Category Archives: Astronomy

Solar System

Posted by on March 11, 2012 No comments

A friend on Facebook pointed out that it was very clear outside tonight. I looked out of the window, and it was indeed – so I grabbed my telescope and decided to try harder at astrophotography (rather than plain old astronomy).

In the past, stuff that has looked great through the eyepiece have been disappointing when captured on camera because of the lower magnification. I need a Barlow lens, but tonight I thought I’d try experimenting using a teleconverter instead.

Here are the pictures I managed to take. If you’re a geek and you like lenses and telescopes, read on to see how I took these pictures. If you’re normal, you might as well stop here ;)

  • Venus has phases, just like the Moon. Here less than half of Venus is illuminated.
  • If you look carefully, you can see a sliver of white around the top of Mars. This is its polar ice cap.

I do have a modern EF-mount teleconverter but the camera refuses to work at all if you don’t attach a valid lens to the other side of it (a telescope mounted via a T-ring doesn’t count). So I was forced to use a 1970s FD-mount teleconverter which is actually better than the modern one, but it does mean I have to use an FD-EF adapter. Mine isn’t great and causes some aberrations.

So here’s my kit list:

With all that teleconverting going on, the f/12 telescope quickly turns into an f/36 monstrosity. That’s pretty slow. Nonetheless, Venus is still bright enough that I could capture it at ISO200, 1/50s.

These images have had very little processing – they haven’t been stacked. I’ve only done an unsharp mask on them. They are pretty low resolution – even with a huge focal length you have to crop an awful lot to make a planet fill the frame.

To improve on this, I would need a larger magnification teleconverter or Barlow lens. Teleconverters generally don’t come in higher powers than 3×, but I have seen Barlows at 5×. Trouble is, I’m already running into stability problems with the telescope’s tripod. It’s extremely hard to keep it still as it is! I’m not sure 5× would be possible on my tripod, and it’s well beyond the telescope’s highest useful magnification.

A new view of the Sun

Posted by on February 26, 2012 No comments

Recently I wrote about making a solar filter for my camera but it’s taken a while to get round to using it. I had a quick go today, using my Canon 450D and a Tokina 400mm f/5.6 lens.

The finished filter

The results are a bit iffy – there were massive problems with internal reflections in the lens tube. The filter itself consists of two plates of glass with a thin layer of nichrome between them. This on its own gives several surfaces for reflection problems.

Common wisdom says that stopping down a lens increases sharpness (to a point) but in this case, the smaller the aperture, the more area there is of the aperture blades to reflect light back up to the mirrored filter. All of the photos have lots of flare, which I was able to reduce somewhat. Perhaps with practice I will learn techniques to work around this.

This first picture is a straightforward picture of the sun. The red “corona” around the bottom-right hand side is actually lens flare.

The Sun

The second picture is made from the source data as the first picture, but is only showing the green channel in monochrome. It has less flare, since most of the flare was red. It also seems to show off more detail of a band-like structure across the centre of the sun (although this could be down to lens flare, Newton’s rings, or some other kind of defect).

The Sun

Making a solar filter

Posted by on February 16, 2012 No comments

I wanted a solar filter for my camera, for safe observation and photography of the sun. Typically these filters have an optical density of OD 5.0, which for photographers is ND 100,000. This means that only 0.001% of light is allowed through the filter.

You can buy pre-made solar filters, but they are expensive and relatively hard to find. You can also buy solar film and then mount it onto your camera lens or telescope yourself, but it is relatively fragile. If it gets a scratch, you could potentially blind yourself. Crinkled foil could also spoil the image quality.

I wanted a solution that was effective, durable and wouldn’t break the bank. Fortunately, one of my friends works in a chemistry lab and has access to a really cool machine that can coat glass filters with various kinds of metal. (This is the same chap who has the frickin’ laser that I wrote about not long ago).

Choosing a donor filter

I needed a 72mm filter for the telephoto lens I planned to use. I bought two cheap UV filters, although plain glass blanks would have been equally good.

What’s important is that the glass can be removed from the frame. Cheap filters often have the glass glued into the frame. Semi-decent ones have the glass dropped into the outer ring, and then the inner retaining ring is screwed in to hold it. Usually these screw-mounted filters have a screw thread on each side of the filter so you can stack two or more on the same lens.

