Quote:
Originally Posted by jase
In my opinion, I like your previous versions Bert. The background appears mottled in these versions. Perhaps you've pushed noise reduction too hard or used a similar routine. Its great to see this data working over and over with the various tweaks made. Its a testament to your commitment. Keep them coming.
|
The mottling is caused by the RGB data having lack of signal that is far less than the NB data. It then tends to modulate the floor of the NB data in dim areas. I am just glad I can see down to the noise so then I can do something about it.
The noise reduction in Startools is very good. I fact it is very difficult to see any difference before and after. Startools keeps track of the stretching and thus the S/N of each pixel.
I am sure if I only took a small part of the image and optimally blended the RGB with NB the result would be far better. With a huge dynamic range in both RGB and NB over 12+ square degrees it is difficult to satisfy the bright and dim areas at the same time.
I do not want to start messing around with selective area manipulation as it would lead to matching panels accurately for mosaics impossible.
My only motive for putting up all these trials and errors is so it may inspire others to push the limits with their own data.
I know that there is no final solution to any complex problem. There is only compromise and you have to be satisfied with an outcome that is far from perfect.
Thanks for your input Jase, it stops me from being caught in logic loop that I then cannot see my way out of!
Bert
