If you figured out a way to "overwrite" noise, would the imagestill look "normal?"

On 1/29/2017 2:59 PM, RichA wrote:
I used to try all kinds of home-brew tricks to eliminate noise in the "bad old days" when 4/3rds sensors suffered a lot from it. I'm wondering, apart from colour blotches which don't show up now until very high ISO's, if you figured out a way to average out discretely coloured "noisy" pixels, would the image still look normal, or would it take on the appearance of a heavily noise-washed image? Seems to me that noise reduction is crude, with the effect being in severe cases whole swaths of images simply wiped, like smearing paint with a brush. If de-discolouration could be done at the pixel-level, NR might work better. I just don't know if it's possible.


Unless something has changed in the last three years, de noising
software is really image blurring.

--
PeterN

In article , PeterN
wrote:

I used to try all kinds of home-brew tricks to eliminate noise in the "bad
old days" when 4/3rds sensors suffered a lot from it. I'm wondering, apart
from colour blotches which don't show up now until very high ISO's, if you
figured out a way to average out discretely coloured "noisy" pixels, would
the image still look normal, or would it take on the appearance of a
heavily noise-washed image? Seems to me that noise reduction is crude,
with the effect being in severe cases whole swaths of images simply wiped,
like smearing paint with a brush. If de-discolouration could be done at the
pixel-level, NR might work better. I just don't know if it's possible.


Unless something has changed in the last three years, de noising
software is really image blurring.

no it isn't, nor has it been except in the ****tiest of implementations.

On 1/29/2017 9:09 PM, nospam wrote:
In article , PeterN
wrote:

I used to try all kinds of home-brew tricks to eliminate noise in the "bad
old days" when 4/3rds sensors suffered a lot from it. I'm wondering, apart
from colour blotches which don't show up now until very high ISO's, if you
figured out a way to average out discretely coloured "noisy" pixels, would
the image still look normal, or would it take on the appearance of a
heavily noise-washed image? Seems to me that noise reduction is crude,
with the effect being in severe cases whole swaths of images simply wiped,
like smearing paint with a brush. If de-discolouration could be done at the
pixel-level, NR might work better. I just don't know if it's possible.


Unless something has changed in the last three years, de noising
software is really image blurring.

no it isn't, nor has it been except in the ****tiest of implementations.


I forgot you know more than my mentors.

--
PeterN

PeterN:
Unless something has changed in the last three years, de noising
software is really image blurring.

Except in astro-image pre-processing. Aligning and stacking FITS images
causes signal (stars, galaxies, nebulas) to build while random noise in
the background sky cancels out. Noise is further reduce by the
subtraction of dark frames, images made under the same conditions as
the light images, but with the telescope covered. Standard noise
reduction is sometimes needed in post-processing, but it makes no
difference if the black background sky is blurred; it is seen as clean
and black. In this photo the black is not background sky, but
foreground dust clouds obscuring the background stars and ionized
hydrogen http://www.primordial-light.com/gammacyg-toa-larger.html but
the principle is the same. This is a stack of five 12-minute exposures.
Other examples at http://www.primordial-light.com/deepsky8.html.

--
I agree with almost everything that you have said and almost everything that
you will say in your entire life.

usenet *at* davidillig dawt cawm

In article , PeterN
wrote:

I used to try all kinds of home-brew tricks to eliminate noise in the "bad
old days" when 4/3rds sensors suffered a lot from it. I'm wondering,
apart
from colour blotches which don't show up now until very high ISO's, if you
figured out a way to average out discretely coloured "noisy" pixels, would
the image still look normal, or would it take on the appearance of a
heavily noise-washed image? Seems to me that noise reduction is crude,
with the effect being in severe cases whole swaths of images simply wiped,
like smearing paint with a brush. If de-discolouration could be done at
the
pixel-level, NR might work better. I just don't know if it's possible.


Unless something has changed in the last three years, de noising
software is really image blurring.

no it isn't, nor has it been except in the ****tiest of implementations.

I forgot you know more than my mentors.

apparently so.

go read a book or two on noise reduction algorithms before commenting
further.

On 1/29/2017 10:01 PM, Davoud wrote:
PeterN:
Unless something has changed in the last three years, de noising
software is really image blurring.

Except in astro-image pre-processing. Aligning and stacking FITS images
causes signal (stars, galaxies, nebulas) to build while random noise in
the background sky cancels out. Noise is further reduce by the
subtraction of dark frames, images made under the same conditions as
the light images, but with the telescope covered. Standard noise
reduction is sometimes needed in post-processing, but it makes no
difference if the black background sky is blurred; it is seen as clean
and black. In this photo the black is not background sky, but
foreground dust clouds obscuring the background stars and ionized
hydrogen http://www.primordial-light.com/gammacyg-toa-larger.html but
the principle is the same. This is a stack of five 12-minute exposures.
Other examples at http://www.primordial-light.com/deepsky8.html.


I suspect that technique may be somewhat similar to in camera long
exposure NR, which makes a second exposure.

--
PeterN