Astrophotography Rob

This page last updated: 22 January 2017

In March 2015 I became bored with the robot building and thought I would have a crack at astrophotography, something I've always wanted to do after having completely failed at it as a boy.  Well, it wasn't photography then, it was getting a 4" reflector, without an eyepiece, that my father had acquired, to focus on anything, anything at all.  This time I will use the things I've learned in robot building to fully automate the rig, a robot of sorts I suppose.

Near the bottom of the page is a section on the basics, using only simple worms (and with drawings in crayon) for us mortals.  It is followed by a section on the important beginners lessons I've learned.  The remainder of the page is my astrophotogaphy diary, starting with my first attempt at the bottom and with updates on progress in chronological order, most recent at the top.

Finder Alignment And Telescope Focussing
22 January 2017

Alignment of the finderscope and initial focussing of the telescope are rather more difficult to do in astrophotography than with a normal telescope setup, particularly in an automated situation such as mine.  This bit took a lot of patience and several afternoons/evenings of fiddling.  It also required a window through which you can see the sky and distant objects on the ground, with the Linux machine running Indi close by.

To get the finderscope and the telescope basically aligned I set the telescope up beside the machine running Indi and a window with a good view during the day time.  I unclamped the axes of the telescope mount, pointed it at the window and looked for a distinct feature in the distance, something I would be able to recognise the shape of easily; a satellite dish sticking out from a house about 150 metres away in my case.  Using Indi's streamed view I centred this object in the finderscope CCD and tightened the clamps on the axes of the mount.  Then I switched the streamed view to the CCD on the telescope and focussed the telescope (on anything at all that has edges; if all you see is sky or a featureless wall, etc., use the Indi mount controls to move
the telescope around a little).  I did this focussing using Indi and the motorised focusser though, on reflection, it might have been better to set that to mid range and establish focus using the focus knob on the telescope tube itself.  With this done, I slowly and laboriously used the Indi mount controls to move the telescope until the distinct feature was (a) found and (b) centred.  Then I switched back to the streamed view of the finderscope and adjusted the screws until the distinct object was centred in the finderscope streamed view again.  The finderscope and the telescope are now roughly aligned.

Next I waited for a starry night.  Again I set up the telescope beside the PC running Indi and an outside window, one from which I could see some stars, with the mount positioned to point vaguely north.  The nub of the problem now is that the relative fields of view of the two CCDs are vastly different, a night sky is mostly blackness and I had no idea where the right focus position was or what the right exposure duration would be for either CCD.  So what I did first of all was remove the focuser/CCD assembly from the back of the telescope and put back the optical eyepiece.  The situation is something like the picture below, where red is the field of view of the finderscope CCD, green the FOV of the optical eyepiece and purple the FOV of the telescope CCD.  There are one hundred times more bad focus positions than good ones and in a bad focus position all you'll get is noise or blackness.  And the stars are that dim.  They eye is much more reactive for this kind of thing and the field of view of the optical eyepiece is more forgiving.

Fields of view of finder scope CCD, optical view and
          telescope CCD

In the Indi mount controls I set the mount to run sidereal tracking.  This was not going to be very good, as the mount is not properly aligned, but it was better than nothing.  I unclamped the axes of the telescope mount and pointed it vaguely at some stars, then I used the Indi controls to try different exposure durations on the finderscope until some stars appeared; 0.2 seconds seemed about right for my QHY5-II CCD.  Then I looked through the optical eyepiece of the telescope and adjusted the focus knob on the telescope tube until something came into view.  Given the differences in field of view, and with the finderscope not yet being aligned, it is possible you might not see any stars at all, so this bit may require several attempts and some unclamping/wiggling of the telescope.  When you've finally got something in focus, and I was luck enough to see two stars of an easily recognisable spacing to give me some context, you can pick one star,
unclamp the axes of the telescope and, very gently indeed, centre this star in the main telescope eyepiece.  Then I went around this loop:
You must go around this loop quickly, as the star will be moving out of position all the time.  You'll probably need to unclamp the axes, centre the star again, and repeat the whole thing a few times to convince yourself it is right.

