Hacker News Reader

Comment by invisible | Hacker News Reader

On the face of it, a composite image of space that has no adjustments being ruined by 2x the number of satellites does seem pretty bad and damaging for researchers. Beyond that, it still might be worth it to actually be able to provide internet to people that either don't have access, have an insane cost to access, and/or have a bad provider currently.

The argument I've seen is that this stuff can be adjusted for and that most researchers have already had to adjust for it (since there are ~1500 non-Starlink satellites already). Is that just an utter fallacy?

View on HN

I turned up the person who actually made the image, and their choice for how they made this image makes me confused:

https://twitter.com/GrumpyOldAstro/status/139637061036883149...

Essentially, they chose to multiply any light from all of the frames.

(Disclaimer: I know ~nothing about astronomy)

The other claim I've seen is that once the satellites reach their operational orbit (~4 months after launch), they are no longer visible.

I don't know if this has been proven/disproven yet. Curious if anyone in-the-know can comment.

They are no longer visible to the naked eye. The satellites are initially inserted at ~half their final orbit. They then boot up, deploy their solar panels horizontally to reduce drag, and then fire up their engines for the boost to their final orbit. This takes about a month or so. Once there, they rotate their panels to point away from the earth, which greatly reduces their apparent brightness.

You're correct that they are no longer visible to the naked eye. They're reduced down to (right now) approximately 6th magnitude and SpaceX is trying to get them down to 7th magnitude. The human eye in a perfectly or almost perfectly dark sky can see down to roughly (humans vary quite a bit) 6th magnitude but it would be very hard, especially for a moving object.

In order to see 6th magnitude you need a very dark sky (no moon or clouds, low humidity, no human populations anywhere nearby), you would need to dark adapt for over 30 minutes with no viewing of any sources of light, you'd need to be pretty young, and then you'd need to use averted-vision viewing where you look near an object, but not directly at it such that you can use the higher density of rods in the rest of your eye in order to see the object. You'd then have a chance of seeing current Starlink satellites, but you'd see a ton of other satellites as well as there's many satellites above 6th magnitude. (In 2017 or so I was in rural Nebraska the night before the solar eclipse and it felt like there was a half dozen visible satellites in the sky at any given time.)

However that doesn't mean they're not visible to telescopes.

Image processing is not magic.

Obviously you can remove the satellite traces from the image, but each of them causes a degradation in quality, i.e. a reduction in the signal-to-noise ratio, because the light coming from the sky zones obscured by the satellites has not been captured by the sensors.

Regardless how well you might process the sky images, the large increase in the number of satellites will reduce their quality.

It makes sense that each frame would suffer from signal-to-noise problems and be degraded, but given that this image was created over 2 hours -- is there really any loss in quality if sampled correctly? I could honestly believe there are a bunch of complexities beyond this, but: is this already having to be done _anyway_?

> Regardless how well you might process the sky images, the large increase in the number of satellites will reduce their quality.

Right, this is the true statement of the issue. However is a _marginal_ loss in image data quality really not that bad a thing to ask for for getting internet across the planet effectively (in human history terms) overnight?

Reply on news.ycombinator.com