This was my first chance to image comet Lovejoy (C/2014 Q2) - well into March and long after its best - but its never too late. (Too cloudy much of the time and no access to my observatory in February)

The field features the open cluster NGC 457 on the right (known sometimes as the ET cluster) and on the left is the unusual asymmetric planetary nebula Sharpless 2-188 (recently nicknamed the Firefox nebula ). The expanding gas from the planetary nebula is colliding with ambient gas in the interstellar medium. The nebula is actually circular in shape but is much brighter to the southeast (lower left) because the central star is moving rapidly in that direction. The upper right side wasn't captured in my image.

-

Note on update B: I switched blending mode of the comet layer from "screen" to "lighten" to reduce the streaking in the sky background due to star trail artefacts in the comet layer. I don't like "lighten" blending as its not a proper colour mixing mode but the streaking wasn't pretty so I compromised. Also I added a radial blurred difference mask to selectively sharpen the edges of the ion tail. The radial blur was set at 20 degrees and centred on the comet head. The difference image was biased to mid-grey (brightness of 128 on the 0-255 scale) and mixed with the main image using "overlay" mode. This is the same technique used to enhance solar eclipse images and sharpen up the radial coronal streamers.

[color= #87CEFA][TUTORIAL: How to de-streak a comet aligned stacked image][/color]

Note on update C: I've developed a reliable way to reduce the residual star streaks in the "comet only" (CO) layer. This is all performed in Photoshop but can easily be translated to other applications if a dust and scratches filter or equivalent is available. The other operations are just straight forward pixel math. The final blending mode for the comet only image has been restored to "screen" mode.

Duplicate the original CO image and apply a dust and scratches filter to reduce the star streaks. The parameters should be adjusted so that the streaks are blurred far more than the comet features - i.e. keep the filter radius low enough for the residual star streaks but not too high that the comet tail streamers get too damaged - this gives a filtered CO image (FCO).

Now subtract the original CO image from the FCO image to get a new image that contains mostly only the streaks (the S image). Due to the blurring effect of the filter the comet coma will leave a residual too - this can be cloned out with background if desired if no obvious streaks cross the coma area, to ensure the coma is fully preserved.

[color= #87CEFA][TIP: This "streaks only" image will be very dark. I suggest that you place a temporary levels layer on top and adjust it to boost contrast if you wish to work on removing the coma artefact in the S image. Pixel maths-wise we have S = FCO - CO][/color]

Now subtract the S image from the original CO image to create a de-streaked comet only image

(DCO). This now greatly lessens the residual star streaks leaving only faint blurred remnants.

[color= #87CEFA][Pixel maths-wise we have DCO = CO - S][/color]

The final de-streaked comet only image (DCO) now has the black point set in the "levels" tool so the that darkest point in the background is set to RGB (0,0,0). It is then copied over the "stars only" image and set to "screen" blending mode. Finally another "levels" layer is placed on top and the black point again adjusted but this time so that the background sky is set to RGB (20, 20, 20)

[color= #87CEFA][So the overall pixel math is DCO = CO - (FCO - CO)][/color]

This extends the Hubl method used for removing comet remnants from the "stars only" image* to an additional process for removing the star remnants from the "comet only" image by applying a narrow linear feature filter and applying a back subtraction technique. This is effectively a description of a novel "Double Hubl" method.

* (Hubl uses Gaussian blurring and clone-stamping to remove stars followed by back subtraction to remove the motion blurred comet)

- [color= #87CEFA][Final adjustments][/color]

Note on update D: The de-streaking actions described above softens the edges of the tail streamers. This effected has been mostly reversed by the following two actions.

First a de-speckle filter was applied to the whole image to remove fine grained dark pixel noise introduced during the streak subtraction. This is important as it prevents noise enhancement by the following step.

Second a masked local contrast enhancement (LCE) pass was applied to just the tail streamers were they emerge from the coma. The rest of the image was masked to prevent LCE being applied elsewhere where it mostly gives stars far too sharp edges.

Notes on update E: A star mask was applied and a little further colour saturation added to the stars. A 50:50 blend of image D with a vertical de-banding filtered version of D was used to generate E to remove the small streaking residuals that could be seen in the comet tail.