Light Pollution Advice

[Arran townsend]

Hi all, I have recently gone to the trouble of building my own computer controlled barn door tracker, which I'm happy to say works well even with a zoom lens, anyway here in the UK near Birmingham light pollution is terrible. I recently stacked 138×20 second exposures of the andromeda galaxy, which worked great by far my best photo yet, but I'm somewhat disappointed that 2/3 of the galaxy is missing as a result of bad light pollution. 

What I'd like to know is, is it worth taking hundreds of light frames to resolve the fainter detail or is this simply not possible, I read somewhere that you can't get any light fainter than the light pollution it's self, therefore pointless trying. But others say hundreds of exposures is the way to go with light pollution.

Can someone guide me on what I should do tips and tricks would also help as I'm a beginner and willing to learn thanks. Arran

 

 

[cletrac1922]

Arran

I do not have your problem as much were I live

Few weeks ago, one of our club members gave a talk on narrowband filters, and light pollution

Badder make a selection of narrowband filters for AP

https://www.baader-planetarium.com/en/blog/baader-narrowband-filter-why/

John

[vlaiv]

2 hours ago, Arran townsend said:

I read somewhere that you can't get any light fainter than the light pollution it's self,

That is simply not correct. You can get decent images from heavy LP if you get enough total imaging time. I'm guessing you took above image from Dudley?

Here is what andromeda looks like with ~20h total exposure from about same LP levels (maybe tiny bit less LP, but very small amount):

As you can see - capturing faint detail was not a problem in this image. Granted, above image was captured with scope and cooled dedicated mono astro camera with filters, but still same applies - get enough exposure and faint details will show.

Here are couple of pointers.

When did you take above image? Is it recent work or some time ago? Always go for objects that are near zenith or crossing meridian at the time of your session. Don't go for objects too far east or west. Image "seasonal" objects.

If you have more LP in certain direction - East for your location as Birmingham is that way - avoid shooting in that part of sky - wait for target to clear the worst LP.

What sort of camera and lens are you using? Your image looks like it was processed in black and white, but I'm guessing you are using a sort of DSLR that should be color? Did you process your image properly? What stacking software did you use? Did you record raw files or have you used jpeg compressed ones?

Use raw / unaltered files for stacking. If using color camera make sure you get bayer matrix settings right - this will impact color if you get it wrong. Use fast lens and make it wide open (don't stop it down) if that lens gives you acceptable stars. If you have to stop it down to get sharp stars - do it minimally. Maybe consider getting better / faster lens if your lens is too slow.

What settings are you using for your DSLR? You need your ISO to be about 800, but that will depend on camera model.

[Demonperformer]

Stretching your image, it looks to me like it was positioned badly ... you seem to be imaging right above a light source? If that is the case, then I would have to agree with vlaiv that choosing your target for its position is the way to go. Living along "streetlight alley" I feel the frustration that this can be and have tried many things. You can "tinker around the edges" with things like a decent dewshield (which also helps reduce stray light), but the bottom line is that trying to image things with a light source nearby is always going to be frustrating.

Narrowband will deal with the light pollution, but will be useless on objects like galaxies (but are great for emission nebulae).

[almcl]

An alternative to shooting at home if you have transport, is to head west from Dudley to the blue zone in the image below.  If you can do this the skies there can be quite amazing.

 

[alacant]

6 hours ago, Arran townsend said:

tips and tricks

The easiest way would be to use a light pollution filter e.g. this one.

HTH

 

[Demonperformer]

LP filters depend very much on the sort of LP you have. The old orange sodium streetlights were very effectively blocked. The LP from the  new LED streetlights and private "insecurity lighting" is almost unaffected by them.

[Arran townsend]

12 hours ago, vlaiv said:

That is simply not correct. You can get decent images from heavy LP if you get enough total imaging time. I'm guessing you took above image from Dudley?

