Roger Samworth

Solar Image processing. Myths and Legends.

I have just read an (unattributed) article in April’s Sky at Night magazine on solar imaging, and I have to say, from my own experience I disagree with a lot of it!

It says

“— requires a monochrome high frame rate camera set-up” and “use of a colour camera is inefficient”

Who cares? There is plenty of light from the Sun, efficiency isn’t a problem!

It also says that Ha features change quite quickly. True. As do the atmospheric “cells” that cause image wobble. It suggests you take a 1000-1500 frame avi. The time this takes immediately cancels any advantage there might be from a high frame rate. When you stack all these, you get a blurred image. The only reason you would take so many frames is to reduce noise. Again there is plenty of light from the Sun, so this isn’t an issue.

It suggests that you might need a flat frame (possible) and that you take a defocussed 500-1000 frame avi to achieve this. Why? It is much easier and more accurate if you need a flat frame to simply blur an image you have already acquired.

My images use a £50 colour camera with a not particularly high frame rate. I find a good compromise is 200 frames. This takes around 7 seconds.

Click on “Solar” on the blog and judge for yourselves!

While still in Victor Meldrew  mode, in the same magazine there is a review of a new Skywatcher 20” goto dob for £5499. I am sure that this is a splendid scope, but following my earlier post it is worth remembering that it is only 1 stop faster than Rob’s new 14”! I am pretty sure Rob didn’t spend that amount on it! In fact, in the review there are pictures of M42 and the Trapezium. There is also a picture of M51 of recent discussion. They look nice, but I would invite you to compare the pics with these window-sill images with a scope costing £100 ish.

Moral – Just because something is in print does not necessarily mean it is correct. This is a hobby, it is whatever floats your boat. You can spend a fortune if that is what you want to do, but you don’t HAVE to!

Some Musings on Aperture and DSOs

Rob’s recent post on his new scope led me to reflect (!) on my own experiences. My main interest is deep-sky objects, or “faint fuzzies”. I have a rule of thumb that says that to see any appreciable difference in object  brightness, you have to go up 2 stops in focal ratio (f-numbers). This translates to a doubling in aperture (doubling the aperture increases the light grasp by 4X). So when I upgraded from my old 4”, I went to 8”. Doing the same again would suggest a 16” would be needed. For me, a 16” would be unmanageable, so I went to a 12”. This is only 1 stop advantage to the 8”, and visual results were a bit disappointing. This more-or-less coincided with getting the PD camera, and even using this “live” as an electronic eyepiece gave views so superior to the 12” that the 12” virtually never got used.

So I wondered why this was, so here is a bit of basic physics.

Firstly, the eyeball. This iris opens to about 7mm when fully adapted to the dark. Or maybe 5mm in an old fogey like me. Given the size of the eyeball this is around f/3. See http://www.faculty.virginia.edu/rwoclass/astr1230/human-eye.html

To get the brightest image from a scope, the magnification from the scope must produce an exit   pupil (the beam width leaving the eyepiece) pupil less than 5-7 mm. Otherwise some of the light leaving the scope doesn’t enter the eye and is wasted. This defines the minimum magnification you can use for a given aperture. For example with the 8” and a 5mm exit pupil, this minimum magnification is 200/5 = X40. (The SCT has a focal length of 2000mm, so this is an eyepiece of 50mm FL). A lower magnification than this is not harmful, it is just that you are then not making use of all the available aperture. 7X50 binoculars are known as “night glasses” for exactly this reason – their exit pupil is 7mm, making the best use of all night-time light. Therefore, going up in aperture does not necessarily make the view brighter, but rather allows you to us a higher magnification for the same brightness. This might be a huge advantage for small objects like planetary nebulae, but less so for extended objects. Bigger aperture also improves resolution, allowing you to split closer binaries, but this is usually not the critical issue for faint fuzzies. The other issues affecting brightness are the eye sensitivity (more of this later) and its “integration time”, or the time period over which the eye sums the image it sees. This is its “shutter speed” and there is some literature that suggests in the dark, this is about 0.2 seconds. Again. see http://www.faculty.virginia.edu/rwoclass/astr1230/human-eye.html

Now, from here, I have NEVER conclusively seen any galactic spiral arms visually, although sometimes I have persuaded myself I can. So when I first coupled up the PD camera to the 8”, and turned it on M51, this is what I saw, live, with no processing at all:

Bingo! Spiral arms!

