Debunking the myth about light pollution, and understanding how it affects SNR.

[@CuivTheLazyGeek]

Excellent video, well done, extremely well explained, AND it dismisses the "LP is good because read noise can be swamped" fallacy. NO! Dark areas ALLOW you to expose longer (and ideally you should), but you don't HAVE to! 30s exposures in Bortle 1 that don't swamp the read noise at all will still be a million times (* not to scale 🙂) better than 30s exposures in Bortle 9 that swamp the read noise completely (even assuming infinite well depth). It's exactly the same with narrowband filters - I often see wider bandpasses being suggested as better because brighter thus requiring shorter sub exposures... NO, narrower NB filters (dismissing relative velocity red shift and focal ratio blue shift effects - which compound each other despite the names) will still have all the signal, but less LP and thus less LP noise - you can thus expose longer (and ideally should), but you don't HAVE to!

[@Butzemann123]

Yesss this! The argument against narrower bandpass because of the "need" of longer sub exposure. I read that way too often. As if it would be a bad thing to have such low amount of LP noise, that your read noise actually becomes relevant. Its like having a small problem and instead of dealing with it, you just add so much worse problems, that the initial problem is irrelevant. Wait... thats how politic works 🤠

[@deepskydetail]

Thanks, Cuiv! This is a great comment!

[@deepskydetail]

@Butzemann123 And it's not like the read noise magically goes away either! It's still there with the LP!

[@Butzemann123]

Ive never understood why someone would believe that adding more random noise to random noise would make things any better. It could only be rooted in a misunderstanding. I think you did a really good job explaining how lightpollution actually works.

[@deepskydetail]

Exactly! Trying to overcome noise with more noise just isn't it.

[@gubigm]

That is because people cannot differentiate two phenomena: read noise and quantization noise. In signal processing theory there is a technique called dithering (no, this is not the same dithering we use to combat walking noise), that can help overcome quantization noise, and detect signals bellow 1 lsb. But in our cameras read noise (an analog noise added to the signal before quantization) is actually do the job for dithering. You don't need light pollution to do that.

[@khoatran9804]

Thank you SO MUCH for making this video. I live in one of the worst Bortle 9 zones on the continent, and I always joke that worrying about read noise in the city is like worrying about how loud your car stereo is when you’re driving on an F1 track.

[@deep_space_dave]

I think you read my mind as I have been trying to find ways to combat light pollution where I live which is about 20 miles from Chicago. I am wrapping up a crazy fast imaging project of M 81 where I have taken almost 80,000 subs for which I was only able to keep 50K. I wrote a python program to sort the subs by MAD, so I may have to start culling and keeping images with lower MAD. Just amazing to see the random noise dance around my faint object mocking me when I do a fast blink through my subs. Thanks for this great explanation on LP as not only do I have to deal with extinction and seeing, LP is the one noise to rule them all!

[@deepskydetail]

😲80,000 subs! Holy cow! What Bortle zone are you in? Are the faint areas starting to come out? I've got so many questions ;)

[@deep_space_dave]

@@deepskydetail My Bortle was 8.5 but now closer to 9 😞 The project is something I am doing for the PixInsight community to test their new FastImaging process. But now I am realizing that probably half of my subs are no good as they are contaminated with LP

[@whyf16uy]

When I grow up I want to be both entertaining and interesting like you.

[@deepskydetail]

1) There's no way I could build my own harmonic drive mount! So, if you were to grow up and be like me, you'd be going backwards! I can't wait to see your mount capture stunning images! It was so exciting to see it being tested :)

[@whyf16uy]

@@deepskydetailHaha. And you're nice too

[@petef.4361]

I am new into astrophotography, thank you for explaining about the histogram, it helps me picture in my head what it all means. Can you do a video about exposure time, and dumb it down for us? I keep hearing different things. Let's say I want to take subs of a bright object vs a faint object, how do you determine the correct exposure time? What if you put in a narrowband filter to get something like the horsehead, veil, or crab nebula, do you want to increase your exposure time, or does it not matter as long as the length of exposure time is equal? How does being in different bortle zones affect the recommended exposure time, like being in bortle 2 vs a bortle 6 area on the same target, using the same scope and filters? I hear some say it does not matter, for example if you take ten 5 minute exposures, vs fifty 1 minute exposures, as long as you stack them since that equals 50 minutes integration time total, but is that true?

[@deepskydetail]

Thanks! Good suggestion :)

[@Mr.Volcanoes22]

Another fantastic informative video as usual

[@deepskydetail]

Glad you enjoyed it!