Here is what andromeda looks like with ~20h total exposure from about same LP levels (maybe tiny bit less LP, but very small amount):

As you can see - capturing faint detail was not a problem in this image. Granted, above image was captured with scope and cooled dedicated mono astro camera with filters, but still same applies - get enough exposure and faint details will show.

Here are couple of pointers.

When did you take above image? Is it recent work or some time ago? Always go for objects that are near zenith or crossing meridian at the time of your session. Don't go for objects too far east or west. Image "seasonal" objects.

If you have more LP in certain direction - East for your location as Birmingham is that way - avoid shooting in that part of sky - wait for target to clear the worst LP.

What sort of camera and lens are you using? Your image looks like it was processed in black and white, but I'm guessing you are using a sort of DSLR that should be color? Did you process your image properly? What stacking software did you use? Did you record raw files or have you used jpeg compressed ones?

Use raw / unaltered files for stacking. If using color camera make sure you get bayer matrix settings right - this will impact color if you get it wrong. Use fast lens and make it wide open (don't stop it down) if that lens gives you acceptable stars. If you have to stop it down to get sharp stars - do it minimally. Maybe consider getting better / faster lens if your lens is too slow.

What settings are you using for your DSLR? You need your ISO to be about 800, but that will depend on camera model.

Hi thanks for your reply, I took this picture from the dudley area using a canon 1200d with a 200mm lens at f4.5, I'm unsure why every picture I take is in black and white, I used deep sky stacker and then used a free software called fotor which I used to sharpen and mess around with it. ISO was at 1600 I took this image recently at around 9PM I always shoot in raw format with astrophotography.

[vlaiv]

20 minutes ago, Arran townsend said:

Hi thanks for your reply, I took this picture from the dudley area using a canon 1200d with a 200mm lens at f4.5, I'm unsure why every picture I take is in black and white, I used deep sky stacker and then used a free software called fotor which I used to sharpen and mess around with it. ISO was at 1600 I took this image recently at around 9PM I always shoot in raw format with astrophotography.

Use following default settings for DSS:

Use regular average stacking method (don't bother with sigma clip or more advanced stacking methods at the moment). Use per channel background calibration.

Maybe post your resulting tiff for inspection in this thread or start new topic?

[Arran townsend]

1 hour ago, vlaiv said:

Use following default settings for DSS:

Use regular average stacking method (don't bother with sigma clip or more advanced stacking methods at the moment). Use per channel background calibration.

Maybe post your resulting tiff for inspection in this thread or start new topic?

 

andromeda stacked 2.TIF

[vlaiv]

This was rather hard to process - SNR is poor, but color is there now. I had to crop it to be able to remove background with any success.

Also, don't apply stretching in DSS - do it after, when you are done with linear stage manipulation (like binning / denoising or whatever you need to do to image).

[Arran townsend]

17 hours ago, vlaiv said:

33 minutes ago, vlaiv said:

This was rather hard to process - SNR is poor, but color is there now. I had to crop it to be able to remove background with any success.

Also, don't apply stretching in DSS - do it after, when you are done with linear stage manipulation (like binning / denoising or whatever you need to do to image).

That looks pretty good with the colour, when you say it has a bad signal to noise ratio how do I improve that? And when you say don't apply stretching I'm not sure what you mean because that's how it came out DSS I didn't apply anything like that to it.  Also what software did you use for this? 

[vlaiv]

Sorry if I mistook this image for stretch, it's been a while since I last used DSS. I remember it having histogram manipulation and basic stretch and curve on the image below indicates non linear transform (that is what we call stretch).

Here is what I've done in terms of processing:

I imported your TIF into ImageJ - convert it to 32bit, binned 3x3 to recover some SNR, cropped it and applied basic background subtraction - it usually does not produce very good results, but in this case it worked well because there is a lot "flat" background around M31 (it causes trouble if target occupies much of the image).

After that I saved it as 32bit Tif and opened it in Gimp 2.10 for saturation adjustment and slight stretch to bring a bit of fainter galaxy parts into view. I also did selective hue adjustment, as is green was significant in the image so I shifted green hue to yellow.

