• Welcome to Orpington Astronomical Society.
 

News:

New version SMF 2.1.4 installed. You may need to clear cookies and login again...

Main Menu

On adding "diffraction spikes" post-aquisition

Started by Rick, Nov 27, 2007, 08:47:04

Previous topic - Next topic

0 Members and 2 Guests are viewing this topic.

Rick

Quote from: Fay on Nov 30, 2007, 14:13:31Well, I think that when doing astrophotography as a hobby & you use any sort of enhancement, be it curves, levels, filters RGB Ha & the convoluting ones etc etc. it is false, but how else can us humans get an idea of the shapes & different textures of space?
All that processing has been worked out to enhance the information in your original images. There are good reasons for taking those flats and darks as well as the actual images, and doing all that tedious manipulation. If you do the work well then you'll produce an accurate image that's a true representation of your subject within the limits of your imaging equipment.

Quote from: Fay on Nov 30, 2007, 14:13:31A lot of things are a personal preference & some of the professionals go a bit too far, using so many tools & colours. Personally I think you have to pull the reins in a bit, not because I can't do it as well, but you have to be a bit realistic & allow objects to be seen as they most probably are & not defininitely as they are not, if you know what I mean.
Yep. Get as much out of the image as you can, while being sure that everything you've drawn out of the original is actually real. If you do the work well then the image will be both pretty and a good scientific record.

Quote from: Fay on Nov 30, 2007, 14:13:31If imaging for scientific reasons, they still enhance to bring out the features that they may be studying at the time so as to get a better knowledge of what they are looking at, but that is ok as it is done for a different reason.
You can most definitely enhance an image in different ways. Take a look at the work Mark did on his Comet Holmes images in order to pull detail out of the comet's coma. The surrounding background went to pieces, but that was a price worth paying to pull the coma detail up out of the gloom.

I would very much like to see what your image would look like without any added spikes.

I'm used to seeing diffration spikes on astronomical images, and I use their presence to work out what I'm looking at. Take a look at Friday's APOD:

http://antwrp.gsfc.nasa.gov/apod/ap071129.html

The image was taken using a Takahashi Epsilon 180ED. (See http://www.astromart.com/articles/article.asp?article_id=571 fo a review of the instrument.)
Even in the small-scale version on the main APOD page you can see that some of the stars have quite sharp diffraction spikes while others do not. Look a little more closely at the image, and you'll find that the reason is that some of the stars are being obscured slightly by dust. Notice how they're the ones with less well-defined diffraction spikes. So whether a star shows diffraction spikes or not can be used to work out a little about whether you're seeing it directly, or whether its image is being blurred a bit by dust. Now, if you have a look at the review page you'll see an image including a globular cluster and a bright star. Note the prominent spikes on the star. So genuine spikes can tell us whether an object is stellar (at least within the resolving power of the telescope, which is probably better than the resolution of the imaging chip).

If you try to add spikes during processing you've only got the resolution of the imaging system to work with, and you have to try to work out which objects are stellar (and so might have spikes) and which are diffuse (and wouldn't have them). With the best will in the world the processing software isn't going to get it right all the time, and it's got less to go on in the first place. It's also, inevitably, going to mess the image up a bit. When folk look at the image, and they don't know the spikes have been added during processing, they could be fooled into mis-interpreting it (as I was when I first looked at your image), and that's the main reason why I think that adding spikes during processing is a bad thing to do.

Rocket Pooch

Quote thingy "So whether a star shows diffraction spikes or not can be used to work out a little about whether you're seeing it directly"

Rick, nope, what is happening here is the stars with smaller diffraction spikes are dimmer, granted they could be obscured by dust, or the might be less bright, you only know this by checking the stars ADU's.

You could however measure the dark points of the sky and then compare it to the dust cloud and check the difference.  In the image you have used as an example the dust clouds are a lot brighter than the sky background.  And where the dust is there in abundance there are no stars.


Quote thingy "If you try to add spikes during processing you've only got the resolution of the imaging system to work with, and you have to try to work out which objects are stellar (and so might have spikes) and which are diffuse (and wouldn't have them). With the best will in the world the processing software isn't going to get it right all the time"

Wrong, this is actually piss easy, in optical system diffraction spikes are caused by the interference of the secondary supports or fluids in scope systems in relation to the brightness of the object and exposure duration.  If we assume there is an ADU level where spikes are caused and visible, using that ADU value we could very easy define which stars would cause diffraction spikes and highlight them.  If I took an image and said all stars over 40K ADU have spikes and 1/3rd of the brightest stars on my images have them this proves they have an ADU value over 40k.  I could even write a routine to put a sliding scale on the application of diffraction spikes based on the ADU level and the size of the spike and simulate a scope, as with Mr Carbone.

You have given us a brilliant example of someone spending £8k on his imaging kit to take good photo's and techno babble again, well done.

Ian

Guys,

I'm going to lock this now. Time to agree to differ.

Ian