The filter in this picture is the wrong type, where the glass is glued into the metal frame.

Glued filter

And here’s a close up of the kind with a screw-in retaining ring inside. This makes disassembly really easy with a pair of screwdrivers (be careful!).

Filter with retaining ring

I bought two UV filters made by Neewer, who are a relatively well-respected budget filter maker. Ordering from China, I paid less than £2 for each filter.  It was then easy to remove the glass from the frame.

Coating

In order to achieve an optical density of approximately 5, you need a total metal coating thickness of around 80nm. Trouble is, these metal coatings are very easily scratched – a fingernail would easily remove them. So I decided to coat each of the two filters with 40nm of metal and assemble the filter with the two metal coatings sandwiched together between the two pieces of glass.

The coating process is interesting. It’s done by boiling your chosen metal using an electric filament until it becomes a gas. At this stage, it will float around and stick to everything. However, you have to do this in a sealed bell jar that contains a vacuum.

First I cleaned the filters using a jet of compressed air, followed by a wiping with butanol and acetone. Any dust or grease on the filters will prevent the coating from sticking properly.

Then with the clean filters placed on the coating platform, I loaded a filament made of tungsten, wound with nichrome wire, and placed the bell jar over the whole lot. The small lump to the upper-right of the filters is a thickness sensor.

With the equipment set up, I started pumping the air out of the chamber. There’s a conventional rotary pump that can get the pressure down to a thousandth of an atmosphere, but after that you have to use a diffusion pump. For the coating to work properly, you need a pressure of just 10-6 millibars – that’s a billionth of an atmosphere. This takes about 20 minutes.

When the pressure is low enough, the chamber is sealed, and we can turn the filament on to gently begin heating the nichrome. You can’t heat it too fast, or it will all melt and drip off the filament and be wasted.

When the nichrome starts to evaporate, we crank up the current in the filament to some 50 amps. Now the whole filament acts like a light bulb, and is too bright to look at. No photos of the filament in this state, I’m afraid. I was too busy shielding my eyes.

There’s an instrument on the desk that’s hooked up to the thickness sensor I mentioned earlier. This tells me how thick the coating is, in nanometres. Here the readout is showing almost 25nm, over halfway to my 40nm target. I didn’t time how long it took to build up 40nm but it was something like 20-30 minutes.

After a while we can once again look at the bell jar, since the inside of it is partially coated with metal and quite dim. You can see how the filters now appear to be mirrors.

When the thickness meter reckons we’ve got the right amount of metal on the glass surface, we turn off the filament and gently let the air back into the bell jar. Cautiously we remove the filters from the platform using gloves (because they are hot, and because we don’t want to get dust on them).

Assembly

Now it’s just a simple matter of placing the two filters with their coated faces together, and screwing them back into the metal filter mount. Try to avoid the two filters sliding or rotating against each other. If you have dust between them, it could easily cause a scratch. Scratching one of the coated surfaces isn’t a huge problem, but if you scratch both faces then you’ll get a pinhole which allows unfiltered sunlight through, and will do a pretty good job of zapping you in the eye.

The coating is not exposed to the outside world, so this filter is no more fragile than a normal filter. The outer glass surfaces are just regular UV filters and can be cleaned in the usual way.

I found that with two pieces of glass together in one filter frame, the thin retaining ring didn’t seem that secure, so I simply used the spare frame to hold the glass in place. This means my filter is twice as thick as a normal filter, but that’s no problem.

The finished filter

And here is the filter in situ, mounted on my Tokina RMC 400mm f/5.6.

The filter in place

Testing

Unfortunately as it’s winter, it hasn’t yet been daylight during any of my free time, so I haven’t been able to use this filter with the sun. However, by comparing photos of a lightbulb in my house with and without the filter, I was able to estimate the optical density.

Without the filter, I shot the lightbulb at ISO1600, f/5.6 and a shutter speed of 1/4000s. To achieve the same exposure with the filter (keeping the ISO and aperture the same) needed about 120 seconds. That’s 200,000 times more exposure.

