Back To Search Results
Site Logo

 
 

Processing (Pictor 216)


I have owned the 216 for over a year now, and until recently I was taking good images without realising. By the time I had processed the image, I had lost so much detail either by having too much noise, or by lowering the black point so far that the background was jet black (you can see that on most of my earlier images).

Now I understand how to get a good image. Firstly focus is critical, secondly as the 216 is a noisy camera the calibration frames have to be averaged from at least 8 individual frames, and lastly the processing must be done correctly (it is no good having a perfectly long exposed image with hardly any noise if you cannot display it to its best.)

So, after reading a couple of books (including Ron Wodawski's The New CCD Astronomy) and a lot of experimenting, I am having some success.
D20NewM5-016


The image to the left is an average of 8 x 20 second dark exposures. Note the gradient from top to bottom.

The image on the right is 1 of 70 images exposed for 20 seconds.

Flat M5-016d


This image on the left is the flat field. It is an average of 40 x 70ms exposures.
I take the flats in daylight inside a white room with no windows. The flat has had a dark frame subtracted that was an average of 8 x 70ms dark images.

The image on the right is the image above right with the dark frame subtracted and the flat field divided. Note the grainy background noise.

M5-085M5-085d

These images are from a set of 27 images with the object of interest moved slightly - this way we will get a larger field of view. Again the image on the right has been, dark framed and flat fielded. Note the grainy background.

M570-unadM570lowproc

This image on the right is an average of 70 frames that have already been calibrated - note the field rotation. It appears all washed out - but there is detail in there.
With a mild amount of curves and levels applied in photoshop the core of the cluster reveals its detail.

M570M5fullypro

Now with extra curves and levels in photoshop the fainter stars are visible but the core is burned out.
The image on the right was the above right image and the left image layered in photoshop. The burned out core in the top layer was feathered and deleted to reveal the detail from the layer beneath.

The dark areas were gaussian blurred and the bright areas unsharp masked.

Now the images are 'stitched' together by carefully matching levels and curves and feathering the join.
A final set of curves is applied to increase contrast.

BM5M5

I then tried a new technique that I had read on one of the yahoo groups. I duplicated the layer and oversharpened until there were black halos round the stars, I then blended the layer using 'lighten'.

The final image is sharp, contrasty, and fairly noise free.