JULY 2023

Untracked Imaging from a dark site

In October 2021, I spent a week in the Isles of Scilly totake part in the COSMOS ‘Scilly Dark Skies Week’.  The islands must have one of the darkest skiesin the UK and we did have some clear, transparent, moonless nights with thenorthern Milky Way from Perseus to Lyra stretching overhead towards midnight.

I had taken my Sony A5000 camera (see article in Digest)along with a superb Zeiss 45mm, f/2, legacy lens used with an adapter. Importantly,this lens is totally rectilinear and does not suffer from distortion – veryimportant if a set of untracked images are to be taken.  As the A5000 had an APS-C sensor with a 1.5crop factor, the effective focal length was 67.5 mm.  The focal length is significant as I did nothave a tracking mount with me and so was taking untracked exposures. Thislimits the exposure time allowed for the taking of a single frame if startrailing is to be avoided.  The formulaoften used is to divide the effective focal length into 500.   However, the trailing is greatest near theCelestial Equator and is least as one approached the North Celestial Pole.  So it is possible to include a ‘Cos Dec’ termto reduce the effective lens focal length. The regions of the Milky Way that I was aiming to image were around Dec45 degrees. The Cos Dec term is ~0.7,  soreducing the effective lens focal length to 47.25 mm and thus 10 secondexposures would be appropriate.  With thelatest high resolution sensors it is suggested that 300 should be used ratherthan 500 and, in the resulting images a touch of star trailing could be seen whenviewed at 100%.  As described below, thisis easy to remove.

In untracked images, as the sky rotates against the sky someof the outer parts of the image will not receive the full illumination that thetotal exposure would provide.  It maythen be necessary to crop the image down somewhat. The use of  untracked imaging may well not work so wellusing very wide angle lenses if they are not totally rectilinear – as most arenot – and blurring of the outer parts of the frame may result so some severe croppingcould well be required.

It is difficult to take a separate set of dark frames with aDSLR or mirrorless camera as the sensor temperature changes as the images aretaken.  When exposures longer than a fewseconds are taken, these cameras will usually follow a  light frame with a dark frame.  In most cameras the taking of these darkframes can be disabled so spending more time imaging the sky but then ‘hotpixels’ may be visible which the dark frames would have removed.  However, for these images, the taking of thedark frames was enabled.

The general advice when astroimaging is to save raw images,but the use of JPEGs will greatly reduce the data storage required.  A single raw frame will have a greater bitdepth than JPEGs but, if many frames are stacked (as one will do when usingmany short exposures to image an object), then the effective bit depth isincreased and there may be little difference.  (In fact, as the internal production of JPEGswithin the camera may involve some stretching and in some circumstances theresult may actually be better.) In this case JPEGs were saved.

I imaged two areas of the Milky Way; one including theconstellation Cassiopeia and the other Perseus. In both, the star colours showed up nicely.  The sky background is not uniform showing thevarying dust regions in the Milky Way. ‘Background Extraction’ would haveremoved these variations and was not used – the site had virtually no lightpollution and is rated as Bortle Class 1.

For the Cassiopeia image I was able to stack 56 frames.  There may have been some water vapour in the atmospherewhen I took these as the brightest stars had become somewhat diffuse – nicelymaking them more apparent in the image. For this image, nothing was done except to stretch the image and removethe hint of star trailing. 

For the Perseus image I was able to stack 65 frames.  The double cluster shows up well as does thelarge open cluster NGC 1348 that surrounds Mirfak, Alpha Persei.  Again, for this image, nothing was doneexcept to stretch the image and remove the hint of star trailing. 

Removing the hint ofstar trailing in the two images

In Adobe Photoshop or Affinity Photo, the image isduplicated and the blending mode set to ‘Darken’. Using the ‘Move’ tool, theupper layer is moved, pixel by pixel over the base layer and the stars will beseen to become circular.  For precision,it may be worth first doubling the size of the image before in order to makefiner adjustments.

There is one proviso. If the image contains nebula regions, this process may cause artefactsand it is then best to separate the nebula and stars using StarNett++ and onlyapply the technique to the stars image before combining them.

© Ian Morison