[@sonofoneintheuniverse]

Happily moderate light pollution does not completely ruin astrophotography. Quite a few of us do not have alternatives as we live in real life and not in an idealized world. Anyway - great video debunking myths.

[@deepskydetail]

Agreed! Thank you!

[@abhijitjuvekar]

Regarding light pollution negation, I was thinking to capture totally out of focused LIGHT frames and use them as DARK frames in stacking program. Out of focused LIGHT frames gives you smooth light pollution sky profile and hence help to remove that 'unwanted signal' from your main data stack. I have not fully done it yet but seems it will work good for long focal length imaging like DSO galaxies compared to wide-field imaging. Have you ever thought on these lines ?

[@deepskydetail]

That's an interesting idea, and you can definitely try it. But I don't think it will work for a few reasons: 1) The true signal is still in each frame. You'll also be subtracting out that signal. Things will get messy. 2) Dark frames are good for getting rid of hot pixels, dark current or fixed pattern noise. A dark frame should not have shot noise in it (i.e., noise from your target or light pollution). An out of focus dark frame will have that shot noise. 3) As your target moves across the sky (or rather as the earth moves), the light pollution gradients are going to change. That's going to throw a wrench into a lot of things. I don't think the results will look good. But if you do try it, let me know how it goes!

Grat info, but as usual need to do some homeworks to fully understand enery concept and I like it !!!

[@deepskydetail]

Thank you! It does take homework for sure! It did for me!

[@TevisC]

Very well explained. Would you infer that a faster scope helps more in a light polluted zone than in a darker one? Example 100mm x 600mm F6 vs 600mm x 150mmm f4. The slower scope in a darker environment, the sensor has more accurate reading of the dso even though it's collecting fewer photons. The faster scope collecting more photons helps the sensor - stacking algorithm weed out the light pollution noise. (I suspect that even an F2 600mm fl scope in light pollution would still have trouble against an F6 scope in dark zones.. =[ )

[@deepskydetail]

My understanding (and I could be wrong) is that the light pollution will be problem with both scopes, but I'd still prefer the faster scope. The faster scope will collect photons faster (of course given equal aperture or equal focal length like in your example). But it will collect LP faster too. With the faster one, you can shoot shorter sub frames to get the same signal *and* noise as the slower scope (this is still good because you need less total integration time). But you'll still be limited by LP with both scopes. If the slower scope has a maximum exposure time of 60 seconds, then the faster scope will just have a maximum exposure time much less than that (maybe 15 seconds; haven't done the math). So, in theory the faster scope still wins. But the faster scope isn't going to get rid of the light pollution.

[@Guido_XL]

@@deepskydetail This is also the reason as to why the large professional telescopes moved to remote mountain peaks. Largeness does not beat light pollution. Of course, a faster scope with a larger opening will create a better angular resolution, increasing the spatial imaging quality, as compared to the scope with a smaller opening. That consideration is not about light pollution, but about the scope's performance that depends on its design.

[@Guido_XL]

This is often being summarised with the "gasoline filter" statement: the best astrophotography filter. Get into your car, with your astro-gear, and drive to a low-Bortle area. No other filter can beat that advantage.

[@4sapphireb]

I left a high paying job in a bortle 8 and took a pay cut to live 45 mins from bortle 1. I never looked back.

[@deepskydetail]

Very cool! Doing what you love to do is so important! :)

[@gregmac8268]

Again, as a dummy, you're info makes me feel smrt...

[@deepskydetail]

I hope it helps! And none of my viewers are dummies :)

[@Guido_XL]

Revealing the desire to learn means that you are not stupid, by definition. Stupidity feeds on ignorance, which amplifies itself. Genuinely smart people are not necessarily the ones that know a lot, but the ones that want to learn, regardless the social status that the knowledge may entail. So, if you feel a natural instinct to ask questions that increase your understanding, you are definitely not stupid.

[@tostativerdk]

Very nice video, keep them coming :)

[@EnvironmentalConservers]

Awesome video!

[@deepskydetail]

Glad you enjoyed it!

[@AmatureAstronomer]

Light pollution is bad.

[@syrupusurper3774]

"Light Pollution" is actual civilization, you know people not starving, being safer, not living in pre-industrial revolution poverty. Sorry that spoils your hobby.

[@DBFIU]

You ok?

[@Guido_XL]

Ehhhm, so, according to you, civilisation requires a side-product, like light pollution, since this is how we can measure how civilised we are? We could just become more aware of light pollution, being a waste-product that needs mitigation. We want to illuminate streets, yes, but not necessarily the environment. The ecological consequences aren't even mentioned yet in this context. Plants and animals rather experience a natural succession of night and day, in which artificial illumination has no place.