  • Transmittance: 0.0005%
  • Neutral density: ND 200,000
  • Optical density: OD 5.3

Given that I was aiming for an OD of 5.0, I’m pleased with OD 5.3. If my measurements are correct, that will be totally safe for viewing by eye through the camera’s viewfinder. Anyone fancy a sweepstake? :D

Soon, I hope to be able to publish some pictures of the sun taken with this filter – obviously depending on there being sun during the day at the weekend while I’m free. Cross referencing my personal calendar with the forecast for Bristol, it might be a few weeks…

Astrophotography tutorial: shoot the Moon

Posted by on January 28, 2012 No comments

Whether or not they have an interest in astronomy, at some point most photographers are likely to take a picture of the Moon. There’s a lot more to astrophotography than you might think, so I’ll walk through this step-by-step guide on shooting the Moon.

Capturing the best source image

It’s important to capture the best raw image data that we can, which will make the whole process easier. In this guide, I am assuming that you have a DSLR with a removable lens. Most of the advice also applies to film SLRs and digital compact cameras though, so don’t worry if that’s what you have.

Equipment

The lens has a lot more to do with the picture than the camera. For shooting the Moon, I recommend you use a telescope or a telephoto lens. The longer the focal length, the better. For most people, this means using a 70-300mm zoom lens, although if you have a 500mm camera lens, or a telescope which can be anywhere from 700mm up to 2000mm, then you can use that. On most crop-sensor DSLRs, a focal length of about 1400mm will make the Moon fill the viewfinder.

Tokina 400mm f/5.6 on Canon 450D

If you want to get a bit more reach, you could use a teleconverter. This fits in between your lens and your camera, and effectively multiplies the focal length of your lens – at the cost of losing some of the light and sacrificing some sharpness and quality. Common teleconverter sizes are 1.4×, 2× and 3×.

Kenko Teleplus MC7 2x Teleconverter

At these long focal lengths, the tiniest vibration will make your picture blurry. A tripod is absolutely essential – the sturdier the better.

Always use a cable remote to trigger the shutter without touching the camera. All SLRs support these, but most compacts probably won’t. If you can’t use a cable remote, the self-timer is your second choice. Set the timer, carefully press the button, and hope that the vibrations have died down by the time the photo is taken!

Canon Remote Switch RS60-E3

If your camera has mirror lock-up, you should always use it. This means the first time you press the button, the mirror flips up and the viewfinder blacks out. Then you wait a few seconds for the vibrations to die down, and then press the button again to fire the shutter. Mirror movements are a common source of vibrations and are probably the biggest cause of blurry Moon photos.

Technique

Forget autofocus. If your camera has a manual focus option, use it. It will be more accurate, faster, and will prevent your camera from re-focusing on each shot you take. If your SLR offers live view, use that and magnify the view if possible. Once you’ve set the focus, leave it alone :)

The Moon is bright (brighter than you think) and it is set against a dark sky. This really confuses the camera’s auto exposure, so it’s best to use full manual mode (usually marked M on your camera dial). You’ll need to tinker with the settings, but if you set the ISO to 200 then some reasonable starting settings for the full Moon might be a shutter speed of 1/250 and an aperture of f/11. Take a few snapshots until you get an exposure that looks about right. The most common mistake is to overexpose the Moon. We often think of it as being white, but it should be grey in the camera.

Keep your shutter speed fast. If you let it get too long, you will start to get motion blur. Shoot at least as fast as 1/250.

Choose a middling value for your aperture. Most telephoto lenses have a maximum aperture of around f/5.6 at full zoom. Usually this gives poor image quality, so it helps to stop down a few stops. Usually f/8 or f/11 is OK. Much smaller than that, and you start to lose sharpness again due to diffraction. If you’re not sure where your lens’s sharpest aperture is, check some reviews. Failing that, a good rule of thumb is the the sharpest aperture is 2-3 stops down from the widest.

Feel free to set the ISO as high as you like. You may be aware that higher ISOs cause more noise in the picture – especially in low light conditions. This is absolutely true, but in this case it doesn’t matter. We will discuss effective noise reduction techniques for astrophotography later in this guide. If choosing ISO 1600 enables you to keep a fast shutter speed and to use the sharpest aperture, so be it. Don’t be disheartened by the grainy pictures that come out – these are not the end product.

