Hi folks,
A couple of weeks ago, during the new Moon, we were blessed with a really clear and frosty night so the telescope, camera, comfy chair and notebook computer were bundled into the back of the car and I drove to a spot several miles away from any street lights. With any large centres of population (Witney, Oxford, Cirencester Cheltenham etc.) well over the horizon I had real hopes of going deep so I left the Hutech sodium street light filter behind. Otherwise the set-up was as described before but, for new readers, here is a quick reprise.
My EOS 40D was piggybacked on top of the telescope. The telescope optics weren't used but this arrangement allows me to take advantage of the motorised mounting's ability to track the stars, so allowing longer time exposures without the star trails. The notebook computer was used to take advantage of the 40D's remote Live View functionality and the EOS utility software's "intervalometer" was used to run the exposure sequences. I used my 85mm f/1.2 lens wide open at f/1.2 and the camera was set to ISO 400 and a sequence of twelve separate four minute exposures were made of the constellation of Orion. Each exposure was followed by a three minute rest to allow the camera's sensor to cool down and air temperature started just above freezing and fell to several degrees below during the session, as my feet kept reminding me.
Of the twelve exposures six were taken of the top half of Orion and six of the bottom half.
It had been my intention to take another six exposures in precisely the same way but with the lens cap on. However, the cold proved too much for my notebook's battery and it died leaving me with just the one "dark frame".
Processing is critical to success for this type of shot. To illustrate this here are two thumbnails from the original twelve, unprocessed apart from resizing, with the top half of Orion on the left.
..
Not too promising at first glance and I confess that thought was partly responsible for the delay in processing the images, though I will also blame Christmas! That orange glow is the result of street lights many miles away and
not sensor noise. So the profligacy with which the local authorities waste electricity by trying to light up the Universe means I still have to consider using my Hutech filter, despite the fact that it introduces diffraction spikes into the image around bright stars.
Because I wanted to mosaic the top and bottom halves of Orion together I decided to start processing by opening each of the RAW files in Canon's DPP and correcting for lens distortion and vignetting before saving once more in RAW. I also did this with all of the ancillary files needed to complete processing using
ash's "Diamonds from the Rough" primer for Christian Buil's excellent
IRIS software. My first deviation from this methodology was to align the "lights" using manually selected stars. It takes more time but the results are more predictable.
With two images (top and bottom of Orion) produced from each stack of six it was time to do some serious image post-processing and here I stepped away from ash's primer. Although DPP had done an initial correction for vignetting the subsequent processing needed to bring out faint details means that more needs to be done. I followed the
BIN_DOWN method described
here using two passes. This was done for each of the two images. I then used the
Modified Equalisation function to enhance the faint nebulosity. The two images were then exported from IRIS as TIFF files so they could be opened in Photoshop CS2.
The first task in Photoshop was to use the
Levels function to get the background sky as dark as possible without losing any significant nebulosity. I then fired up its
Photomerge dialog and pointed it at the two files. The result was a
very pleasant surprise as not only did Photoshop do all the required geometry calculations but it also equalised the backgrounds. There are a few artefacts if you look very carefully but only the dedicated pixel peepers will be aware of them. With the mosaicked image assembled the final step was some intensive cosmetic tweaking to the colour and a mixture of sharpening and blurring. Here is the final result (apologies for exceeding the normal recommended image height but at least I kept the width down!):
Clickable if you want to view the full size JPEG (4MB)
Many of you will be familiar with Orion but with so many stars visible it isn't so easy to navigate so here is a handy image I've plucked from the Web:
Though the resolution in this map is necessarily poorer and not so many stars are visible it is possibly worth pointing out that there is slightly more of the red nebulosity visible. This highlights a problem with using unmodified DSLRs for astrophotography as the filter just in front of the sensor substantially reduces its sensitivity to the Hydrogen Alpha red emission line which is often a major feature in nebulae.
Bob.
Login
Register
FAQ
Search
and one of the test shots was a 30 minute exposure at f/16 and ISO 100 to capture some star trails.
In order to try and make a fair comparison I shot at f/16 and ISO 100.
and who want to try your hand at star trails I think that ISO 100 and f/16 are much too conservative 
