taking long exposures with heavy light polloution

[dph1nm]

Another thing to remember is that all these images people are showing of light polluted skies are jpegs, and most software scales these so that the background looks bright! Depending on you camera, you may find that the RAW image is nowhere near as saturated. My 1000D has an extra 2 stops in RAW (a factor of 4). So even if a single sub looks bright orange on the back of the camera, one can still exposure 4 times as long before actually saturating the image (or perhaps more to the point, there is still plenty of dynamic range available).

NigelM

[oldpink]

the red X is my location, as you can see i'm in a pretty bad area for LP

[Alien 13]

Another thing to remember is that all these images people are showing of light polluted skies are jpegs, and most software scales these so that the background looks bright! Depending on you camera, you may find that the RAW image is nowhere near as saturated. My 1000D has an extra 2 stops in RAW (a factor of 4). So even if a single sub looks bright orange on the back of the camera, one can still exposure 4 times as long before actually saturating the image (or perhaps more to the point, there is still plenty of dynamic range available).

NigelM

I totally agree and i have found it much easier getting data out of bright subs than trying to drag stuff from close to the noise floor in dark ones.

Alan

[Thalestris24]

the red X is my location, as you can see i'm in a pretty bad area for LP

I'm right in the middle of Glasgow !!

[Dubhar]

the red X is my location, as you can see i'm in a pretty bad area for LP

I'm in the opposite side of Glasgow to you, the light pollution is definitely annoying, though i think i may have less than you. I've found you can process some of it out to an extent. I found this interesting, using a custom white balance, not tried that yet (ive no idea what im doing i just got my camera ), may help. And filters as suggested, you might get good results with a narrowband filter if your looking at DSOs, if you've ever watched the virtual star party Gary Gonnella takes some amazing images in a heavy light polluted area.

[Thalestris24]

I totally agree and i have found it much easier getting data out of bright subs than trying to drag stuff from close to the noise floor in dark ones.

Alan

Hi Alan and Nigel

Yes, indeed. However, the sub I posted does look the same in RAW but obviously raw files have a much bigger dynamic range (14bit compared to 8bit jpg). I've got used to seeing my bright subs! After stacking many subs with flats and darks the bright blue (caused by the lp filter) goes away and I can get a fairly decent intermediate image - further improved by processing, of course.

Louise

[Ptarmigan]

Hi Oldpink, Alien, Thalestris et. al
I feel your pain !

only at about 5s / 800iso could I even see the stars

You have probably gone too short !
Has anyone said about not worrying about the glow, the stars are in there but you just cant see them YET ! see below**

Has anyone said about the histogram of a test exposure ? see below** !

Firstly, I am no expert astro photographer,
I hope I am not about to appear to teach granny about eggs !

I thought I would pop in and offer this site for your consideration (and interests on cloudy nights, some facinating reading) :-

"Minimal Exposures (Can we use really short sub-exposures for our stacking?)"
http://www.samirkharusi.net/sub-exposures.html
"Ever Shorter Sub-Exposures (What about even shorter than "minimal"?)"
http://www.pbase.com/samirkharusi/really_short_subs
He has lots of interesting other pages in those two domains.

**
Take a series of ever shorter duration exposures until you get to some in which the histogram of all the orangness is mostly contained in the lefthand half or one thirdish. (dont worry about looking for stars yet, but it does give you confidence if you can see one or two lol! )

Take a guess on one of those lefthandish exposures, set your camera to it and take several shots of your subject.
Not too many to begin with else DeepSkyStacker will drive you nuts !
Dont yet bother with bias and flats etc.either, that can come later.
but a dark or 3, as described in that vid. wont go amiss till you are familiar with DSS

With luck the orangness will become darker (random stuff/noise averaged out of the picture) and stars will magically start to appear out of the murk (repetative wanted signal) !

If you did insist on seeing lots of stars in your original test exposure then now lengthen the exposure a bit

and do it all again

Oh I nearly forgot,

as they have said, take your pics in RAW,

then process to TIFF (or other losless format, I cant remember which others DSS can handle) and then import to DSS.

Cos DSS isnt always up to date with cameras RAWs and sometimes produces strange results. Use your camera bundled sw to process the RAWs

( at least till you get familiar with DSS !)

phew!

hth!

good luck

[Ptarmigan]

Oh ooo hang on,

PS.

yes take a good number of darks even up to equal as many as the lights

I forgot that my camera does its own sort of dark on each frame

clang ! sorry, told you I was no expert !!

[oldpink]

out of curiosity I put 40 subs and 5 blacks from the batch I ran as a test, the picture above was from that set
DSS came back with an error saying it couldn't find any stars and only one frame would be stacked
any idea why this happened as there is clearly some data in there, is it possibly because I used the Raw files

and DSS had an issue with them ??

[frugal]

Make sure that you have the latest beta version of DSS, otherwise it will only use the left hand 1/3 of the image on the recent few versions of the Canon cameras. Quite why they never promote it out of beta I have no idea.

If you have washed out subs, then you might need to set the Star Detection threshold value lower than the default of 10 in order to pick out the stars.

