- This topic has 5 replies, 4 voices, and was last updated 2 weeks, 5 days ago by
.
-
Posts
-
Attached are examples of the effect of the numerous satellites that leave light trails on long exposure astronomical images.
Micheal Jaeger posted this exceptional image of C/2025A6 Lemmon taken on 20251026. It is a mosaic image showing the total extent and structure of the ion tail.
The first image shows the image marred by the satellite light trails and the second image is the result of 6 hours work on the image to digitally remove the satellite trails.
I know there are some who argue that the ability to digitally remove these trials means that the problem is diminished, but as the number of satellites increases the work to digitally remove them is going to become increasingly difficult, if not impossible.
What a shame that for all the good that technology can bring it comes with a huge downside for astronomers and epsecially imagers.
Micheal has given us permission to publish his images within the BAA.
Here are his comments:
After six hours of work removing the satellite tracks, we can now show the finished result from Sunday evening.
The tail shown here is longer than 15 degrees.
C/2025 A6 Lemmon 2025-10-26 18.00 UT mosaic 15/15min filter blue Leica-Apo-Telyt f-180/4.0 QHY600,
Michael Jäger, Gerald RhemannMichael’s images are fantastic but I wonder why he doesn’t take more, shorter exposures although I’m not clear what his sub exposures are. It says 15×15 minutes in your description but that is almost four hours so that can’t be right. CMOS sensors have very low read noise and can be read out very fast so doing short exposures is not normally a problem and it allows you to stack statistically to remove sub frame pixels affected by satellite and aircraft trails. My image from here last night:
https://britastro.org/observations/observation.php?id=20251027_224004_df7ddf0b98656b8c
had multiple satellite trails per frame but none are visible on the final stack of 21x30s frames.
Jeager and Rhemann are undoubtedly the best comet images in the world so I must be missing something about their technique. It is unfortunately the case that comet imagers are badly affected by the mega-constellatons since we are imaging with wide-field, fast optics in twilight but there are solutions which don’t involve manually removing trails from each image.
This was discussed in my talk at the recent comet section meeting around 26 minutes in here:
https://britastro.org/video/2025-10-04_Comet_Section/03%20-%20Edinburgh_NDJ.mp4
Hell, that is outrageously good and the stuff at the dim end of the tail must be seriously dim. Its a lovely piece of work.
So what are they doing that everyone else isn’t?
Can I assume they take superb flatfields, remove the stars, remove sky gradients and then work on the tail with a very non-linear stretch and noise suppression before draping it back on one of the original images?
Perhaps they takes the longer exposures to catch enough signal from the sky (that the tail is similar in brigtness to) so when they subtract the sky theres enough signal left to avoid it being horrible grainy.
I sort of assumed they did short exposures for the bright in-close fine detail and long exposures for the extended tail…
Sneaky is best and all that.
They certainly do produce extraordinary finished images and they are far better than anything that we have seen in the past so I think we can conclude that the “good” tech (sensors, mounts, optics and processing software) is currently beating the “bad” tech (LED lights, LEO mega-constellations). The comet imagers of old would have been astonished by the quality of images produced today.
We have to live in the world as it is rather than the world we would like so dealing with mega-constellations is, for now, just another hassle, like the weather or light pollution. If these images are processed by separating the comet from the star background, processing each one differently, possibly with some clever, statistical noise reduction and then putting them back together then using statistics or AI to remove the satellite trails is just another step.
If these images are processed by separating the comet from the star background, processing each one differently, possibly with some clever, statistical noise reduction and then putting them back together then using statistics or AI to remove the satellite trails is just another step.
Not science though without the raw data and enough details of the processing steps to allow others to reproduce it.
Robin – I would agree with that but I don’t think that “science” is the main objective in these cases. Jager and Rhemann are very careful with what they do but there is always a concern that so many complex steps can lead to artefacts. I don’t think that this is the case here but I have seen plenty of examples online from less experienced imagers.
Personally, I think the whole idea of separating the comet from the background and then processing each one separately is a bit dodgy but I know that others will have a different opinion!
- You must be logged in to reply to this topic.