That gives us alignment to the green circle, next we need to get the purple one sorted, the telescope CCD field of view.  I removed the optical eyepiece and put back the focusser/CCD assembly.  I used Indi to set the motorised focusser to mid range and I turned the telescope tube focus knob all the way out (fully anti-clockwise).  The next bit needed two people, one at the Indi controls and one at the back of the telescope.  The procedure went as follows:
  1. press the exposure "set "button on the Indi tab for the finderscope CCD to get a new finderscope image,
  2. use the Indi mount controls to make sure a star is centred in the image (repeating from (1) until done),
  3. press the exposure "set" button on the Indi tab for the telescope CCD to get a new telescope image (you will need to try different exposures; I found that 0.7 seconds was about right for ASI 120 MC, you want to see some noise but not too much),
  4. turn the focus knob on the telescope tube half a turn and look for a star appearing in the image,
  5. if it does not appear, repeat from (3) but, every 10ish attempts, return to (1) as the star may have drifted from centre.
Eventually, and it took several attempts, we got a star to appear in the telescope CCD image.  The moment we managed to do this for the first time the sky clouded over, so I don't have a saved image to share.  But we did achieve both focus and alignment, which gives me the confidence to try outside again.

Note: you'll notice that I used the Indi streamed view for the day time work but not for the night time work.  This was because it didn't seem to work at night, which was down to a bug in the way exposure settings were passed around in Indi that has since been fixed.  That said, viewing the images separately in a FITS viewer window is probably a better idea as there is a little "cross-hairs" button you can toggle, making it a lot easier to know when a star is centred.

2 January 2017

Having had some issues with the QHY CCD driver that required Chinese intervention, I've spent the last 6 months being distracted by other things.  Only tonight have I finally got back to trying out this darned setup of mine.

Astro ready to go

I haven't got very far yet but I have learned a few things this evening:
I've managed to do all of the above but, so far, only with the finderscope; my main camera seems to be generating a lot of noise and it is not sufficiently well aligned with my finderscope camera that I can tell where I am.  Anyway, for what it's worth this is Betelgeuse through [the by now frosted glass of] my finderscope camera:


Yes, yes, I know, not particularly impressive but I do know that it is Betelgeuse, not just some random star, it stayed put in the view finder despite the world going round, I've lined up my Indi world with the real world and, despite it getting down to -2 C outside, I've stayed toasty warm.  Now I just gotta do this with the actual telescope rather than the finderscope.

Waiting For Clarity
22 May 2016

I have re-built the SSD image on the Hummingboard Pro (leaving out Siril this time as I have Ethernet to the lawn now), fixing the Hummingboard Pro into the box with nylon bolts, and verified that the whole rig works off power over Ethernet; no fan needed, no overheating issues.  All I need now is a clear sky [drums fingers on desk].

China Deliveries
14 May 2016

Last week the components I was waiting for from China (IP68 trailing Ethernet socket and power over Ethernet extraction module) arrived. I've now got everything wired up and have tested that it works, then I secured everything into place in the box.

Computing box, top view

Ethernet input and power over ethernet output
12 V
                power input (and output, since they are all connected
                together) and USB
                rest of the USB connectors

Unfortunately, one of the only two mounting holes that are accessible on the Hummingboard Pro rubs right up against the SSD and so, in attempting to secure the Hummingboard Pro, I must have shorted some of the layers of the SSD's PCB.  Another one has been ordered; I'll have to set that up again and use a nylon bolt next time.  I may also need to think about cooling, perhaps a small fan.  Anyway, very nearly there now.

Internet To The Lawn
2 May 2016

Last week, while in the midst of finding the stand, I had a moment of inspiration.  I was pondering whether my Wifi connection back to the house would be sufficiently reliable and it occurred to me that I could run an Ethernet cable out to the telescope if this was an issue.  Then I thought a bit more and looked up power over Ethernet.  A quick measurement of the power requirements of my whole telescope, with motors running, showed that it was less than 2.5 Amps at 12 Volts, which is what the latest 802.3AT power over Ethernet standard can deliver.  A quick search on the internet found an IP68 trailing Ethernet socket, a power over Ethernet extraction module that could be mounted inside my existing plastic box, a power over Ethernet injector unit and a Brabantia plastic ground tube with cover (intended as the base for a rotary washing line).  The first two items are coming from China, so I'm not yet completely operational, but this weekend I put all of the wiring in place.

