Astrophotographs

Evolving set of astrophotographs from various locations
All photos, Copyright (C) Bill Caid, 2013. All rights reserved.

Back to Bill Caid's Home Page

Background

I have always been fascinated by photographs of the world around us.  I think that is why I got into photography back in the late 1960s, back when film was king, before the PC and Photoshop. Taking most photos of the stars requires the ability to follow the motion of the sky.  It was only recently that relatively low cost hardware was available to accomplish this task.

Equipment

In the old days, film was used for photos.  Today, it is digital all the way.  I use a variety of digital cameras, but the one currently best suited for the task is my Olympus OM-D EM-5.  This is a micro four-thirds (MFT) format camera that has "live view" that allows focusing the camera/lens or camera/telescope combination with ease.  Autofocus, even the best autofocus, sucks in very low light, so being able to focus is critical.  The Olympus camera is much smaller, and much less expensive than my Canon but it has exceeded my expectations in almost every regard.

The other digital camera that I use is a top of the line (in 2008) 1Ds Mark 3.  The Mark 3 is a 21 mp camera that is built war-zone tough.  But, that toughness has a cost (besides the wallet-busting price) which is weight.  The build quality is top drawer, but the camera body alone is about 4 pounds.  Add a 3.5 pound lens on that, now you have some meat that you have to manage.  And the weight of the camera/lens combo is beyond the capacity of some of the low-end equipment for tracking (like the Vixen Polarie).

Lenses are a critical element to any photography action.  Make no mistake about it, there are OK lenses and there are excellent lenses and the difference is cost.  It is certainly true that there are some reasonable cost lenses that are sharp and produce acceptable photos, but that is the exception rather than the rule.  On my Mark 3, I use Canon L glass, their pro-grade lenses.  These are totally worth the money and I have gotten many amazing shots because of the ability of the camera to rapidly focus on objects.

On the Oly, things are a bit different.  There are very few really good native MFT lenses (my opinion), but there are many fully acceptable lenses.  Plus, the MFT camera design is such that it is possible to use existing excellent lenses from other cameras on the MFT bodies (with a lens manufacturer specific adapter).  When using these non-native lenses, you have to run the camera in either Aperture-priority mode of fully manual.  But, for astrophotography, that does not matter as you normally run the camera in manual mode anyway.  And, any auto-focus in the lens will not work either, but again a non-issue as it is virtually impossible to auto-focus on stars.  In my lens bag I have several Olympus native lenses that can be set to manual focus mode.  I also have a Voightlander 17.5mm that is fully manual and a Rokinon 7.5mm manual fisheye.  Finally, I use several Zeiss lenses migrated from my Leica M8 (28, 35mm) and a Canon FD mount 50mm and 300mm lenses.

Click here to see the tracking mount used for my Astrophotgraphs.


Technique

For the photos that I have taken, there are basically 2 different setups: guided and unguided.  For the unguided case, it further breaks down into two sub cases  that either assume that the motion trails are intentional (a star trail photo, for instance) or the exposure times are too short to be a factor in the final photo quality (for example a full moon photo).  In the unguided case, the world is simple.  Put the camera on a tripod or mount, point it at the target, determine the focus, aperture and exposure length and pull the trigger.  What results in the photo that you show to others.

In the guided case, things are a bit more complex.  First, you must obtain a sufficient polar alignment of the tracking mount.  This can be performed by a number of methods including use of a calibrated polar alignment scope, use of automated alignment software or the "drift alignment" technique.  Each of these techniques have advantages and disadvantages based on your specific objective.  Specifically, if taking short exposure photos with a short focal length lens, sloppy alignment will work almost as well as a precise alignment.  But, as either the focal length ("zoom") of the lens increases or the length of exposure increases or both, a high precision alignment becomes critical to preventing slurring due to motion of the stars relative to the camera.

To help reduce the precision tracking requirements, it is now possible (with the correct software application) to "stack" multiple images into one long-exposure image.  By taking a series of short duration photos of the same object and stacking them together, the effect of the motion slurring can be reduced or eliminated.  Stacking can help bring out subtle details not visible in any single photo while averaging out pixel noise resulting from thermal noise in the CCD imaging chip.

