I'm new to AAVSO. I've been doing astrophotography for about 2 years now, but have been an amateur visual astronomer for a long time. I want to start using my equipment for variable star photometry in addition to deep sky imaging.
I read the AAVSO DSLR observing manual (v 1.4) and I understand most of it. I was following instructions in the appendices to characterize my equipment. When I did my flats test (Testing Flats for Uniform Illumination) in Appendix B, I was shocked at how bad my results were. Nowhere near the even 1% variation along the diagonal. More like 10%.
Is this a problem with field curvature, sensor tilt, or is it a problem with my flat panel? I use one of those LED artists drawing light pads and two tee shirts. Is this saying that the light panel isn't uniform? To answer this, would I take ratios of flats made by rotating the flat panel 90 degrees instead of the camera?
I've been using this for astrophotography. My worry is that my equipment is not suitable for variable star photometry unless I figure out the cause.
Thanks for any recommendations,
I just sent you an email.
In my opinion, the section on testing flats in the AAVSO DSLR observing manual should refer to symmetrical rather than uniform illumination. The latter implies, as I assume you interpreted, a uniform distribution of signal across the field. In practice, as you have discovered, this is not the case. It is usual that vignetting is clearly apparent, with maximum signal centrally and a gradual decrease towards the edges of the field.
Try this experiment. Do the same thing with a science image, preferably one with plenty of sky background, and plot the line profile of the sky across a diagonal. My guess is you will see the same profile (apart from spikes due to stars along the line) that you saw in your flats.
Hi Bob and Roy,
Appendix B of the DSLR manual is designed to test that the device used to illuminate the camera when taking flats is generating a uniform distribution of light. It isn't concerned with vignetting or other imperfections in the imaging system.
If your light source, telescope, camera and everything else in the light path were perfect the image you record would have the same ADU value in every pixel. But nothing is perfect.
However, let us assume for now that your light source is perfect, but your telescope, camera and everything else in the light path have the usual imperfection. In this case your flats will show vignetting in the corners, dust donuts and small random variations due to noise. The top part of Figure B1 in the DSLR manual show all these.
If you perform the Appendix B test then the image you get after dividing one master flat by the other will show the same "average" ADU value in every pixel but there will be some random noise from pixel to pixel.
However, your light source is very unlikely to be perfect. It may have a gradient from one side to the other, or even splotches of uneven illumination. The Appendix B test will show these as sloping straight lines or wavey lines, respectively. If the variation is too big then you'll need to look at alternative light sources.
Once you have a satisfactory light source the master flats you make will take care of the vignetting and dust donuts from your imaging system. Cheers,
There were two issues in my previous post, which were not properly separated.
The first was the typical vignetting that is seen in most imaging. The vignetting occurs in science images of the sky, and is also present in flats. The second was the testing of the uniformity of the light source for flats.
The following relates only to the second point. My contention is that the testing for the uniformity of illumination of the light source for flats, as described in Appendix B of the AAVSO DSLR observing manual, actually detects any asymmetry within the light source (e.g., one side less bright than the other). The reason is that the test involves rotating the light source (or the camera) prior to taking a second set of flats, then diving one set into the other.
However, if (for example) the centre of the light source is brighter than the periphery, but the central bright area is symmetrically placed, the test as described in the manual will not detect the non-uniformity.
I now understand your concern and I agree with what you are saying. One way to check if your light source has a central bright spot and radial gradient would be to move the light source off centre between making the two master flats. That assumes the light source is significantly larger than the aperture of the lens or telescope.
Appendix B could be updated, but I think the whole DSLR manual should be updated. I'll leave that for someone actively involved in DSLR photometry now. Cheers,