I used outdoor-grade Ethernet cable (UV-proof) but, for good measure, put it inside a 20 mm plastic conduit.  I dug a spade-depth (~200 mm) slot in the lawn by driving the same metal pole that I used when inserting the legs for the stand into the ground, wiggling it from side to side (and repeat) then digging out the soil with a spade.  I cut the bottom off the plastic ground tube so that water could drain away and I connected the conduit with the plastic ground tube about half way up, to prevent any water that might sit in the ground tube from running down into the conduit.  I needed to use some extra soil to fill in the slot, patted it down hard, and sprinkled grass seed over the top.  The plastic ground tube is positioned in the centre of the tripod mounts and the cable runs back to the power over Ethernet injector box, and thence the router, in the house.  A very satisfying bank holiday weekend project.

Plastic ground tube with cover
Digging the slot

Finding The Stand
30 April 2016

Now that I have all the bits, after some practice with KStars/Indi/Ekos, I want to get out and do stuff.  This left me with the problem of finding the stand mounts in the lawn after 9 months of moss/grass growth.  I had expected this problem but had assumed that I could find the large lumps of metal using one of those DIY things designed to find metal in walls.  However, those devices seem more oriented towards detecting changes in density and the ground is just one big dense thing so no dice.  I considered making a metal detector using the instructions here but didn't fancy vero-boarding a design with so many discrete components.  I had a 555 timer lying around so I tried making one according to the instructions here but it is too simple and therefore too insensitive; you really need two oscillators beating.  So I tried building a Velleman K7102 kit metal detector, which was very satisfying, but again no dice; it is not sufficiently sensitive and, in any case, it is quite a tiny thing to drag over several square metres of lawn in a structured fashion.

After making a few attempts to beg, borrow, or hire a proper metal detector (and failing) I bought the cheapest kids' metal detector that Maplin had to offer, the Beginners Metal Detector With Analogue Display and, finally, that did the trick.  I now have measurements of where the mounting points are engraved into the nearest paving slab, along with a big arrow.

So beware, mark your mounting points clearly if they are going to be consumed by grass.

23 April 2016

After a long wait in customs, my focuser has now arrived.  I chose to use Moonlite Focusers; really helpful people.  They manufacture all their own products and can make mods as necessary and sell you as much or as little as you need.  Here's a video of them making their focusers; they even have their own pick and place machine for the circuit boards.  In my particular case they machined the flange, making it slightly higher so that it wouldn't foul the focus knob on the back of my 6" reflector.  I went for their "CS 2 inch format SCT" focuser with stepper motor and their V2 controller, though this latter was an accident as I had originally intended to make my own controller, powered from USB.  Ah well.

Here's the focuser fitted to the back of my telescope and the V2 controller.

                focuser Moonlite V2 focus controller

Simply unscrew the existing eyepiece from the back of the telescope, screw the focuser assembly into place, remove the draw-tube from the back of the focuser and attach your CCD there.  The V2 controller provides an ASCOM interface over USB, for which Indi has a built-in driver.  I was up and running in no time at all.  Sweet.

So that's all the bits sorted.  Now I need to understand Kstars/Ekos/Indi properly.  I have found an old laptop that I have loaded Kubuntu Linux onto, saving me having to take my Windows machine out of service when I want to do astro stuff and, incidentally, providing me with a computing solution that I can carry to the telescope if I have to.

You never know, I might even take a photograph soon.

Quick Guider
5 April 2016

While my focuser is arriving I took a quick look at how I might set up a guide scope.  An equatorial mount is capable of slewing nicely with the night sky but it will have innaccuracies; these are resolved by locking onto a star, evaluating where it is in the field of view, and feeding correction pulses into a separate "ST4" port on the equatorial mount.  There are three ways to grab the necessary starry image:

  1. mount another telescope on your telescope with a CCD on it for this purpose,
  2. grab a portion of the light coming through your telescope and divert it to another CCD for this purpose,
  3. divert the infra-red part of the light coming through your telescope, which you don't care about, to another CCD for this purpose.

