I've been using ATWT and MMT for a while for both noise reduction and sharpening. I don't think they replace deconvolution but they can certainly be used to enhance the image afterwards. Deconvolution attempts to reverse the blurring due to seeing and other effects using an estimated point spread function. This can actually reduce star sizes and recover details. ATWT and MMT just enhance what is already there.
I don't know if there's necessarily any right and wrong, but my feeble understanding is that as you never know exactly what the correct PSF is and no so much quality data for deconv to work, ATWT and other contrast enhancing algorithms in PI will often succeed with less ring artifacts etc....
Have you looked at DynamicPSF, Rob? It appears to provide a reasonable method for estimating the PSF by measuring a sample of suitable stars from a linear image. It also appears that best practice of using a deringing mask, global deringing and masking low SNR areas them can control artifacts pretty well.
I'm convinced that, in theory at least, deconvolution can provide benefits that the other tools can't. Unfortunately, it's a fairly daunting process in PI. I'm sure it gets easier with practice. And that's what I'll be doing again tonight
Have you looked at DynamicPSF, Rob? It appears to provide a reasonable method for estimating the PSF by measuring a sample of suitable stars from a linear image. It also appears that best practice of using a deringing mask, global deringing and masking low SNR areas them can control artifacts pretty well.
I'm convinced that, in theory at least, deconvolution can provide benefits that the other tools can't. Unfortunately, it's a fairly daunting process in PI. I'm sure it gets easier with practice. And that's what I'll be doing again tonight
Cheers,
Rick.
Only rarely Rick, but the tutorial using lunar images blew my mind away with what someone with decent knowledge of how to use it could improve an image must confess.
I'm definitely getting tighter stars and more detail in the galaxy itself from the deconv. The hard part is knowing how far to push it. I often look back at my efforts afterwards and think, "hmmm... I sharpened/denoised/saturated/whatevered that just a bit too much"
I'm definitely getting tighter stars and more detail in the galaxy itself from the deconv. The hard part is knowing how far to push it. I often look back at my efforts afterwards and think, "hmmm... I sharpened/denoised/saturated/whatevered that just a bit too much"
Yes that is the trick.
I usually use 40 iterations or less in CCDstack positive constraint.
Basically Ken's technique is to do several versions at higher and low levels and then combine them in Photoshop as you see fit.
I believe how much decon you can use also depends on the image. A dim image will not take much but one that is stong will take more.
I judge by the star shapes. But if you want detail in the galaxy you could just judge the galaxy only, not worry about deformed stars and then blend in the galaxy sharpening only in a masked layer in Photoshop.
I'm still working on this image trying lots of different techniques and options, but here's the current version. It's certainly a lot better than my original attempt, at least IMHO
The galaxy is too magenta. Try Photoshop selective colour and tone back the magentas and add some yellow to the blues and you should get that nice galaxy sutley bluish arms.
The galaxy is too magenta. Try Photoshop selective colour and tone back the magentas and add some yellow to the blues and you should get that nice galaxy sutley bluish arms.
Greg.
Ta, Greg. Had a go at that. It's definitely an improvement. Thanks for the lessons!
Quote:
Originally Posted by RobF
Starting to look very nice! Plenty of great data to get that far for sure.
Thanks, Rob. My data collecting ability has always exceeded my ability to process. Maybe it'll catch up one day! It takes a while to develop a good sense of aesthetics for these images, too. There are lots of subtleties that you don't notice at first.