Astrophotography is all about SNR, you want your SNR to be as high as possible. Here are basic guidelines on how it works. Signal to noise ratio (SNR) is ratio of Signal and Noise. You can increase it by either increasing signal or decreasing noise.

Signal is just light from target, in this case galaxy and stars. You can increase it in single exposure by extending exposure length.

Noise is more complicated as it comes from different sources. One is shot noise - fact that light comes in discrete packets / photons, and each pixel that registers some signal will on average have square root of signal value as associated shot noise. Besides this, there is thermal noise of camera that accumulates with exposure (thermal noise is another type of "shot" noise - as thermal signal builds and it consists out of electrons same applies - on average noise associated with thermal signal is equal to square root of this signal). In image this thermal signal also represents sort of noise. It's not random in nature but it is unwanted signal - this is what calibration is for or dark subtraction. There is also read noise - that one does not depend on exposure length and is related to way pixel values are read from sensor and converted to digital values. In the end there is LP noise. Since light pollution is also a sort of light signal (unwanted, but you can't calibrate it out - you need way of removing it, like background removal I applied above) it comes with its own shot noise.

In the end when you stack multiple exposures, SNR increases by factor of square root of number of exposures. So 16 exposures stacked will have 4 times higher SNR than any of single exposures.

To improve your SNR:

- expose longer (stronger signal)

- make sure target is high in the sky because atmosphere attenuates signal (stronger signal)

- stack more individual images (stacking improves SNR over individual subs)

- manage thermal noise / dark current (at this point there is nothing you can do about it since you don't have cooled camera) (lower noise)

- do proper calibration of frames (this is something you can do to an extent because you can't precisely control temperature of your sensor) (removes unwanted signal)

- minimize light pollution (like choosing targets further from strongest LP, and again ones that are high in the sky - LP is usually the least strong there) (lower noise)

- avoid nights when the Moon is out - it acts in the same way as LP (lower noise)

- make sure you image in astronomical night time (check this website https://www.timeanddate.com/sun/uk/dudley)

 

[Rush]

Eventhrough light polution star colours are still visible.

reselect the bad frames and stack again, dont touch the stauration sliders in DSS.

Well, apologise if i tuned it a bit too far ?.

Cheers

Rush

[Arran townsend]

3 hours ago, Rush said:

Eventhrough light polution star colours are still visible.

reselect the bad frames and stack again, dont touch the stauration sliders in DSS.

Well, apologise if i tuned it a bit too far ?.

Cheers

Rush

Ah ok, it does look a lot better than mine with the colour, I did do a picture of the Orion nebula but that was also in black and white, I'm not sure my editing is up to scratch. Did you have to pay for the software you used?

[Rush]

30 minutes ago, Arran townsend said:

Ah ok, it does look a lot better than mine with the colour, I did do a picture of the Orion nebula but that was also in black and white, I'm not sure my editing is up to scratch. Did you have to pay for the software you used? 

If u detail your workflow, will be easy 2 figure out. Yes i use a dedicated  astro Image processing soaftware called PIXINSIGHT.

 

[Arran townsend]

On 06/03/2019 at 16:33, Rush said:

If u detail your workflow, will be easy 2 figure out. Yes i use a dedicated  astro Image processing soaftware called PIXINSIGHT.

 

After I take the photos I go through then to see which ones are unusable then load them into deep sky stacker along with just dark frames, then once stacked I get the contrast right so I can see the target, increase the saturation then save as a 16bit tiff. Then I load the image in to a freeprogram called Fotor, to get my final image I usually increase contrast and definition. At this point I try to get some colour in the image but it never works. I'm guessing the program is not all that good. I do also have gimp 2.0 and star tools which I don't know how to use.