There are pros and cons to each of these, (3) being about the best and the most expensive.  None of them are cheap, though, so having just spent quite a lot of money on a focuser, I went for the cheapest option, which is to turn the small finderscope you get with your telescope into a guider.  There's a very helpful video on YouTube from a guy describing how to go about it:

What it amounts to is this:

150 later and I have a guider; probably one of the cheapest things I've bought so far, astrophotographically speaking. It integrates into Indi really easily; Indi even offers a mode where all the guiding is done in the driver on the server machine and so Wifi bandwidth is minimised.  The CCD guider not only takes the images and is USB-connected/powered to/from my Hummingboard Pro but it also takes back the ST4 correction messages over the same USB port and, through another port on the back of the CCD guider, sends them to the ST4 port on the equatorial mount (the necessary cables being provided with the CCD guider).  Very neat indeed.  Here it is in position, the CCD guider (the silver tube) mounted into the finderscope with USB and ST4 cables coming out of the back:

Finder Guider

This does mean abandoning my red dot finder, which would have mounted in the same foot on the telescope, but now that I'm all camera'd up and going-remote it is less necessary and I can always slot it back in for a moment if I need to.  Of course I will find out the limitations of this thing pretty quickly but, right now, I have a cheap solution that allows me to play with a complete system.  Then I can get it right once I've had the chance to determine what wrong is.

26 March 2016

A long overdue update.  Late last year I got quite a long way with this project but failed to write it up here before being overloaded with work for 6 months.

My plan is to put a small board computer with the telescope and then control that over a Wifi link from a PC in the house. Wifi has low bandwidth, especially when the device is not in the house, so I need to do as much processing as I can on the small board computer and, preferably, only do control over the Wifi link.  I could probably also do with a directional Wifi terminal that I can point at the hub in the house.

In a similar vein, the Raspberry Pi, convenient thought it is, is terrible at IO, really slow.  If I am to do image processing on the small board computer it needs a proper disk interface and lots of storage.

Then there's the software. I need to be able to split the client and the server, preferably with the client running on Windows and the server on Linux on my small board computer.

Then I need to package the small board computer and sufficient IO up in a manner that can be kept with the telescope.

This is what I purchased:

The assembled computing hardware (with the Hummingboard Pro and the PSU board not yet fixed in place) look like this:

Computing box Computing box

Going through the contents, at the top is the USB hub, in the centre is the Hummingboard Pro with the SSD mounted underneath it, and beside that is the small green PSU board. Around the outside are the USB connectors and the two power connectors.  The power connectors are effectively wired in parallel so that power can be supplied through one and distributed (to the HEQ5 mount) via the other [I later added a few more as there were more things to power]. I also bought a 12 V laptop power supply block that can be used instead of the Tracer LiIon battery pack while I'm developing things.

Next there's the software.  The thing for this is Indi, which allows the client and server running the devices to be utterly separate.  As I know it reasonably well from robot work, I chose to run Arch Linux on the small board computer.  In theory, you can run the Indi client on a Windows machine.  However, in practice that doesn't seem to work and the most mature and capable packages for astrophotography, being the pairing of KStar and Ekos, run best under Linux anyway so I resigned myself to using Linux for the client side.  Rather then purchasing a new computer for this, though, I have simply downloaded bootable Ubuntu onto a USB stick and that seems to work perfectly fine.  I can boot my usual Windows PC into Linux when I need to.

I spent some time configuring the small board computer to run an image processing package called Siril, controlling this over VNC, so that I can do local image processing if I want to, and also of course to run the Indi client.

On the camera front I discovered ZWO ASI CCDs.  These are Chinese made and offer great performance at a bearable price.  I purchased the ASI 120 MC.  It also happens to have Indi drivers already written.

My very detailed notes on the computing setup process can be found here.

Setting up such a system is not without its frustrations and setbacks of course.  Many a time I was stumped by a problem but the people on the Indi forums were always able to help and, as of today, I have a system where I can control the mount and the camera from the comfort of my loft with the power of Kstars allowing me to point it anywhere I like in the sky.

The only thing remaining is to purchase a focusser.  I've been putting this off until I've proved that the rest of the system works.  Now I'm off to find one, preferably one that can be powered/controlled over USB.

Making A Stand
28 June 2015

The best position for the telescope in my small garden, the least obscured, happens to be in the middle of the lawn.  So I've made myself a set of stands that can be mounted on the lawn in perfect polar alignment, and dead level.  Here's what I did.

I bought the following bits and pieces:

Stainless steel tube. Stainless steel threaded bar.
Stainless steel threaded connector. Stainless steel plate.

You'll note that I bought at least four of each, because I wanted a spare for the next stage.

I tracked down a colleague who had an interest in old cars (in this case Triumph Heralds) and therefore had an arc welder, a MIG welder in fact, that has the capability to weld stainless steel (thanks John!).

                  and the welding. Welded
Completed stand. Completed stand.