For reasons that will become clear in a minute, once you’ve found the ideal focus and exposure, you’ll need to take a few near-identical pictures. The Moon will naturally drift across the viewfinder – this is fine. You might want to place the Moon in one corner of the viewfinder and repeatedly take photos until it reaches the other side. Anywhere between 3 and 10 pictures is fine – just don’t forget that when shooting repeatedly, you still need to give vibrations time to die down after each mirror lock-up.

This next picture shows how fast the Moon moves across the sky. These exposures were taken just three minutes apart each, using a 300mm lens. Even leaving time to fiddle with the mirror lock-up between each exposure, you ought to be able to shoot at least one picture a minute, which will give you quite a few pictures at the end of the session.

Moon moving across the sky

Post-processing

Now that you’ve taken a handful of source images, we need to work on them to bring the best out.

Stacking

Stacking means taking a set of similar images, shifting and rotating them so they line up, and adding them together. This has the effect of averaging out noise from your camera, and distortions from atmospheric turbulence. The best free piece of software for Windows is called RegiStax. Those using Linux might want to consider ALE.

I wrote about ALE on this very blog not so long ago, but if you need a helping hand with RegiStax then I recommend you read this RegiStax tutorial. For the mostpart, you just follow through the steps it gives you – but there are a lot of scary options.

No matter which program you decide to use, after stacking, you will end up with a single image file which will look like a slightly improved version of a single frame. Now we move on to post-process this image in a more conventional photo editor. If you have Adobe Photoshop and you are familiar with it, then use that. I prefer to use GIMP which is similar to Photoshop but also free. It runs on Windows, Mac or Linux.

Colour channels

If you are shooting the Moon, it is effectively black & white, so we can do a trick with colour channels to improve sharpness at the cost of converting the image to actual black & white. (This doesn’t work if you want to end up with a colour photograph, by the way. If you’re shooting colour images of planets or similar, skip this step). Open your stacked image in GIMP.

Go to the Colours menu, Components submenu and choose Decompose. Make sure colour model is set to RGB, uncheck “Decompose to layers” and press OK.

This will split your colour image into its red, green and blue components, each of which opens as a new monochrome image. Now you can close the original image to save confusion.

Examine the three monochrome images you’ve got. They should be similar, but subtly different. Which one is sharpest depends on how much light pollution there is in your area, what colour it is, how well your lens/telescope performs at different colours and a million other factors. Zoom into each picture at 100% (do this by pressing 1) and have a look at the craters for comparison. When you’ve chosen the sharpest image, close the other two.

Unsharp mask

One of the best techniques for sharpening a slightly blurry picture is to use an unsharp mask. I won’t go into the theory here, but the basis of a blur is that a tiny dot becomes a small circle. Unsharp masking studies the image, and tries to convert the small circles of blur back into dots.

For this to work, we need to estimate the radius of the blur. Zoom into your image as far as possible (1600%) so you can clearly see the individual pixels as squares. Find an area of high contrast – either the edge of the Moon, or the edge of a crater. The edge of a crater or the lit edge of the Moon should be a sharp, defined line, but you’ll see that it is actually a gradual change, a few pixels wide. Count the number of pixels that it takes to cross the boundary.

In this example the middle red line (roughly) shows where the true edge of the Moon is. The outer two lines approximately show where the blur extends to. The distance between the two outer lines is roughly 5 pixels in this example. Work out the equivalent number for your image, and remember it.

 

Go to the Filters menu, Enhance submenu, and choose Unsharp Mask. Set Radius to the number you found in the previous step. Amount is set to 0.5 by default but you can change this if you wish. Numbers between 0.5 and 1.0 seem to work best.

Scroll around in the preview window to look at interesting parts of the image. Repeatedly tick and untick the Preview box so you can see what effect the unsharp mask will have. When you’re happy, press OK.

Colour curves

The name is a bit misleading – colour curves don’t have anything to do with colour in this context. They are a good way of enhancing contrast, though. Bring up the curves window by going to the Colours menu and clicking on Curves. The default “curve” is actually a diagonal line.

Arrange the curves window and your image alongside each other so you can see both at the same time. Drag the shape of the curve into a gentle S-shape. The exact shape and amount of the curve depends on your needs, but have a look at the next two screenshots to see what effect the curve has had. Click on the images to view them larger and use the arrow keys to go back and forth. You can immediately see that the dark patches are darker and the pale patches are paler.

