How flat is your lightbox?

[pvelez (Pete)]

I've posted recently about my travails with flats. I had recurrent blue background around the edges of my images. I had multiple diagnoses - dodgy filters, light pollution, poor calibration routines etc etc.

In the end, I concluded the issue was low SNR for my blue filter resulted in excessive stretching in the R and G which meant I was actually highlighting the noise in the background.

Still need to test this - I'm waiting for the Sydney weather to play nicely.

Anyway, just for fun, I thought I'd test my filters by taking some images with my spectroscope. I have a LISA (a great bit of kit) so I decided to take some images of the sun as it came through the skylight in my study using R, G and B filters. This was just a test really.

I then took images through the same filters using my lightbox as my source.

The transmission curves should be broadly similar as for the sunlight - ignoring absorption lines in the sun's atmosphere and the earth's atmosphere. I expected 3 inverted Us crossing over midway up the curve so that the combined signal of overlapping signals at any point in the curve summed to the peak intensity at the top of the 3 Us.

What I found was that in B, my lightbox has a relatively sharp cutoff in the range 4600 - 5050 Angstroms.

The first image illustrates this. I've superimposed the lightbox curve on the curve for the diffuse sunlight.

The second image shows the same pair as well as the curve for the lightbox for G and R - R is to the extreme right.

The last shows the curve for images of the lightbox taken through B, G and R filters (left to right).

I haven't corrected these with flats or darks nor taken account of the response of the CCD so I know the data isn't necessarily accurate in absolute terms. But it does show the difference between the lightbox and solar illumination.

I don't think this really impacts on the performance of the lightbox - though I'm happy to stand corrected. I guess its a function of the type of diodes used in the lightbox.

Pete

[pvelez (Pete)]

In case you are interested, here is a profile of an image taken through a basic Ha filter using my lightbox. Superimposed is the profile of the lightbox using the R filter. The Ha line is marked in lime green.

The LEDs provide plenty of Ha. This also gives a good idea of how narrow the Ha filter is.

Pete

[pvelez (Pete)]

Looks like there is a drop off in R too

Here is a profile of an image taken of the lightbox through an R filter compared with an image of sunlight through the same filter.

Pete

[Exfso (Peter)]

Peter, not being too technically minded about this, but I would suggest you will find differences with any artificial light source. Obviously nothing is quite as efficient as good old natural sunlight. When I first started building these, I considered using the old standard light bulb, but they tend to give off a rather orange glow and their life was far too short, this is why I opted for bright white LED's. The odd reading may well be a function of the diffuser material I am using as well, this is an opal shade opaque 4mm perspex.
As you say I don't believe that there is any adverse result in the function of the light box. Thus far I have built around 100 of these and as far as I can tell by the lack of negative feedback, everyone is happy with the results that they are getting.

[pvelez (Pete)]

Peter

I hadn't thought about the diffuser material. I also didn't rig this up to my scope, just had it sitting neatly underneath.

Looking at other artificial lights, the LEDs work a treat. Looking at standard neon lights for example, all the LISA reveals is a series of emission lines.

I'm very happy with my box. It's the one constant in all my fiddling with calibration. I recently had a play with dusk flats - both regular and with a tshirt and the results weren't up to your light box. Frankly, you are doing us all a service turning these out at a very reasonable price.

I agree with you - I can't see how LRGB imaging is adversely affected if the light is not flat across all wavelengths.

Pete

[RickS (Rick)]

Interesting and thought provoking data, Pete! I'm also very happy with my pair of Exfso light boxes

Cheers,
Rick.

[pvelez (Pete)]

You're right there Rick.

Its not a test of the lightbox as they work a treat.

I initially set out to look at the filters and was surprised by the drop off in B. I did a paper for my SAO course a couple of years back on flats and recall the importance for photometry of having a light source that matched the SED (spectral energy density from memory) ie it was flat across all the relevant wavelengths. But I suspect that was for serious science for very high precision work.

Perhaps the more relevant point to make is that the range of wavelengths is so even across most of the visible spectrum.

I have an electro-thingy panel too. If i have the time, I might do a comparison.

Pete

[RickS (Rick)]

I have a large electroluminescent panel on the way from Gerd Neumann. A light box for my AG12 was going to be a bit too cumbersome. Will be interesting to see how it compares...