The Photos

These photos have been accumulated over the years on a variety of equipment.  Where known, I have listed the setup and location.  Not all photos are created equal, of course, as my equipment set, equipment and techniques have changed.  In the photos below, where appropriate, I have included the full-sized image, but set the display size to fit a normal web page.  By clicking on the image, you can see the full-size image. Enjoy.


In the 1999 time frame, I purchased a Meade LX-200 10" SCT telescope.  At the time (and even today) that decision was ill-conceived as the scope is big and difficult to move around.  And, given where we live, viewing conditions are never optimal due to marine aerosols and light pollution.  I also discovered that attempting to use the scope visually via the eye piece presents all sorts of discomfort ranging from assuming unnatural positions to suffering the cold and bug bites.  So, my solution was to get a camera to attach to the scope.  I did get an early, low-pixel count camera, but as always that generated other, more complex problems.  Technical issues aside, the camera did work, but obtaining critical focus was difficult.  Above is a shot of the moon that was made after a number of focusing iterations.

Managing contrast, focus and overall exposure proved challenging.  This photo shows less acute focusing and general under-exposure.

The CCD camera, from Santa Barbara Instrument Group (SBIG) required a computer for operation, so the simple act of having a computer in the field made things more complex.  Back then, camera operation was an iterative process that was both tedious and time-consuming.

Digital cameras became commonplace and taking photos became easier.  Above is a moon photo taken in 2007 in Seattle from the balcony of my apartment.  Canon 1Ds Mark 2 with 28-300mm lens.

20121231.  By some quirk of fate, I stumbled upon a blog that discussed taking photos that show star trails.  Reading the blog with interest, I attempted to take a star trail photo with my Olympus OM-D EM-5 digital camera.  The photo above was taken on New Year's Eve 2013 in Borrego Springs, CA.  Rokinon 7.5mm f/3.5 lens on EM-5 camera.  This photo is a "stack" (additive digital superposition) of 15 images, each 6 minutes long stacked using StarStax (a program specifically designed for creating star trails).  The large amount of light in the photo above is due to illumination of the rising moon.  Clouds were also coming over my position resulting in some odd effects.  The really bright track in the photo above is the planet Jupiter.   Also, the dotted light track was due to a plane that passed overhead during the photos. Other than the process of shooting and stacking a number of long exposures, the photo above was easy to produce.  The camera was on a fixed tripod and was not attempting to track the stars.  But, if you want to have the stars appear as points as opposed to smears, you must employ a tracking mount.

I researched and purchased a tracking mount for star photos.  I chose the Astrotrac mount because of size, portability and tracking performance.  We took the mount, cameras and lenses to Borrego Springs and stayed at a friends home.  In the front yard I shot this set of 6 235 second exposures of the Milky Way which were then stacked using PixInsight.  Olympus OM-D EM-5 16mp camera with Rokinon 7.5mm fisheye lens as mounted on Astrotrac tracker.  The light on the right is pollution that is coming over the Santa Rosa mountains from Palm Springs, CA.  I processed this set of images "hard" with PixInsight to reduce the effect of the light pollution.

While the Astrotrac was working on a set of images, I also set up my Canon 1Ds Mark 3 21mp digital camera with 28-300mm lens to attempt to capture some star trails.  The capture worked, but only after the fact did I discover that I had programmed the remote controller incorrectly resulting in interruptions in the trails.  But, the resulting stacked photo was still interesting.  The small city in the foreground is Borrego Springs, CA.  The light coming over the distant mountains is from Palm Springs, CA.


I discovered a web site that specializes in the sale of used camera equipment.  The Olympus OM-D EM-5 I had purchased some months earlier had an interesting property that it could accept, with proper adapters, a wide variety of legacy lenses from many manufacturers including Leica, Zeiss, Nikon and Canon.  I found an excellent deal on a used Canon 300mm FD fully manual lens and purchased an adapter for the FD mount.  With the 300mm lens on the Oly, the resulting "real" focal length is 600mm due to the 2X "crop factor".  To test the camera-lens combination, I setup my tripod in my front yard and got the shot above.  The photo above is a crop of the source photo and is shown at pixel level resolution.

Over New Years in the desert, I got the photo above, again with the Oly/300mm Canon FD lens combo.  The brightness of the full moon makes resolution of the fine detail of the moon's surface difficult and therefore the focus was not fully critical.

