trouble with noisy picture

[ClausDuerr]

Hello to all PI Fans,

I am using PI just since 2 weeks - it´s a great tool :-)
... which takes some time to really be able to use all it´s powerful functions correctly.

I have serios trouble to de-noise the following picture of M31, I really tried hard with little success:

https://dl.dropbox.com/u/83650637/m31%20step%201.fit

could please help me with de-noising the picture!?
Which functions with which parameters would be helpful...?

thanks alot and clear skies!
Claus
 

[gvanhau]

Hello Claus
I Just downloaded your image.


As far As I can see, there is a guiding problem with it, (Al bright star are doubled)
If this is a single image, there is not much to be done.
If it is the result of a stacking, the I suggest you first  to revise the images  and exclude the bad ones.

The image itself is not so noisy.
If you make a luminance mask (extract Lightness) and apply it reversed
and then use atruswavelettransform  enabling denoise of layers 1 and 2 with deffault parameters, almost all noise is gone.

Regards
Geert

[ClausDuerr]

Good Morning Geert,

thanks alot for your quick answer! :-)

regarding guiding: your´re right, there is something wrong ... have to check why this happenend during exposure.

I tried the way you described to denoise the picture, with limited success.
maybe my default settings are wrong or I did something wrong regarding the layers?

here are my screenshot (small size for easy download) including the parameters for AWT.

before: https://dl.dropbox.com/u/83650637/before%20m31.JPG
after: https://dl.dropbox.com/u/83650637/after%20m31%20variante%201.JPG
another try:  https://dl.dropbox.com/u/83650637/after%20m31%20variante%202.JPG

could you please be so kind a also do a screenshot with your settings an results?

Dank u wel!
Claus

[Alejandro Tombolini]

Hi Claus,

As Geert pointed i assume that this is a single image. If it is the result of an Integration, then you should check first the calibration results and also rejection parameters during integration to be sure all hot pixels are eliminated.
Note that hot pixels can not be eliminated with the noise reduction tools.
In the example, before ATWT for noise reduction I have applied:

DynamicCorp
BackgroundNeutralization
DinamicBackgroundExtraction twice
ColorCalibration
CanonBandingReduction
HistogramTransformation
The showed ATrousWaveletTransform for noise reduction
HistogramTransformation for clipping.

Saludos. Alejandro.

[gvanhau]

Hello

here is the result that I propose,  basically what you did wrong is that you only enabled noise reduction in the first layer.
You have also to select layer two and enable noise reduction.

What Alejandro said is also true, bad pixels are very difficult to remove by noise reduction algorithms.

On my picture: left is original at the right the processed image including the settings. Y also included a small image containing an example of the stars with guiding problem.

Regards
Geert

[ClausDuerr]

Dear Alejandro & Geert,

thank you very much for your help!

there workflow was very helpful and also the screenshots.
here I realized that I had a basic problem understanding the user-interface of PI!

I did not know (until you showed me) that I am able to specify different denoising parameters to the single layers by selecting them in the layer-list in the upper area of the dialog.
shame on me (maybe also RTFM) :-)

now the result is more ok to me, knowing that the raw-image was not of high quality (esp. guiding) :-)

so thanks again -you did help me a lot- and have a nice sunday!
Claus