Starlink Satellites

Ha, yes it would and it already exists. 😉 You can get rid of them by using the normal clipping algorithms.

Vincent

Do you think the clipping algorithms will cope with 30,000 of these things filling the sky? My real concern is that there are only 9000 satellites excluding Elon's up there now and could we be be at the beginning of the end of ground based amateur astrophotography? 

@chrispeace I'm confident with enough exposures the clipping algorithms will do the job. The likelihood of multiple satellites ending up in most of your subs on the very same track is virtually zero if you have more than a few exposures. Crossing satellite tracks could perhaps introduce star-like features on the crossing point if you exposure count is too low (3 or so). Just make sure you have at least a handful of exposures, clipping will do the rest.

Happy Days as I quite like what Elon is doing for the world apart from this but I'm reassured that I don't have to go sell my gear just yet. Many thanks for the reply at such a late hour.

Also keep in mind due to the low altitude, it's highly likely that starlink will be completely invisible an hour past sunset. 

@chrispeace I'm not 100% sure it won't have any effect, but I think the clipping algorithms will work just fine. It's a lot of satellites, but they still will be quite far apart when looking at the sky. They will criss-cross through the frames, all at different positions all the time, which will be clipped during processing. You may have to take a bit more data then usual, maybe. And they should be almost gone when the astro-dark time sets in anyway. I think the biggest worry is for radio-astronomy.

@vincent-mod just in case better removal would be needed, here's some thoughts I've had on the topic. I noticed in my exposures at extreme stretch levels that there is just a tad of increased brightness in some cases when satellite or airplane trails were removed. Of course there's always a level where the data are barely not clipped by the outlier rejection. For an "advanced" satellite and airplane trail removal one could identify these trails in the usual way, perhaps use a Hough transform to distinguish between straight lines and other shapes (cosmic rays), and then use a Grow operation on the selected pixels to also clip adjacent pixels to  much more stringent level (say, 1.5 sigma).

Just a thought.

Sounds very interesting, I'm no expert in these kind of algorithms so that one is for Mabula. Thanks for that input! @mabula-admin

Posted by: @ralph

@vincent-mod just in case better removal would be needed, here's some thoughts I've had on the topic. I noticed in my exposures at extreme stretch levels that there is just a tad of increased brightness in some cases when satellite or airplane trails were removed. Of course there's always a level where the data are barely not clipped by the outlier rejection. For an "advanced" satellite and airplane trail removal one could identify these trails in the usual way, perhaps use a Hough transform to distinguish between straight lines and other shapes (cosmic rays), and then use a Grow operation on the selected pixels to also clip adjacent pixels to  much more stringent level (say, 1.5 sigma).

Just a thought.

Hi @ralph , @beercia, @chrispeace, @kingjamez,

For the moment, the adaptive outlier rejection algorithm should be able to remove the satellites if you shoot enough frames (more than 20). Even if you have many in many frames... It becomes a problem though if the satellites would show themselves in most frames on exactly the same pixels. And the chance of that happening is in fact very, very small !

So I would not be too concerned at the moment about this. But I agree, if several other companies also start polluting our skies with 10000s of satellites, things might become awkward at some point... so I will definitely keep this in mind and at some point I will try to implement a feature rejection algortihm like Ralph suggests.

Kind regards,

Mabula