K3's Astronomy - Deep Sky
"When I consider your heavens, the work of your fingers, the moon and the stars which You have set in place, what is Man that You are mindful of him?" -- Psalm 8:3,4

M45 - Detailed study about postprocessing and analysis of captured image
M45 - The Pleiades 11.01.2003 (21:14 - 22:13 UT)
Resolution: 6.1 arcsec/pixel
Camera: Nikon Coolpix 995, f=22.4mm
Telescope: 80mm F5 Refractor, 25mm eyepiece afocal
Result focal length: 358.4mm (~44X Zoom)
Result focal ratio: F4.48
Exposure: 42x20s (ISO400, sunny white balance), Noise Reduction
Camera control: The Force software
Processed by K3CCDTools, Maxim DL Demo, then in Corel PhotoPaint.
Seeing: windy, soft fog, Moon in first quarter, temperature -8C

The above picture is a result of several processing steps. I would like to invite you to a little tour through this steps.

A, Frames capturing
B, Frames integration
C, Post-processing (Going deeper)
D, Integration again
E, Analysis of negative image

F, Reference M45 image taken by Johannes Schedler



A, Frames capturing

1, At firtst I captured frames by Nikon Coolpix 995 Digital camera by means of The Force software. The details about attaching digital camera to my scope are here: Example of using Nikon Coolpix 995 digital camera for astronomy purposes.

2, Then I moved all frames to computer's hard drive.


Example of raw frame resized to 25%.
Click the image to see 50% size.


B, Frames integration

1, I summed all 42 captured frames in K3CCDTools program. K3CCDTools allows summing virtually unlimited number of frames with almost unlimited (under Win2k or WinXP) frames size.
By summing the captured frames we will obtain one result frame with much higher dynamic range (32 bit depth for each color channel) and less noise. Here is the result image (without any further processing, only resized to reasonable size):


Result of integration of 42 frames resized to 25%.
Click the image to see 50% size.
The result is very similar to raw frame, but it is less noisy.
Note: The difference between raw and result frames is reduced by JPG compression used for Internet.

2, The result of summation has dynamic range higher than 8-bit per color. As normally used picture formats as BMP/JPG/PNG and similar store only 8-bit information per color, the result image should be stored in FIT format. I saved result picture in FIT16 format, which is suffcient for summation up to 256 frames. For higher number of frames I use FIT32 format.

3, K3CCDTools offer Histogram stretching tool, which enables fast post-processing of the result picture. I used it very often for basic operations. For advanced post-processing I use FIT export and then I process the image in specialised astronomy programs (e.g. Maxim DL Demo).
Here is an example of simple histogram stretching and applying gamma 2.00:


Result of integration of 27 frames resized to 25%.
Histogram stretching (separate for each color channel) + gamma 2.00
Click the image to see 50% size.


C, Post-processing

As I didn't capture Flat Field frames (I had not enough time for it), I was forced to use more advanced software for post-processing which enables background flattening. Furthermore, in order to get more from picture I decided to use gray scale picture, which is not so sensitive to strong histogram stretching.

1, I loaded FIT images (exported by K3CCDTools) into Maxim DL Demo (MDL) and I stacked them into one grayscale FIT image.

2, Flatten background command applied to suppress uneven illumination (caused by vignetting)

3, I applied gamma factor 2.00 (in MDL it is 0.5)

Going deeper

4, Then I started to play with histogram stretching. Less stretching gave me picture with less noise, but also less faint details. Here are examples of 5 various histogram settings (starting with less stretching):


42x20s, gamma 2.00, histogram 1
Comment: Decent picture

42x20s, gamma 2.00, histogram 2
Comment: Still decent picture, more stars visible

42x20s, gamma 2.00, histogram 3
Comment: Some brighter stars start to merge, increased background noise

42x20s, gamma 2.00, histogram 4
Comment: More stars merged, background noise easily visible

42x20s, gamma 2.00, histogram 5
Comment: Many stars merged, rather high background noise, BUT hint of nebulosity arround Merope star is visible .

It is very interesting to see, how much information is hidden in result summed picture (exported to FIT files).
The following picture shows the above 5 picture in one animation (resized to 2/3 of above size):


Click the picture to see full size animation.

Conclusion:
Higher histogram stretching reveals more details in the picture, but some details are also lost (merged stars) and image noise is increasing.
The most annoying thing (at least for me) is merging bright objects.
So there is question - Is there a possibility to show faint objects while bright objects are not merged?
If you look again at the top picture of M57, you will see, that answer is YES.


D, Integration again

1, I saved all pictures as result of histogram stretching from 1 to 5

2, I load the pictures into K3CCDTools and summed them. No further processing.
And here is the result:


Click the picture to see full size photo.


E, Analysis of negative image

At the end of our tour I would like to show you negative images, which are useful for analysis of captured details. Black stars on white sky are easily detectable.

Here is negative image of Pleiades with rather strong histogram stretching:


Click the picture to see full size photo.


After even stronger histogram stretching the hint of nebulosity around Merope star is visible:


Notice hint of nebulosity around Merope and around other stars.
The darker background parts in corners are residues after background flattening. This is normal for such strong histogram stretching.

Only for comparison here is the negative image of Pleiades without applying Flatten Background command:


The effect of vignetting is clearly visible. But very soft hint of nebulosity around Merope is still detectable.


F, Reference M45 image taken by Johannes Schedler

And here is beautiful picture of Johannes Schedler (Panther Observatory). I chose it as a reference image of M45 for looking for features in Pleiades:


Click the picture to see full size photo.
M45 - The Pleiades 04.10. 2002 (00:30 UT)
Taken by Johannes Schedler during ITT star party
Camera: Canon D60
Telescope: 105 mm TMB refractor F6.2
Exposure: 4x300s (ISO800, sunny white balance), autoguided, DF subtracted
Processing: darks subtracted, averaged, curves, unsharp mask, color balance, size reduced to 40%; cropped, all in Photoshop
Seeing: mag 6 sky, some haze present

Even hints of IC349 visible south of Merope (click the picture to enlarge):

For those, who are interesting, here is a raw frame (click the picture to see full size raw frame):

My thanks belong to Johannes Schedler for providing his beautiful photo and information.

 

Back to M45 page


Computer generated images, real images, drawings and texts, if not mentioned otherwise, are property of the author and may not be reproduced or used without permission of author.


Home

Last Update: 15.01.2003