In Borrego, using the Astrotrac mount, I got the photo above using my EM-5 camera and Zeiss 28mm M-mount lens with adapter.  The photo is a stack of 4 six minute exposures.  The crop factor of the EM-5 makes this lens appear as a 56mm focal length lens.  Orion's Belt is in the center of the photo with the yellow star to the top being Betelgeuse.  The bright star at the bottom center is Rigel.  Just a bit less than 1/2 way between Orion's Belt and Rigel is the Great Orion Nebula (the fuzzy object).

.

A crop of the photo above shows pixel-level details not easily visible in the full sized photo.  Note the Horsehead nebula present near the left-most star in Orion's belt and to the left of center.  The Great Orion Nebula is the bright area below and left of center.

Upon seeing the photo above on my laptop, I elected to switch to the 300mm lens for a closer look.  While the longer lens did provide more magnification, it also pointed out that to use long lenses on the Astrotrac, you have to be precise about polar alignment of the device.  Given the slurring in the image above, it is clear that my alignment technique needed to be improved. 


The image above was taken with my Voightlander 17.5mm f/0.9 lens on the Oly EM-5.  This shows a full frame image that was an integration of 5- 235 second exposures.


This is crop of the prior image with emphasis on the galaxy.


20130130. A shot of M42 was taken from my parent's place in Tucson (in the middle of the city with plenty of dirty light pollution) using the Astrotrac, Oly EM-5 camera and an Oly 14-150 f/4 zoom lens.  The photo above was a stack of 12 2 minute shots at ISO 400, f/4, 150mm (300mm actual) and then stacked with PixInsight.  Now, this image is not too bad, but it does show that the quality of the lens is the real issue.  There is some star slurring mostly visible at the top, but the bottom stars are not distorted.  Also note that the diffraction "star patterns" show the lower quality of this len.  Contrast this with the Zeiss lens several photos above.  That said, the camera/lens/mount combo did what I asked and produced a respectable photo.  I also know that the focus as not fully on the money which is something that I attribute to nature of the lens.  Note that the capture process did resolve some color in the nebula.


20130130.  The photo above consisted of 72 3-minute exposures at ISO 200 and 28mm.  As shot from the roof deck at my parent's place in Tucson.  The stack was created with PixInsight using the "max-additive" integration method.


20130302.  The photo above was taken at a SDAA Tierra del Sol star party and is only one of several that we took.  You can see both M42 and M43 in the photo above.  180mm Leica lens on Oly EM-5 camera.  240 seconds at ISO 400.


20130302.  This is an integrated stack based on the TDS photos.  There is not much, if any, additional detail in this photo when compared to the one above it (which is un-integrated).  180mm Leica lens on Oly EM-5 camera, 240 sec at ISO 400.  Note the chromatic abberation (purple fringes around the really bright stars).


This is the Horsehead nebula, but not a very good image.  This was taken with a Canon 100mm f/2.8 lens at 2.8, 240 sec at ISO 400 with Oly EM-5 camera.  The purple rings are chromatic aberrations in the lens.

This image was just so-so but it did capture a meteor that flew by us while we were shooting Orion.  It is the green streak at the left of the photo above.  The red streak at the bottom was a reflection from tail lights of a near-by car.

20170515 - Given the upcoming solar eclipse, I got the wild idea to get a solar telescope.  I got a Lunt Solar Systems LS60 scope.  I first tried a Coronado, but was displease with the focus acuity and a few other aspects of the scope, so I exchanged it for the Lunt.  The Lunt is a much better designed and engineered scope.  There was (is) a rather steep learning curve on this setup and inititally I had "plumbing problems" which were purely mechanical - things like adapters for camera mount, correct adapter lengths, etc.  But, with some resolve and a credit card I was able to get the adapters that will work with the newer mirrorless cameras.  Standard DSLR hardware will not allow the camera to focus as it is behind the critical focus point of the scope.



This is one of the first shots to come from the Lunt mounted on the AstroTrak mount.  Olympus EM-5 ISO 200 at 1/1600 shutter speed.  This is an unstacked photo and was intentionally underexposed to show the structure of the chromosphere.  The dark spot is a sun spot, and these move around day by day, sometimes hour by hour.  20170520



The shot above was also taken with the EM-5 camera, but in raw rather than jpeg.  The additional detail shows the solar flares on the surface of the sun.  Look at the edge of the solar disc at about the eight o'clock position.  Keep in mind that this flare is probably 20 earth-diameters tall.  Sadly, when there is enough light to see the flares clearly, the center of the disc is nearly over-exposed and saturated.  No amount of fancy digital photo processing will recover saturated portions of the photo.  20170527.