So if we are now talking about imaging, rather than the eyeball, what comes into play for image brightness?

3 things.

  • Focal ratio (not aperture)
  • Integration time. The PD single image is 1/50 second. Its “senseup” parameter allows it to internally stack up to 1024 single images, giving an integration time of about 20 seconds, or about 100 times the eyeball.
  • If the CCD had the same sensitivity as the eye, the brightness of the CCD image would be the same as the eye if a senseup of 4 were used. An experiment is called for!. I attached a lens to the camera, set it to around f/3 (similar to the eye), then in a darkened room compared the image from it to that I could see with my eye for various senseups. Although this is a very crude experiment, I reckoned that about senseup=2 was about right-not miles away from the predicted value. So a senseup of 1024 suggests a sensitivity about 8 stops faster than the eyeball.

There is also another factor involved, and that is contrast. This is the brightness of the faint fuzzy compared to the sky “background”, and I have found that with our local skies, that is the controlling factor. You can improve this visually using filters (UHC or OIII for example). These tend to dim the whole view, but the right one can improve contrast. Light pollution filters used to be good in the days of low-pressure sodium street lighting but are not much use with LED lights. On the other hand, the camera has a “gamma” setting that allows you to “stretch” the contrast, live, or if you post-process, the sky is the limit, as they say! For example, stacked 11 of the basic frames of M51 (11264 frames in total), processed it with GIMP, and here is the result. It is flipped vertically to get the orientation right (See http://www.thornett.net/Rosliston/Astrophotography/DSO.pdf for the details). Stacking 11 frames with a senseup of 1024 gives another 3 ½ stops faster than the eyeball or 11 or so altogether.

Interesting as all this might be, let’s remember that this is a hobby – and you do whatever you enjoy!

As a final thought, Lord Rosse used a 72″ aperture reflector to first identify the spiral nature of M51!

M46 & M48 – wide field images

Again taking the opportunity of a half-hour imaging session, here is M46 in Puppis from the window-sill. Since it is a wide field view, the included planetary nebula, NGC 2438, (a line-of-sight effect, it isn’t in the cluster) appears pretty small, but after a bit of processing you can see it just above centre and slightly to the left. For those of you trying to see it visually, here is a quote from Stephen James O’Meara’s splendid book “The Messier Objects”:

“There is yet another illusion with M46. It appears to contain a tiny planetary nebula. NGC 2438 – – –  But the cluster and nebula are not physically associated because the cluster is 5.300 light years distant, whereas the nebula is 6,250 light years away. Positioned just a few arc minutes north of the cluster’s centre, this 11th magnitude planetary measures only about 1’ in diameter. I suspected it at 23X but 72X shows it clearly as a ghostly mote among the multitude”

(I see from my notes that I observed it visually and sketched it at 01:10 UT on 23/12/2001 with my 8″ SCT at X266)

Since there was no moon, I had another go at M48, showing a few more stars than the last one I posted.

For completeness with the Messiers in that region I have also included the recent wide-field image of M47

Wider angle view of M47

Managed a 10-minute observing session from the window-sill before the clouds rolled in. M47 is quite large so this time, in order to get a good context, I used a focal reducer. Using a reducer on an f/5 refractor is not optically very good – and it was rather hazy, so the image is not brilliant. You can compare it with the one without the reducer at http://roslistonastronomy.uk/m46-and-m47

Roger