First processing M31 - posterizing?

[Starface]

Hi everyone, data collected in February. Finally some time for processing. I used tutorial http://www.budgetastro.net/photoshop-basics.html

Lightpolluted skies
HEQ5 Pro Synscan, evostar ed80 pro, canon EOS 600D, no autoguider
56 lights x 50s, ISO 800, 15 darks, 20 bias
stacking DDS

When I started using levels, I saw a huge increase of black grainy particles/layers in the image (I think posterizing?)

Photoshop 2:
http://www.astrobin.com/242553/

I tried to  restack the data and use a more recently version of photoshop (PS 5). But I have the same problem when using levels (see attached picture).

Is this because of my computer? Not good enough for processing? Requirements: AMD Athlon II Dual Core M300 - 64 bit; 4,0 GB Ram, ATI Mobility Radeon HD4530 series
Way of processing?
Insufficient data?

I'm very curious. Because I want to know what to improve for my next images.
Thank for reading!

If possible I could always use a good stack (DSS) form somenone, so I can process this in Photoshop and see where I get?
 

[Oddsocks]

Hard to tell from your posted picture where the problem lies, it could be many things...

When you acquired the data were you saving Raw files or compressed jpg? For best quality you should be acquiring the data as Raw in the camera and import these directly into DSS.

The output from DSS should be saved as 16bit tif with no compression if you are post processing in Photoshop and either 32bit fit or tif with no compression for post processing in dedicated astro processing software.

In Photoshop after you import the stacked image open the "Mode" menu on the top menu bar and make sure you are working in 16bit mode, 32bit mode doesn't give you access to all the available Photoshop tools and 8 bit mode leads to the kind of artefacts your image shows.

Your computer specs are ok, it might be a little slow and it shouldn't cause poor quality images but do check the graphics card quality/resolution settings are set to best/highest when carrying out the post processing in Photoshop so you can see what is going on.

[Marky1973]

Hi Starface - did you increase the levels a little at a time or try all at once? If you push it too much, too quickly you can get a lot more noise creeping in a lot faster.

The other issue with Andromeda is needing longer subs to get the detail. There is data in your shot though, so keep at it, they will only get better as you get more experience (which is funny coming from someone who doesn't really know what they are doing!)

You can try a quick stretch in Deep Sky Stacker as well to give you an idea of what might be in your image. I wrote a very basic tutorial for what I do HERE.

The main thing is to just keep playing around and see how you get on and keep checking out tutorials - everyone has a different way of doing things and each target seems to have a different method too!

[wornish]

I have found that stacking in PS and stacking in DSS does produce different results.  Not sure one is that much better than the other it  seems to depend on the data.

If you post your images where they can be downloaded I am happy to stack them in DSS for you.

 

 

[Starface]

Hi everyone. Thanks for the advice! I stacked my lights, darks, bias in DSS.

I use RAW files. 32 bits converted into 16 bits in photoshop. I've increased the levels carefully.

Could I get a test autosave.tif from anyone ? So I now for sure if my processing is ok.

Kind regards

[wxsatuser]

13 hours ago, Starface said:

Hi everyone. Thanks for the advice! I stacked my lights, darks, bias in DSS.

I use RAW files. 32 bits converted into 16 bits in photoshop. I've increased the levels carefully.

Could I get a test autosave.tif from anyone ? So I now for sure if my processing is ok.

Kind regards

Here is an Autosave tif of the Californian Nebula.
It's 168Mb 32bit and you can download direct off my site.
This is 25x300sec Canon 60Da, Canon EF 70-200mm L IS 11 at f/2.8 105mm ISO 800.
The 25subs were dithered with a 12pixel offset in a spiral pattern.
No flats, darks or bias.
My processing is not that great and no doubt others could do a lot better than me.

http://www.weathersatellite.info/astro/Autosave.tif

Here is the 16bit tif result out of Pixinsight, it's 84Mb, this was hit with ABE three times to kill horrendous gradients.
The PI processing was pretty simple, crop, LRGBcombination, stretch and a little noise reduction.
You may find using Photoshop you will struggle with the gradients.

http://www.weathersatellite.info/astro/Autosave_ABE_ABE_ABE.tif

After a bit of tweaking in Photoshop, Deep space and Space noise reduction actions, tweak curves and colour.
This is a reduced jpeg.

 

[ollypenrice]

You could try this method in Ps to stretch your stack. Do a basic gentle Levels stretch till your background sky measures 23/23/23 (ideally) in each colour channel. This should be rechecked in Levels to be sure that the histogram dataline goes right back to the black point and right over to the white point. In the image below that's what I started with before opening Curves. Note that there are four points on the curve you see which I'll call 1, 2, 3 and 4 with 1 at the bottom.

- Put the cursor on the background sky and Ctrl click. This will put point 2 onto the curve which is at this stage a straight line.

- Put in point 1 by hand to stop the curve bending later.

- Lift the 3rd point by hand.

- Use the 4th point to make the curve into a straight line at the top. Apply.

This prevents you stretching your faint background sky above the noise floor but picks up and stretches the faint galaxy signal above it. Flattening the line at the top reduces the stretch given to the stars.

Olly