I then did a polar alignment of the mount in position on the lawn and marked where the feet were. I used a long metal bar and a sledgehammer to make three holes at those positions, into which I pushed the stainless steel tubes, using the plastic bungs to block off the hole afterwards to prevent them filling up with gunk.

                  in the ground. Stands in the ground.

And there we are, a perfectly aligned, removable, telescope stand in the lawn.

Tripod in position.

One thing: I added the short piece of boxed section tube so that I could insert a screwdriver, or some such, when screwing the plates down into the tubes. Do watch out if you just spin the metal plate with your fingers, the edges of the plate will be sharp.  And, see also Finding The Stand above.

First Light Two
21 May 2015

After a little discussion, Neil at Green Witch was happy to accept back the Celestron telescope and replace it with a new selection as follows:

My intention is to put together something with good foundations for astrophotography (the equatorial mount and power supply) plus a sufficiently good, yet not expensive, tube.  Once I have the automation going I can think more about what I want to observe and purchase the right telescope for that application.  Very excitingly, it all arrived today.  I assembled it in the garden:

setup two

...and pointed it at Jupiter, whereupon the planet and string of moons became visible.  Having not set anything else up as yet, I used my mobile phone camera to prove it:

            phone photo of Jupiter and its moons

Now for some alignment and then some automation.

Beginners Lessons
12 April 2015

Here are the things I have learned so far, roughly in order of importance:
  1. You will never be able to get it right at the beginning.  It's not because you're being stupid, don't think that, it's because you don't know what questions to ask.  At the very least, find a "newbie" forum on an astronomy site and ask everything that does come into your head without embarrassment.  Don't give up.
  2. Astronomy is very expensive.  Even the simplest things cost 150 or more.
  3. It is very easy to buy the wrong thing, or a thing that won't work with another thing.  Check, check, and check again, preferably with others who've done the same... thing.
  4. Astronomy products are generally designed for those who are happy to stand around in the cold and the dark.  If, like me, you want to sit in the warm and automate the rest, most astronomy products are not designed for you and so your job in selecting the right things is even harder.
  5. Astrophotography is subtly different to astronomy.  Sure, a camera can be attached to anything but the optimisations are different (choice of focuser, choice of mount, choice of automation platform).  So when asking for advice be clear what you want to do.  A good resource is
  6. Telescopes for looking at planets are generally not ideal for looking at galaxies and other deep space objects.  Planets are really tiny so require high magnifications (large F-numbers), while galaxies and the like, though far away, are really big and require low F-numbers (otherwise you only get to see a part of them through your telescope).  So again, be clear what you want to do when asking for advice.
  7. Astronomy equipment manufacturers don't do computing. It may seem like they do, but they really don't, believe me.  They don't know what they've got or not got, they have probably bought it in from someone else and they very likely can't support you with it adequately.  For anything computing related, go find the computing community who've done astronomy and rely on them only.  Good resources are and kstars.
  8. Most telescope manufacturers are in the USA so direct support can be a pain.
  9. Interfaces for controlling stuff aren't generally open.  Check that, if you want to automate something, someone else has reverse-engineered the interface already (see or be prepared to do so yourself.
  10. If you aren't going to leave your telescope in position permanently, think about how heavy it all is: the telescope, the mount and the tripod can be anything from 18 kg to 30 kg when assembled.  If it's heavy and difficult to assemble you just won't want to do it.
  11. Consider planning your astronomy for the entire year ahead (though any one viewing session will of course be cloud-permitting). This way you can assemble the telescope during the day and then cover it, disassembling it during a subsequent day once you're done.
  12. If you wish, you can separate the decision on which telescope to buy from the mount/tripod it stands on.  For astrophotography the choice of mount is as important as the choice of telescope.
  13. The earth, and hence the sky, rotates quite quickly.  You need a motorised mount with automatic control to keep anything in the view finder for more than a few minutes.
  14. Planetary photography requires only an alt/az mount, i.e. one which has two motors, one for up/down and one for left/right, with an automatic control system to keep the planet in few.  Then you basically take a video of the sky and use software to chose the best frames and integrate them.  But there are only two interesting planets (Jupiter and Saturn) and the moon, so you'll run out of planetary objects pretty quickly.
  15. Alt/az mounts are quite compact but this often means that when the telescope is pointing directly upwards the eyepiece is close to the top of the mount.  You will not have room to add an optical focuser because it will foul the mount.  On Smith-Cassegrain telescopes you can focus the mirror instead, which is fine, but the image will shift from side to side, often out of view, when you do this, which makes the operation rather fiddly.  You will need a motorised focusing mechanism if you're going to be able to sit in the warm.
  16. To look deeper into space you will need longer exposures (and a clearer sky).  For exposures longer than 30 seconds software won't help you compensate for the moving thing and so you need an equatorial mount, i.e. one which can track across the sky at the same angle as the earth rotates with a single motor, no zigging and zagging.
  17. Equatorial mounts have to be polar aligned so that their motor can then track objects in a "straight" (if you see what I mean) line.  Note that it is the mount which needs to be aligned.  This is different from discovering the alignment of the telescope, which lots of clever software can do for you automatically.  Aligning an equatorial mount is quite doable with a little practice and, if you're always placing the telescope in the same position, you can mark it out and you'll not have a problem.  But these mounts are more expensive and tend to be heavier.

