to look through a lot of text to find the actual answer (see Are answers that just contain links elsewhere really good answers? for more). $\endgroup$
    – chirlu
    Commented Sep 23, 2017 at 13:39
  • 1
    $\begingroup$ Links can (and do) break, at which point your answer will become useless. @chirlu is right. Can you take some time and explain at least the key points of your answer? $\endgroup$
    – uhoh
    Commented Sep 23, 2017 at 13:52
  • $\begingroup$ Yes, the curvy ones do help. $\endgroup$ Commented Sep 23, 2017 at 14:15

tl;dr: Using one vane may not provide any optical advantage over three vanes since the primary mirror is already so highly segmented and "edgy", and it certainly provides a lot of mechanical disadvantages in terms of stiffness against lateral vibration translation and breaking of symmetry during thermal expansion!

note: I'll be updating this answer soon, once I get access to some SPIE papers.


Spacecraft, even space telescopes contain many active sources of vibration. Pumps, valves, refrigerators, thrusters, reaction wheels, actuators (JWST has a movable solar reflector for trimming radiation pressure) movable antennas and solar panels (Hubble at least, not sure about JWST, will check...) and even the science payloads may have movable mirrors or focusing mechanisms.

A three-vane mount provides substantially more stiffness against in-plane movement, since flexure of one vane would result in tension and compression in the other two vanes.

For a single vane; you have a mass on the end of a stick. My first telescope was an Edmund 4¼ inch Newtonian with a big glass right-angle prism on the end of a metal rod. You could tap the tube near the top and that thing would ring, making the image a blur.

Three vanes vs one will also help in axial vibrations (toward/away from the primary) but not as much as transverse, but small transverse oscillations are much worse than small axial vibrations because the secondary is curved. Move it sideways and the image will slide back and forth across the image sensors, blurring across the pixels. Tens of microns of lateral motion of the image will kill the resolution.

However, tens of microns of axial motion would hardly make a difference in focus in a long focal length and high f/no. system.

Symmetry under thermal expansion

With the huge multi-layer Sun shield JWST shows a major effort to keep the temperature of all of the optics and detector system very low so that it does not contribute thermal infrared radiation and so the sensors will have low thermal noise.

However you would also like to keep the large structural elements as thermally stable as possible to minimize drift in focus and position of the image. A three-vane spider offers some amount of lateral symmetry, and if there are temperature changes it might reduce the amount of lateral drift, compared to a single vane where expansion-contraction obviously would just cause the secondary to move back and forth along the vane's axis.

Diffraction, PSF and Hubble

The Hubble Space Telescope does suffer from diffraction effects from its four vanes as well as from its giant secondary and three mirror through-holes for mounting, but a major effort in management of the HST's point spread function (e.g. "optical modeling using Tiny Tim") seems to handle this well.

From this answer in Photography SE:

(see also) @scottbb's (excellent) answer... For scientific analysis of images from Hubble, one has to account for all aberrations, diffraction, and pixilation. There is a nice writeup on this: 20 years of Hubble Space Telescope optical modeling using Tiny Tim.

enter image description here

enter image description here

enter image description here

Vane number

It's important to note that WFIRST went the other way, and had six vanes!

From Introduction to WebbPSF for WFIRST:

enter image description here

JWST's Point Spread Function - don't forget the segmented primary!

The following image is taken out of context (at the moment, I'll retrieve it later today) from the paywalled 2014 SPIE paper Updated point spread function simulations for JWST with WebbPSF. Considering that the primary mirror presents much more total "diffracting edge" transverse to the incoming wavefront than the secondary's three vanes, it's possible or even likely that the spider's contribution even with three vanes is small compared to that from the segmenting of the primary.

enter image description here

From webbpsf Documentation Release 0.8.0 (click for larger size)

enter image description here

Fig. 1: Sample PSFs for JWST’s instrument suite, all on the same angular scale and display stretch.


In other words, using one vane may not provide any advantage over three vanes since the primary mirror is already so highly segmented and "edgy", and it certainly provides a lot of mechanical disadvantages in terms of stiffness against lateral vibration translation and breaking of symmetry during thermal expansion!


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .