What am I doing wrong?

[Star Gazer]

I've been out with my unmodified 7D trying to get some nice frames of M31. I took exposures at 10, 20 and 60 sec @ ISO 800. Not sure if that was going to give me enough dynamic range

My First question is should I be able to get a decent first result keeping in mind I didn't take any darks or flats etc.

At the same time I am trying to get to grips with Astroart. I have a demo version of AA5.

If I try to open any of the exposures, or even stack them the results look awful. The noise is terrible. What am I doing wrong?

Here's a screen shot from AA. I just opened one of the raw files, so no processing

I know I have a small amount of LP in the village where I live,but not this much.

If I try to stack the frames I don't see much difference in the end result in AA.

I tried to stack the same frames in DSS. Here the result looks much different basically it looks like a black and white image after I stretch the histogram, I've no idea why this happens.

If any one is feeling like they would like ago with they favourite software here are three of the 25 subs I took

http://dl.dropbox.com/u/16060721/Astro%20Help/M31_LIGHT_10s_800iso_20111019-21h30m04s481.CR2

http://dl.dropbox.com/u/16060721/Astro%20Help/M31_LIGHT_20s_800iso_20111019-21h33m21s337.CR2

http://dl.dropbox.com/u/16060721/Astro%20Help/M31_LIGHT_60s_800iso_20111019-21h34m56s020.CR2

beware the files are 20MB each.

Any advice gratefully received.

[Stephen]

(snip) didn't see you had already tried DSS

[Ju_Cooper]

Hi, I suspect you will always do better by taking darks, flats and bias frames. The darks and bias will particularly help with the noise. There was remarkably little, if any, gradient in your images probably reflecting the low LP in your village (lucky you!).

Although it probably doesn't make much sense (with diff exposures), I stacked your three raw files in DSS and opened the 32 bit autosave file in PixInsight (I've got no experience of AstroArt). After an autostretch and no other processing it looked like this:

As I said, there's little gradient and noise looks par for the course especially with out calibration frames The horizontal banding is marked and may be Canon related—I often see it with my 450D, but usually not so obvious. It disappeared using PixInsight's CanonBandingReduction script, anyway.

After fairly quick processing in PixInsight and Photoshop I got the final result here:

I'm sure by stacking consistent length long exposures and taking calibration frames you can get excellent results. There's almost always more information you can tease out

[Star Gazer]

Ju_Cooper,

Many thanks for your efforts in helping me; much appreciated!

I know need some calibration frames, and I know these will improve things a lot.

One thing you say that is confusing me. You say that it probably isn't worth stacking frames with different exposures. As I understand it M31 is a big dynamic range target, it has very faint dust clouds (needs a very long exposure) and a very bright core (needs a short exposure). So are you saying you don't need different length exposures? Or are you saying if you do have them, stacking will just ignore them?

I should say that I now know that a 60s exposure of M31 isn't long enough for the faint stuff.

I just tried stacking the three frames in DSS just to check my results with yours. I' must say my result is quite different from yours. I opened the autosav.tif in windows picture viewer thingy and the result is black and white. I tried to open it AstroArt but AA didn't like the file format. Any ideas why it is blab and white?

The reason I am using AA is that I am trying to evaluate it; may be the files I am using aren't exactly the best test going. I see you're using PixInsight. I took a look a some of the tutorials on the web I was amazed at what it does. However I dismissed it because of the strange workflows and user interface that it has. My thinking was I would only every be able to do what I was shown in those tutorials.

Do you find PixInsight to be easy to use? How did you learn to use it?

Once again many thanks for your help

[Black Knight]

The mono output from DSS is probably down to saturation. Have a look at this thread for some suggested DSS settings from AndyUK. They may help you out.

http://stargazerslounge.com/imaging-image-processing-help-techniques/150121-dss-problem-ngc7000.html

[Ju_Cooper]

The DSS output looks black and white for me if I try to open it in Windows Photo Viewer—it is also squashed side to side and has a grey bar down the right side. I think this is a side effect of trying to open a 32 bit image in a program that only just copes with 16 bit images.

When I open the same file in PixInsight it still looks very dark with only the merest hint of colour—it is only when the histogram is stretched can the details start to be seen, and then there is a lot of claening up to do to get rid of noise, gradients etc.

I may be wrong, but as I understand it there is no advantage in combining different exposures (at the initial stacking stage anyway). Astronomical images are exceedingly faint and the best general strategy is to have long exposures to increase the chance of capturing photons then stack these to "average out" the noise. M31 is a difficult object to process because of the large dynamic range, but I don't think there is much danger of losing the highlights in a 32 bit image. The main advantage of having the image at 32 bit is that it is possible to recover detail in the bright areas with appropriate processing as here on a partially processed M42 (HDR Wavelets before and after):

It's clear form the AstroArt website that it will work with 32 bit colour images (although it calls them 96 bit!).

There are ways of combining astro images of different exposure to give an HDR type combination, but I've never tried these and I suspect you would stack the lights of each exposure together and combine them later. Your dark calibration frames should be of the same duration as your light frames so combining different durations in DSS would confuse the program I suspect.

PixInsight is difficult to get into, but the documentation is improving. I, like many users, started out with Harry's videos without which I would have got nowhere. They can be found here:

http://www.harrysastroshed.com/pixinsighthome.html

[LukeSkywatcher]

Hi, I suspect you will always do better by taking darks, flats and bias frames. The darks and bias will particularly help with the noise. There was remarkably little, if any, gradient in your images probably reflecting the low LP in your village (lucky you!).

Although it probably doesn't make much sense (with diff exposures), I stacked your three raw files in DSS and opened the 32 bit autosave file in PixInsight (I've got no experience of AstroArt). After an autostretch and no other processing it looked like this:

[ATTACH]73041[/ATTACH]

As I said, there's little gradient and noise looks par for the course especially with out calibration frames The horizontal banding is marked and may be Canon related—I often see it with my 450D, but usually not so obvious. It disappeared using PixInsight's CanonBandingReduction script, anyway.

After fairly quick processing in PixInsight and Photoshop I got the final result here:

[ATTACH]73042[/ATTACH]

I'm sure by stacking consistent length long exposures and taking calibration frames you can get excellent results. There's almost always more information you can tease out

I dont think anything else needs saying.

WOW. What a difference.

This has given me great hope. Usually i keep exposures down to 5s because then the orange glow kicks in. This has shown me that i can take longer exposures and pretty easily subtract it in post processing.

I already know that stacking greatly improves images.