So, what am I doing wrong :-)

[szymon]

Last night I set up the 50D with MagicLantern as an intervalometer, pointed the scope at M31, and took some 70 lights, 15 darks and 15 flats.  DSS shows a total of 48m30s exposure in the stack.  My raw images look like this:

Doing some simple processing on this (just increasing the contrast and strengthening the colour levels) shows that there is something clearly there:

The light at the bottom is actually all over the image, but it's much worse at the bottom -- there is a clear delineation.  I am assuming that this is light pollution from the great smoke.

My issue is that, after stacking, I can't seem to pull out any more data.  Here's the best I've been able to come up with so far:

Can anyone explain what I'm doing wrong? :-)

-simon

[szymon]

If anyone wants to have a go at processing, here is the original tiff output from DSS (I have many different ones from trying different options in DSS).

06.TIF

-simon

[Daniel-K]

well it looks like something is blocking your light path and your focus is off. do you use a long dew shield?

[Leveye]

First crop off the edges. I see stacking remnants. That will really mess up the histogram when stretching later. Lots of LP there.Use a gradient mask or filter.  Your over tweaking the levels in that crazy edited  one. Big no no. Less contrast the better.

[Mukv]

Greetings

Sorry to say but this is light pollution at it's worst, I suffer a little worse than this but it seems slightly less when the air is dry in a frost it seems less.

Andy

[szymon]

well it looks like something is blocking your light path and your focus is off. do you use a long dew shield?

Interesting.  I do have a long-ish dew shield, it's a home-made one that I got with the scope (bought second hand).  That said, it hasn't seemed to cause any problems with other images.  Regarding the focus, yes I know it's a little off -- I stupidly forgot to refocus when I swapped the 6D out for the 50D.  But it's not *that* far off is it?

First crop off the edges. I see stacking remnants. That will really mess up the histogram when stretching later. Lots of LP there.Use a gradient mask or filter.  Your over tweaking the levels in that crazy edited  one. Big no no. Less contrast the better.

Oh, interesting -- didn't think about cropping off the edges.  Yes, the over-tweaking was just an example, I've been playing with lots of things.  I'm considering investing in a filter (deciding UHP, OIII, LP - which do folk think would be better for this particular problem?)  What is a gradient mask?

Greetings

Sorry to say but this is light pollution at it's worst, I suffer a little worse than this but it seems slightly less when the air is dry in a frost it seems less.

Andy

Yeah, I feared that was the case.  Ruddy London.  I think I'll try again in a few months when it's higher in the sky :-)

-simon

[sharkmelley]

Your image is certainly defocused but that is not the main issue.  The raw image you showed at the beginning shows obvious light pollution.  Almost certainly a light pollution filter would help here. 

But I think your main problem is to do with DSS settings and I don't have much experience of DSS I'm afraid.  I downloaded your TIFF file only to find that it has been saved as 8 bit i.e. all the fine granularity of your data has been thrown away making it impossible to draw out much detail after removing the gradients.  Also, any colour information in the file has been more or less wiped out - it is almost monochrome.  I strongly suspect some kind of log scaling has been applied - you really don't want that to be done.

Sorry I can't be of much more help but I hope that gives a few pointers.  Save your image as 16bit and don't let DSS wreck your data by applying scaling.

Mark

[szymon]

Ohh, sorry -- the TIFF file was processed using GIMP as an 8 bit tiff, not sure why it wasn't using it in 16 bit.  That was dumb of me, I was testing something.  I'll generate a fresh one from DSS.

The focussing I know (I used a Bhatinov mask, but focussed the 6D, then swapped cameras and then my daughter called me over to show her M13 in the Dobsonian and when I came back I saw the 50D on the newt and turned it on without refocussing!).

-simon

[szymon]

Ok, if anyone would like to try rescuing this, here is the first TIFF I got from DSS.

https://dl.dropboxusercontent.com/u/3614054/first_dss.tif

It's 16bit, not cropped, as is.  Would be interested if anyone can see anything at all! :-)

[sharkmelley]

It still looks like some severe dynamic range scaling (log scaling maybe) has been applied.  This scaling has not done you any favours and has washed out the colour.

The other problem is that distinct cut-off in the light pollution gradient in the top third of the image accompanied by posterisation.  It is possible this is also partly caused by the scaling. 