Saving

That’s it! You’ve now finished all the basic editing in this tutorial. Save your image, but make sure you do Save As and choose a different name, so you don’t overwrite your original.

Summary

This guide touches upon a few of the most common techniques in astrophotography. It is by no means the ultimate guide. If you’ve got any questions, extra tips or if you spot any mistakes in this guide, please comment and let me know. Also, I encourage you to post your Moon photos at the bottom of this page, share your work and show off what you’ve done.

Further reading

On this blog

Elsewhere

Polaris, the pole star

Posted by on October 1, 2011 No comments

I’ve taken star trail photos before, but this time I made a special effort to include Polaris, also known as the pole star or the North star. It is very close to the celestial North pole, meaning all the other stars appear to rotate around it.

At first I was annoyed that two or three aeroplanes crossed the frame in the 45 minutes it took to make this picture, but actually I think the lower, continuous aircraft track adds something to the picture. I don’t like the dotted line, though.

Jupiter II

Posted by on September 29, 2011 No comments

Well after last night’s astrophotography session with a 400mm f/5.6 lens, and 3× teleconverter, I decided to try again with an even longer optical system.

I rummaged a 700mm f/11 telescope and used the same 3× teleconverter, for an effective focal length of 2.1m. This provided almost twice the magnification of last night’s setup, albeit with a much more unwieldy telescope which is extremely difficult to aim accurately, and takes about 30 seconds to stop wobbling after I adjust it.

Jupiter

As you can see, this image is larger than the previous attempt. If I want to improve further, I will need all or any of the following:

  • A lens with greater magnification, to achieve greater resolution on the sensor
  • A faster lens. At this magnification using a shutter speed any slower than 1/30 causes Jupiter to be blurred
  • A camera with a higher pixel density sensor, to achieve greater resolution
  • A motorised tracking tripod, to ease the process and to allow me to shoot slower than 1/30 and collect more light
  • A camera with a lower-noise sensor. I’m using a several-years-old Canon 450D, which is not renowned for its high ISO performance

All of these are expensive, so don’t hold your breath ;)

Jupiter

Posted by on September 29, 2011 1 comment

Last night the sky was clear and I noticed that Jupiter was shining brightly. I decided to try and take a photo of it. Being so far away, and not having a proper telescope, this is the best I could manage.

Jupiter

It’s blurry but it’s clearly recognisable as everyone’s second-favourite red planet ;)

If you want to see more and better pictures of Jupiter, send me some money so I can buy a telescope! :D

Crater face

Posted by on September 13, 2011 No comments

Sorry, I’m running out of original titles for pictures of the moon. Last time I posted one, I said it was the best ever. This one is loads better, though.

Composite of 4 frames, each shot at ISO800, 1/60s, f/11 on Canon 450D with Tokina 400mm f/5.6 lens and 3× teleconverter. Aligned and stacked with ale, green channel extracted, unsharp mask applied with radius of 20 pixels, and slight contrast enhancement. Simples!

My best moon picture so far

Posted by on August 15, 2011 1 comment

Usually when it’s a clear night and the moon is out, I’ll have a go at photographing it. Last night there was actually a full moon and it was the first clear night in a week or two, so I got snapping.

Technically, this is my best moon picture so far, using a Canon 450D camera, a Tokina RMC 400mm f/5.6 lens and a Vivitar 3× teleconverter – all mounted on a beefy tripod. I shot 5 frames and stacked them using a piece of software called ale. The overall result is that there’s lots of surface detail on the moon.

However, I don’t think it’s my best moon shot compositionally. When lit from the side, the moon’s many craters are well-defined and give a striking impression. Here, hardly any of the craters are visible because the moon is lit from the front. I’ll keep trying!

The moon again

Posted by on May 10, 2011 No comments

Yep, it’s the moon again – but this time through my new lens. It’s a Tokina 400mm f/5.6 prime for Canon FD mount, with 2× teleconverter, mounted on my Canon DSLR with a corrective adapter.

For those who don’t know or care about lenses – this is my zoomiest lens so far.

I also stacked a few separate exposures with ale to reduce the noise and increase sharpness. Other than this, there was very little post-processing.

All in all, this is an excellent lens :D