The Basics
11 April 2015

Having muffed it the first time, I decided to spend a little more time consulting on the right equipment for my second attempt.  I went back to the basics, which I understand are as follows.

The hole in the end of the telescope furthest away from your eye, the bit at the starry end, is the APERTURE.  The bigger the aperture
(measured as the diameter of the hole), the more light comes into the telescope  More light is good because you can take shorter exposures with a CCD camera and that both enables the "take a video and select the good frames" approach for planetary stuff and means that any blurring effects due to the earth (or your telescope) moving are minimised.

Then there is the length of the telescope, the distance from your eye to the aperture.  Lets say we have a shorter telescope, relative to the aperture.  It might look like this:

Large field of

The field of view (the area inside the red circle below), i.e. the number of stars/planets you can see (and they are, of course, way more dense than I've drawn here) is quite large:

Large field of

You can see quite a bit, you can get quite a lot of light in, it is easy to tell where you are pointing in the sky because there is more context, but each individual object is quite small and, considering using a CCD camera with its array of pixels, the resolution of any one object on that camera is not high.

If we go for a longer telescope, relative to the aperture, it might look like this:

Small field of

...and the field of view will be correspondingly smaller:

Small field of

Now, when that image falls on the CCD camera, the number of pixels occupied by any one object is increased and you can see more detail.

Of course, the telescope is not an empty tube like some sort of enormous pin-hole camera, it has lenses and/or concave mirrors.  These manipulate the light rays that come in the front of the telescope so that, rather than most of them bouncing pointlessly around the tube, they all land at your eye in focus.  Hence, the length of the telescope is pretty much the same thing as the FOCAL LENGTH of the telescope.  There is a really useful number, the single number you need to define the shape of your telescope.  It is the F-number (not to be confused with the focal length), which is:


For example, a telescope 50 cm long with a hole in the end 5 cm in diameter has an F-number of 10.  The bigger the F-number, the smaller the field of view, the more detail you get.  But you still want the aperture to be large for the light thing and hence there are compromises to be made.

On that lens/mirror system: making large lenses of good quality is difficult and expensive, hence there are reflector telescopes. These use mirrors to focus the light instead of lenses; large mirrors are relatively easy to make.  Reflecting telescopes also bounce the light back and forth
within themselves a few times.  So while the length of the tube is N, the light travels along it three times and they are equivalent to a refracting telescope (i.e. one with lenses) of length 3 x N.


This means you can get a larger aperture for a given F-number without a hugely long and unmanageable tube.  But this doesn't come for free: the reflector does have obstructions at the starry-end (a small mirror with struts to hold it) and a hole in the large receiving mirror where the eyepiece is, or maybe a further small mirror with struts suspended there to push the light out the side.  You won't see these things in the image as they are not in focus but all of them reduce the image quality somewhat.

And finally, there is the stuff going on at the eyepiece end.  Here be [more] lenses.  The telescope has brought everything to a nice sharp image at some point slightly beyond the end of the tube, but this is a really tiny point and what you want is something that sits nicely across your CCD.  The image is also distorted, the rays of light have been modified by the lenses/mirrors in the telescope and they need to be nearly parallel again, as they were on their way in, to present an undistorted image.  This is what the eyepiece does:


Just like the telescope, each eyepiece has a fixed focal length, which is usually stated in millimetres.  Once it is fitted to the telescope, the entire system now has a


For example, a telescope of focal length 2 metres with an eyepiece of focal length 25 mm gives a magnification of 80 times.  You've taken the nearly parallel rays of light coming in at the front (or a subset of them at higher magnifications), processed them and spat them out just as parallel but at a size suited to the CCD at the back.  The smaller the focal length of the eyepiece, or the larger the focal length of the telescope, the higher magnification you get.  Bare in mind that high magnification also means that there is less of the sky in view (so things can be very difficult indeed to find as you are without any context) and also the slightest vibration (and the earth's rotation) will have a significant effect.