You need to check your DSS settings I think.  I would expect a stacked image to look very similar to your original raw with the same colouring etc.  What is actually coming out of DSS is currently looking very different.

Mark

[szymon]

And here is the last tiff I got DSS to produce -- this one you can actually start to see the swirling of the galaxy merged in amongst the light pollution and hidden behind the out of focus stars ;-)

https://dl.dropboxusercontent.com/u/3614054/first_dss.tif

-simon

[szymon]

This is a very interesting point.  Perhaps I've done something very wrong somewhere.  Will look.  Thanks.

You need to check your DSS settings I think.  I would expect a stacked image to look very similar to your original raw with the same colouring etc.  What is actually coming out of DSS is currently looking very different.

[szymon]

Ok, so using a newer version of DSS has given dramatically different results from stacking:

https://dl.dropboxusercontent.com/u/3614054/FINALLY.tif

Perhaps now I can actually try to pull something out of it...

-simon

[Atreta]

Had a go with your image, not the last one.

if you want i can upload the tiff file later.

[szymon]

Thanks!

Yeah that seems to be the option -- getting rid of the LP makes the nebula look like a bright fuzzy similar to what you can see through the scope. Would be interested if you can get any more from the last tif, I think the new DSS did a better job (same settings).

Sent from my iPhone using Tapatalk

[sharkmelley]

The last one FINALLY.tif has not worked so well.  The stars in the bottom left have not stacked properly.  I'm guessing DSS didn't recognise them as stars.

I'm still not sure of the reason for the posterisation though I have seen it from time to time when using sigma stacking on a load of frames where the light pollution gradient is not the same from frame to frame.

Mark

[hazegood]

am totally new to imaging and oily doing widefields at the moment,  at the moment,  but one thing I have worked out is iwhen ligh polloution is giving the orange cast, change the cameras white balance setting to tungsten, seems to remove the colouring, doesn't remove the light though

Leigh

[szymon]

Indeed, the WB can be adjusted in a number of ways (in PP too) but it affects all colours not just the LP.

I've ordered a Skywatcher LP filter let's see if that can help as an inexpensive starter fix.

Sent from my iPhone using Tapatalk

[szymon]

The last one FINALLY.tif has not worked so well. The stars in the bottom left have not stacked properly. I'm guessing DSS didn't recognise them as stars.

I'm still not sure of the reason for the posterisation though I have seen it from time to time when using sigma stacking on a load of frames where the light pollution gradient is not the same from frame to frame.

Mark

I think the bottom left is the fact that I was OOF. The LP is still the major issue though. We'll see if the LP filter helps.

Sent from my iPhone using Tapatalk

[Astrosurf]

What length were your subs? The arms of the galaxy are faint so need quite long subs. The core is bright so some take shorter subs for that and then combine the two sub lengths.

Alexxx

[szymon]

Indeed, I've received the same advice from a couple of people in messages. Subs were 30 seconds, more than that and my mount starts trailing because I'm very new to polar alignment and balancing. I'm working on a finder-guider project which should help, I'll then take some longer subs and combine :-)

Sent from my iPhone using Tapatalk

[Astrosurf]

That's the way to go! That galaxy is a very difficult subject. Why not have a go at brighter, easier ones such as M27? You'll get more satisfaction that way.

[carastro]

I managed to get a little bit of the dust lanes showing by doing a gradient exterminator, but there is something seriously wrong with the file, it has long streaks from left to right.

Obviously you know about the trailing (as you're not guiding), and it's also definitely out of focus.

Rome wasn't built in a day.  Unfortunately this target needs longer subs which you can't do at the moment.

Carole 

[szymon]

Carole, that's better than anything I managed to achieve with this data -- very well done! I'd love to know what your processing steps were.

Yes, I know this image is a write-off, my purpose here is to learn. I've had some success with M57 and plan to image M13 tonight...

Sent from my iPhone using Tapatalk

[sharkmelley]

What length were your subs? The arms of the galaxy are faint so need quite long subs. The core is bright so some take shorter subs for that and then combine the two sub lengths.

Alexxx

Yes, taking some shorter subs to ensure the core is not burnt out is a good idea.  Take long subs for the faint arms but only if you have a perfect sky.  Once your subs are long enough for the light pollution to be obvious (as in this case) then subs of longer length will be of no help at all in revealing the faint parts.

Mark