Quote:
Originally Posted by Rigel003
That last image is fantastic Trevor, a real improvement. I'm not sure about this resampling near the end thing though. Resampling enlarges your image and for the extra pixels that have been added it interpolates values from the surrounding pixels. It will make a bigger picture but it can't improve resolution or add extra detail. I would rather have a smaller but sharper image than a bigger blurry one.
|
Thanks Graeme,
I am very pleased with the gain in resolution from re sampling.
This technique is quite new to me and will require more experimentation to establish the most effective way to employ it.
The image scale from my 16" F4.5 Newt is very nice to start with and I have found that with good data, even the resampled image is very nice, definitely not blurry.
Graeme, I freely admit that I do not understand the nuts and bolts of how resampling and for that matter deconvolution or other sharpening filters actually do their stuff. But I can see the results of my experimentation and at the end of the day, that is far and away the most crucial thing in processing, at least in my humble opinion.
I am doing minimal processing in Registax. I think this is where I have picked up resolution with the image posted here. The majority of and probably most crucial part of my processing was done on the upsampled image. The upsampled image, after final processing would stand alone on its merits, however down sizing that image back to my original image scale really gains so much, I was quite surprised as my data was very good to start with.
Obviously there are limitations to what can be extracted and it would be a wast of time and effort to do this with poor data, crap in crap out is how I look at it. However, with good data to start with there are real gains to be made. The proof of the pudding really is, in the eating.