Sony A7RM2 jpeg mode.  20170526.  Note the large flares at the 6 o'clock position and more around 11:30.



Sony A7RM2 in raw mode 20170526.  This shot was intentionally underexposed to provide details of the surface structure.  Note the additional sun spots.



Olympus EM-5 in raw mode.  20170527.

As I said, the learning curve is rather steep.  But now that the attachment plumbing problems have been addressed, what remains is to get a few sunny days (not so easy in San Diego in May) and try a number of combinations of over/under exposure with the Sony and the Olympus cameras.  Then, when I have that addressed, then attempting to do the same thing with a 2x Barlow lens in the optical path.  Finally, coming to terms with the various stacking programs that are available (like PixInsight, Registax and Autostakkert).  Stacking should allow elimination of the fuzziness in the images that are due to atmospheric scintillation.  The good news is that Registax can process video and use the individual frames as the basis for stacking.  The bad news is that the video resolution is so much worse than a still photo. 



It seems that the crux of this whole affair is the processing chain that is used.  This shot was originally a set of Olympus raw files (ORF) that were handled by PIPP.  PIPP then crops the images based on object detection and wrote a set of FITS files (to preserve the dynamic range).  These 30 files were then processed by Registax to perform the stacking, wavelet enhancement and histogram normalization.  The original files showed streamers, but these were either averaged out or I used the incorrect set of parameters.  The output file is 1500x1500 pixels which is larger than my (self-imposed) 1200x1200 maximum so the browser reduces the resolution for display but the full sized image is served during page load.



This was taken with my Sony A7R2 using the Astrotrac mount at ISO 100 at 1/40 second exposure.  Stacked with Autostakkert2.  Note at the 7 o'clock position a solar flare is visible.  This was a monochrome conversion done by PIPP at the front end of the processing pipeline.



This is a re-process of the Olympus ISO 200 at 1/800 second exposure.  Stacked with AS2.  The Olympus has less than 1/2 the pixels of the Sony.  This photo is color with RGB channels histogram normalized individually within PIPP.



This photo was also from the Sony but was 2x2 binned.  The Sony has sufficient resolution to allow binning which reduces each pixel dimension by a factor of 2.  Note that some solar flares are visible at the 9 o'clock position.  Preprocessed with PIPP to do the binning, stacked with AS2.



20170613 - The photo above was produced from a stack of 120 photos from the Sony A7RM2 taken at ISO 200 @ 1/100 using a 2x Barlow extender.  The weather was clear, so good resolution was obtained.  Note the large flares at the 9 o'clock position.  PIPP was used for photo preprocessing producing a 3000x3000 pixel image that had independent normalization applied to each color channel.  Autostakkert was used for stacking and sharpening.  20170613.




The photo above is a crop of the previous photo to emphasize the solar flare.



20170619 - We had a clear day yesterday so I thought I would grab a solar photo to see what was going on.  This image is the result of a stack of 200 photos from my Sony camera, ISO 200, 1/100 exposure in RAW mode.  Oddly, I did not have high expectations for the photos because it was quite hot and the heat shimmer was significant.  Each individual photo was blurry and taken as a single photo, I would have tossed them out.  But, when combined as a stack, the atmospheric variations average out and you can “see through” the shimmer.  Note that the quantity and location of the flares have moved; they change day-day, hour-hour.  The dark spots move as well due to the rotation of the sun, and like the flares, they come and go.  Preprocessed with PIPP to cropped TIF files then stacked with Autostakkert and finished with Capture One.



20170619 - The second photo is a zoom-crop of the first one.  Note that there are two kinds of flare in this photo, the first has loops due to magnetic field lines and the second has huge, wispy plumes of ionized, glowing gas of a slightly different color than the balance of the flares.  These flare are measured in “earth-diameters” in size and are multiple diameters high.


Hope that you enjoyed these shots. I will add more images to this list as they are taken.  We are looking forward to the solar eclipse on 21 Aug 2017.


Back to Bill Caid's Home Page