[frugal]

As a quick comparison of how much light pollution you can remove from an image, here is a single 3 minute sub from last Friday, along with the processed version:

It is still a bit blotchy as I am still learning how to remove gradients, but there are a couple of orders of magnitude more stars in the final version than appear in the unprocessed version.

[Ptarmigan]

there is a slider to adjust the threshold of detection have you found/tried that, on one of the tabed puldowns,

Ive forgotten what camera you are using, my daughter had troubles like that with a 650d till we got the threshold set right and the RAWs processed with DCRAW

Having said that, we dont have your dreadful LP to try it out on !

Will look in here and DSS some more later but now my domestic staff have arrived looking hungry

ah I see Frugal is here with interesting suggestions, good,

[Thalestris24]

out of curiosity I put 40 subs and 5 blacks from the batch I ran as a test, the picture above was from that set

DSS came back with an error saying it couldn't find any stars and only one frame would be stacked

any idea why this happened as there is clearly some data in there, is it possibly because I used the Raw files

and DSS had an issue with them ??

Hi again

Used to have to use the beta version for CR2 files but the latest DSS 3.3.4 works fine. Yeah, as mentioned, use the threshold checker before registering and adjust down if necessary. I think you need something like at least 40 stars detected? Not sure of exact figure. Unfortunately there can be several reasons for DSS refusing to stack. I think it likes stars to be fairly sharp (in focus) and round. I don't think it cares about light pollution as long as it can detect a decent number of stars. I have been saving raw + jpeg via APT so that I could have a quick look at the jpeg versions of the subs and maybe a quick auto-level to see what's there. Obviously if I have obviously eggy stars there's no point in doing anything further with it anyway...

Louise

Louise

[oldpink]

Ok I converted them to TIFF and I'm getting somewhere
I managed to process them and got an image where I can actually see some more detail and take out a lot of the clutter

[toxic]

hi oldpink

wouldnt you be just aswell going to rowardennan on loch lomond i went there a couple of years ago and it was clear and what a site galaxy arm from horizon to horizon beautiful but didnt have any gear with me not even a camera   .

and i wish that was my light pollution i will take a pic tonight and you can see even with the lp i suffer you can still do long exposure of  upto 15 minits with a cheap lp filter much much longer with a real good filter.

[oldpink]

hi oldpink

wouldnt you be just aswell going to rowardennan on loch lomond i went there a couple of years ago and it was clear and what a site galaxy arm from horizon to horizon beautiful but didnt have any gear with me not even a camera   .

and i wish that was my light pollution i will take a pic tonight and you can see even with the lp i suffer you can still do long exposure of  upto 15 minits with a cheap lp filter much much longer with a real good filter.

my plan is to go to the Galloway Forest to the south of me (about 1 hrs drive)

its listed as a Dark Sky Site and has a couple of great locations for astronomy

[Ptarmigan]

here is a single 3 minute sub from last Friday, along with the processed version:

Niceone frugal, good demo of the power of stacking !

It'll be interesting to see how far down oldpink can get the sub exposure times

there is an eventual trade off with read noise, but that is the least of the probs at the mo !

and if whatsisname in the vid can do 1.6s without hitting it then ?? I didnt note what camera/sensor he was using.

I think, do I remember,  there may be a less of a gain with nebulea than stars ??? or am I getting confused with gradient removal ?

[Spikey]

Here's a 10 minute sub from my back garden without any LP filter - taken with an un-modded Canon 1000D DSLR

and this is the same target, same night, just popped the camera out to attach a £30 2" skywatcher LP filter.

Quite a difference !

[oldpink]

my poor initial efforts with eggy stars
managed to clean it up a bit playing in DSS with no idea what I was doing except aligning the RGB and moving them left & right to see what worked best
this was 12 subs 0.8 and 1.6s and 5 darks so a total exposure of about 8 seconds

[oldpink]

just checked the raws, DSS see's no stars when I check but finds them in the TIF files

[Thalestris24]

Hi again

I'm afraid your stars are well out of focus which won't help....

[Ptarmigan]

just checked the raws, DSS see's no stars when I check but finds them in the TIF files

Thanks for the confirmation, so both the 600 and 650 cause DSS problems,

daughter will be pleased to know that it isnt just her !

And your stacked pic : that's good news

In the style of the Beatles "It's getting better all the tiimeee, getting so much ,,,  "

[Thalestris24]

Thanks for the confirmation, so both the 600 and 650 cause DSS problems,

daughter will be pleased to know that it isnt just her !

And your stacked pic : that's good news

In the style of the Beatles "It's getting better all the tiimeee, getting so much ,,,  "

There shouldn't be any problems with using DSS on images from any of the recent or current canon camera models. Problems usually arise as a result of image quality issues.

Louise

[Thalestris24]

my poor initial efforts with eggy stars

managed to clean it up a bit playing in DSS with no idea what I was doing except aligning the RGB and moving them left & right to see what worked best

this was 12 subs 0.8 and 1.6s and 5 darks so a total exposure of about 8 seconds

ps don't align the colours in DSS - the image will come out in monochrome!

Oh, here's a webcam view of what my scope sees... Couldn't focus the webcam properly for some reason. It never gets dark here!

[oldpink]

Ah thats why they are coming out monochrome thanks for that