[vlaiv]

1 minute ago, Arran townsend said:

After I take the photos I go through them to see which ones are unusable them load them into deep sky stacker along with just dark frames, them once stacked I get the contrast right so I can see the target, increase the saturation then save as a 16bit tiff. Then I load the image in to a freeprogram called Fotor, to get my final image I usually increase contrast and definition. At this point I try to get some colour in the image but it never works. I'm guessing the program is not all that good. I do also have gimp 2.0 and star tools which I don't know how to use.

First thing: don't adjust contrast in DSS or do anything to color there. Save as 32bit Fits format (or tiff if you find fits hard to work with). Use gimp 2.10 - it works with 32 bit data (and is free) - do all your processing there.

[Arran townsend]

14 minutes ago, vlaiv said:

First thing: don't adjust contrast in DSS or do anything to color there. Save as 32bit Fits format (or tiff if you find fits hard to work with). Use gimp 2.10 - it works with 32 bit data (and is free) - do all your processing there.

So as soon as it's done processing in deep sky stacker I don't do anything to the image at all, I just save it as a 32-bit fits? 

[vlaiv]

17 minutes ago, Arran townsend said:

So as soon as it's done processing in deep sky stacker I don't do anything to the image at all, I just save it as a 32-bit fits? 

Yes, or better 32bit TIFF - I can't remember how DSS stores colors in fits files. Try fits - and when importing into Gimp tell it to use multiple images in fits as channels - I think that should work but I never tried it.

[Rush]

20 hours ago, Arran townsend said:

After I take the photos I go through then to see which ones are unusable then load them into deep sky stacker along with just dark frames, then once stacked I get the contrast right so I can see the target, increase the saturation then save as a 16bit tiff. Then I load the image in to a freeprogram called Fotor, to get my final image I usually increase contrast and definition. At this point I try to get some colour in the image but it never works. I'm guessing the program is not all that good. I do also have gimp 2.0 and star tools which I don't know how to use.

Yes, dont adjust anything in DSS. Grab the 16bit Tiff file and do the rest in Gimp. Split the RGB channels clean it up and recombine them using layers. Mask  stars  and ROI and increase saturation  softly.

Red channel you can strech and use as Lum , Dont remeber lighten or Screen mode in layers. Opacity sliders you can use to your personal likes.

Im not capable of using Gimp or PS effecient..  Long back tried PS and gave up. Might be someothers around  here who can help you .

Sorry.

Rush

[Arran townsend]

11 hours ago, Rush said:

Yes, dont adjust anything in DSS. Grab the 16bit Tiff file and do the rest in Gimp. Split the RGB channels clean it up and recombine them using layers. Mask  stars  and ROI and increase saturation  softly.

Red channel you can strech and use as Lum , Dont remeber lighten or Screen mode in layers. Opacity sliders you can use to your personal likes.

Im not capable of using Gimp or PS effecient..  Long back tried PS and gave up. Might be someothers around  here who can help you .

Sorry.

Rush

Thanks for your help here, I will take what you've said and research each step, thankyou

[Arran townsend]

On 06/03/2019 at 12:52, Rush said:

Eventhrough light polution star colours are still visible.

reselect the bad frames and stack again, dont touch the stauration sliders in DSS.

Well, apologise if i tuned it a bit too far ?.

Cheers

Rush

Hi, would you be able to send me the full image maybe in a zip file or somthing so I can add it to my archive?

[Rush]

Hey ,

Im so sorry. didnt check this post.

After a qucik fix i discarded the Image and  the stacked Image file.

If you send me a stacked IF , will redo and post it. is that OK ?

CS

Rush

[Arran townsend]

34 minutes ago, Rush said:

Hey ,

Im so sorry. didnt check this post.

After a qucik fix i discarded the Image and  the stacked Image file.

If you send me a stacked IF , will redo and post it. is that OK ?

CS

Rush

Ok, I mean if you are willing to do that, thanks, I have another image of the Orion nebula which you can mess around with, i have tried processing it but it didn't turn out all that brilliant

andromeda stacked 2.TIF Orion Nebula.TIF