You will note that the focus is always fixed, there is no such thing as a zoom lens in astronomy, which I'm guessing is because such complex lens systems suffer from all sorts of distortions/losses which are acceptable in a hand-held camera but not when viewing things at very low light levels.  There is, however, a thing called a Barlow. This is a diverging lense which you put between the eyepiece and the telescope.  It is a multiplier on the magnification of the eyepiece.  So a 3 times Barlow would increase the magnification in the example above to 240 times.  This can be useful in spreading the image across your CCD, using as many of those pixels as possible.  And Barlows can be bought in a form where their magnification is adjustable.  The thing to watch out for with Barlows is that they are far longer than a normal eyepiece (~15 cm as opposed to ~4 cm) and you need to make sure they don't end up fouling the mount at some angle of the telescope.

One last thing: focus.  I don't know about you, but I remember being taught at school that things sufficiently far away, like stars, are so far off that the light rays getting back to earth are virtually parallel, as I mentioned above.  If that were true you would have thought that focussing were irrelevant, just like when you set your hand-held camera's focus to infinity for things that are more than a few metres away.  But it seems that's not completely true, or maybe its these large F-numbers, or maybe it's variability in the telescope that your compensating for, or maybe it's just that Jupiter is a helluva lot closer than any star.  Either way, a focus knob is required.  For refracting telescopes this means moving the eyepiece back and forth. For reflecting telescopes one can do that also, or one can move a mirror instead.  However, I'm told that when doing mirror-based focusing the position of the image in the frame tends to move randomly left/right/up/down at each focus position, making taking high resolution pictures a bit fiddly as the telescope may have to be moved slightly to regain the image.  So even for a reflecting telescope, focussing through moving the eyepiece is considered superior.

If you want to get a better feel for all of this, there is an interactive telescope simulator here:

There's also a really useful Windows application that will show you the field of view of various real telescope and CCD systems (plus you can add your own), downloadable from here:

First Attempt At First Light
Updated 21 May 2015

At the back end of March I visited Neil at Green Witch to check out my options.  I was looking for something that would do planetary astronomy, that a CCD camera could be fitted to, that was powered and had some existing level of automation on the motors.  I was wowed by the Celestron Nexstar Evolution 8". In particular, the mount can be controlled over Wifi from an Android app, which I thought I should be able to run on my Blackberry 10 phone, and hence I could use that existing automation while I concentrated on automating the photography side of things.  The telescope arrived on 23rd March and on 24th March we were looking at Jupiter and its moons in perfect focus, all controlled from my phone.  Amazing.  I was truly excited, a hey-wow moment of first light.

I planned to create automation (more on this later) so spent the next few weeks working on that while the skies were not so good.  Then on Sunday 5th April I had my next opportunity to get out at night.  On this occasion the app refused to run and subsequently refused to download its star database after reinstalling.  Huge bummer.  The first outing was so exciting and yet now I was utterly stalled by some software outside my control supported on an e-mail based support line to a California-based company.  It took them two days to tell me that my phone wasn't supported.  This wouldn't have been a problem if it had never worked on my phone, I'd have probably gone out and got a cheap Samsung handset to work with it, but since it did actually work first time I found that I no longer trusted it at all.  We get so few clear nights here, being utterly buggered by something out of my control would make me heart-attack fodder I think.  And then the mount failed entirely, wouldn't power up, no lights on it under any circumstances.

So I have retrieved the packaging from the bin, packed up all the kit, taken it back to Neil and sent a letter to Celestron asking for my money back.  I don't think their mount/telescope is bad, in fact it was great, I simply don't feel that I can trust it after this episode.

But now I'm on the cost/specification escalator, as Gordon Aspin, an amateur astronomer who I work with, put it.  My next purchase is going to be an equatorial mount with a separate, beefier, power supply, and it's already looking like it's going to cost 200 more than the Nexstar.  Ah well.

Back to Meades Family Homepage