Building a Flat-Frame Light Box for an SV80S

I’ve been gradually working my way through the many things to learn in astro-imaging. One of the things I encountered in my readings was the concept of using “flat frames” to counteract problems in the camera (dust motes, field vignetting, etc.). It seemed complicated and I wasn’t convinced it was worth the effort.

Well, once I reached the point of actually taking some test exposures, it became clear that it is a necessary technique. Even after a careful cleaning, my camera is showing dust specks and moisture blooms that need to be removed from the images. And, there is a noticeable gradient of brightness from the centre out to the edges that needs to be balanced. So, flat frames it is. I needed a way to photograph a flat, very uniformly-lit neutral surface. While I read of several photographers who have successfully used the twilight sky as a flat frame, most produce their flats using some kind of a “light box” — an artificial target that fits over the telescope and contains a uniformly-lit surface.

I read numerous online articles from others who have built light boxes. The articles that appealed to me the most were

From these, I made a basic plan of what I planned to build.

The idea was to build an oblong box that would fit over the dust shield on my SV80S OTA. At the other end would be a light source and, inside the box, some kind of diffuser surfaces to spread out the light to an even field.

What followed was what I’d call “seat of the pants design” — this was not thought out carefully in advance. Rather, it was a series of little steps that combined what materials and tools I could readily find with experiments to see what seemed to work. I think it worked out fairly well, and I offer this photo-journal of the construction to encourage others to try this simple and useful project.

I had in mind an oblong box, width sufficient to slip over the dew shield on my SV80s, and long enough to cover the shield sufficiently for stability, and with enough interior space for diffusers for the light source.

At the left are two panels with holes large enough for the OTA. The two panels are to provide support — I’d place them far enough apart that they would hold the box without sagging. Then I planned a third panel with a slightly smaller hole; the OTA could “butt up” against this panel to ensure that the box doesn’t slide down too far. At the far end of the box would be a light source of some kind, and between the light source and the OTA, a couple of diffuser panels of some kind. I thought that adjusting the position of the two diffuser panels would allow me to adjust the light to a suitable flat field.

I decided to use 5mm (3/16″) foam board for construction, because it is stiff but very light, and because a local craft store has a ready supply.

I used two 600mm by 914mm (2-by-3-foot) sheets, costing about $6.

The outside diameter of the SV80S dew shield measures 111mm (4-3/8″), so I decided to use

  • 178mm (7″) square panels as the basic construction unit. This allows a good amount of room for holes, including diffusers somewhat larger than the OTA diameter.
  • 113mm (4-7/16″) holes to pass over the OTA;
  • 104mm (4-1/8″) hole in the depth-stop panel — larger than the glass diameter, but smaller than the physical diameter, of the OTA.
I cut 6 of the 178mm squares out of foam board, using a very sharp exacto-knife, and centre-marked each.
To make the centre mark available from both sides, I poked a pin through the centre. (This because the blade on my circle cutter isn’t long enough to cut through the foam board without cutting from both sides.)
A circle cutter from a local craft shop was an invaluable investment. It was about $14, and made cutting neat, clean circles easy.
I cut circles in two boards large enough to pass the OTA snugly,
and verified that they just fit over the dew shield.
A third board has a slightly smaller circle that doesn’t fit over the OTA, and will act as a depth stop.
Next, I worked on how to put a light in the other end of the light box.
I found these nice little Sylvania LED lights at Canadian Tire. You push the lens to light the 3 bright white LEDs.
I cut a circle exactly the size to allow the light to pass through,
and stuck a round disk of plastic (from a bulk pack of CDs) to the back of the light with Velcro.
The plastic disk supports the light, stopping it from falling all the way through the hole. A little Velcro tape on the foam board and the plastic disc holds the light securely in place.
Next, I wanted to be able to switch the light on and off without access to push on the lens.
Opening the back shows there is no room at all to sneak a little switch in. (Those are 3 AAA batteries.)
I found a battery case at Radio Shack ($4) that holds 4 AAA batteries and has an on-off switch.
I moved the 3 batteries to this case, and put a steel screw in place of the 4th battery, changing this to a working 3-battery holder.
A volt meter verifies this works — 3 batteries is 4.6 volts.
And a quick test showed which leads in the light’s battery compartment were “live”, lighting the lamp when fed power from the battery case.
I drilled tiny holes in those two leads,
and soldered the wires from the battery case to them (passing them through a hole drilled in the back of the light case).
Putting everything together with some more velcro, the result is a nice compact unit where the light can be switched on and off from behind.
Note I had to cut a little slot in the plastic disk to make room for the wire.
I made two diffuser panels by taping sheets of onion-skin tracing paper to each side of boards with 127mm (5″) holes in them.
You can see slight diffusion with the diffuser sitting right on the light.
And diffusion improves as the panel is moved away from the light.
I discovered that one diffuser close to the light and a second farther away gave a very uniform diffused panel.
To more formally evaluate and measure this, I stood the light panel and the two diffusers upright with alligator clips, and adjusted the distances until I was happy with the flat light on the second diffuser.
This shot taken without flash shows the good even lighting of the second diffuser panel.
Then I measured the distances that had worked.
This allowed me to fill in all the dimensions on my original rough sketch. (Click for an enlargement of this plan.)
I cut side, top, and bottom panels, and laid out the position of the internal baffles.
Remember that two of the panels must be wider than the other two, by twice the thickness of the foam board.
I then did a test assembly using tiny strips of masking tape to hold pieces in place.
You can see the diffusion working,
and a good even field on the second diffuser panel.
The light is suspended well on the rear,
and the whole thing fits OK on the OTA.
Then I took it all apart and assembled it permanently using a hot glue gun.
Light end, showing the battery and switch box and the velcro that holds the plastic disc to the box.
OTA end. There is a second hole this size inside, and both slip over the OTA to provide 2-point bracing.
Here you can see the two OTA holes, the third “depth stop” hole, and the diffuser panel.

OTA view of the illuminated diffuser.

Note: The light looks very blue here, but it isn’t — that’s just poor white balance in these low-light photos taken without flash to see the illumination. It also doesn’t matter – the objective is to have a uniformly-illuminated surface – the colour of the surface doesn’t matter.

And a test image, on scope. shows a good even field (and some dust spots on the CCD to be cleaned.)

Total cost was about $30.

How well does it work? Here’s an image of M13 without and with flats produced by this box.
(And, to be fair, some additional processing of the “with flats” version, since I lost the image that should go here and have to use a processed one).


  1. LEDs tend to have a pretty peaky light spectrum. Does that affect the calibration of the camera? Or is flatness of the field more important than R,G,and B being the same level?

    1. No, the colour doesn’t matter – as long as it’s nicely inside the range that your camera can detect. Flat fields are about luminance, not colour, information. So, as you say, only the flatness is important, not the RGB balance.

  2. Thanks for sharing your design. I built one using your plans for my Orion ED80 works perfectly!

  3. I used your plan design as a basis for a box to fit my 8″ Newtonian. It works well and was not expesnive to make. Thank you for making your design available.

Leave a comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.