What does it mean for Webb's cameras to be working well?

With the upcoming James Webb launch, I thought it would be helpful to talk about how space cameras work.

Well, let's be honest. At this point my brand is to add to your nightmares about the JWST launch by sharing mine. Let's assume the launch and all the deployments appear to go well. Then, the real fun begins. Do the cameras actually return good images?

Let's start with a simple case: we'll take a picture of a star. If your camera took perfectly sharp images it'd look like this.

We can represent this in a cross section like this - the vertical direction represents brightness.

But any real camera blurs the image somewhat, like this

so the cross section looks something like the image below. The technical name for this blurring (spreading) of a perfectly sharp point is the `point spread function'.

Now let's suppose we look at two bright sources A, B and one faint one C, where a perfect image would have the following profile:

If the camera is working well, you get a sharp image with only a small blur:

But if the mirrors are misaligned or misshaped, you can get a blurry image. What you'll see is the black line which is the sum of the three sources. The blue and green sources are merged together and the faint red source is totally lost because it's swamped by the blurred edges (`wings') of the other sources.

In a slightly less severe version of this you still get a nice sharp core but it's only a small fraction of the light, most of it is blurred. This is what went wrong with the Hubble Space Telescope. Still can do science with sharp bright stars, but faint fuzzy stuff is hopeless. Thank goodness the servicing missions found a way around it.

Another big problem is background. One issue is `stray light' - light that gets in the camera from the wrong direction and adds a glow across the field of view. This reduces the contrast for the faint source (red) and it won't be as easily detectable. This is a problem that the Gaia spacecraft ran into a few years ago.

Another kind of background is instrumental `noise' caused by things like radiation affecting the camera. That's really going to wipe out the faint sources. Quite a few space telescopes have had unexpectedly high background that reduced their capabilities.

Another issue is stability: if the source appears in a different place tomorrow, did it move, or did the camera change? If it appears twice as bright tomorrow, did it really get brighter, or did the camera change?

The spectroscopic instruments have all these issues and some additional ones. I think of it as like doing a really precise color calibration for a high end monitor used by people making CGI for Hollywood.

Hopefully this gives you some idea of how hard it is to get a telescope working right. If Webb can get the deployments all done right, then there will be an extended period when the team is checking out the instruments and we'll be paying close attention to words like 'point spread function' and 'background'. But this has all been well tested on the ground and it SHOULD be fine. Fingers crossed...