|
1.
Load the photo
Luminance
noise in this excerpt, especially the shady parts, is quite
strong, but the “poppy seeds” ruining the photo will be
soon gone.
True, there's color noise as well to some extent, but let’s
neglect it now. Those who wish to immerse in color noise,
please click on one of the links at the bottom of the page.
We can only get in the near of luminance noise if we can
set it apart from color noise data.
|
|
2.
Head for the lab!
Nothing
can be easier. This is exactly what LAB color mode is for.
Click on Image/Mode/Lab Color menu, and
bang on you are in Lab mode. Colors have not changed and
seemingly nothing at all happened, but if you click on Channels
palette, you’ll see that we received a Lightness
and two color channels (a and b).
Select Lightness Channel as that will be
used later on. This contains all the details of lightness
in the picture, as well as luminance noise. If you succeed
in sorting out Lightness channel, luminance
noise will also be gone.
|
|
3.
Washday
To
achieve this, choose Filter/Blur/Surface Blur
effect.
We set Radius at 3 for
our fairly bad noise. In case of a not so conspicuous noise
a value of 4-5 may well be appropriate.
It’s not a good idea to exaggarate it as it ruins contrast
(it affects it anyways).
After
you are done with the settings above, increase the Threshold
value carefully. Preferably go step by step and check the
results in the Preview (the Preview option
should be selected). If the value is increased, the program
will blur the areas with less contrast. Be careful: if you
go up the slider too much, the photo may look like a stonewashed
watercolor painting. We should find a value where the details
of the noise are quite blurred but not yet obscured. We
stopped at the value 8.
|
|
4.
Noiseless
Finally
go back to RGB mode – Image/Mode/RGB Color.
The end-result is shown on the right side. The original
picture part is to the left. Though blurring removed the
fine details yet the result is still photo-like, and what
matters is that luminance noise has also been wiped out.
Thank you for
visit our blog.
Hope you can learn something new from this tutorial. You can share your thought
& suggestion with us though comments below.
Thank you…
Reference: digiretus
|
|
|
No comments:
Post a Comment