color of spring

On Fri, 20 May 2016 09:48:26 -0400, nospam
wrote:

In article , Eric Stevens
wrote:

the point is that what dehaze does under the hood cannot be replicated
by anything you (or anyone else) can do in lab space (or any other
space) by tweaking with the sliders.

In many cases I can do better in Lab.

you might think you can, but you can't model haze by simply adjusting
sliders, lab space or otherwise.

Who says?

mathematics.

"The map is not the territory".

In any case I can adjust everything that's available for
adjustment.

sure, but you can't match what the dehaze filter can do.

that adjustment is not available to you any other way.

I'm not going to argue with you until we can compare photographs.

it' not a question of comparing photos.

the dehaze filter mathematically models haze, something that is
*impossible* to do by manually adjusting sliders.

What kind of haze? What is the source of the haze? Or are you going to
argue (of course you are - silly question) that a sulphate haze from
coal burning power stations is identical to the haze caused by forest
fires or aerosols naturally created by a forest or the haze of nitrous
oxides and aerosols associated with a large city? I could go one but
the point is that there is no such thing as an ISO standard
atmospheric haze.

it might look ok, but it won't ever be as good.

http://forum.luminous-landscape.com/index.php?topic=101244.40
"The Dehaze technology is based on a physical model of how light is
transmitted, and it tries to estimate light that is lost due to
absorption and scattering through the atmosphere"

Conceptually, Dehaze estimates two quantities from the original photo
1 a constant haze color (AKA the airlight color
2 a light transmission map which acts as a per-pixel blending
parameter between the true image (without haze, not observed) and the
estimated haze color that leads to the observed image (with haze).

So the software makes estimates and adjusts accordingly.
So why can't I? You could be right and my adjustment may not be as
good but in difficult cases I prefer my own estimates.

With these estimates, one can run the process of removing haze and
adding haze.

and for something more technical:
http://students.cec.wustl.edu/~jwaldron/559/project_final/assets/defog_fattal.pdf
--

Regards,

Eric Stevens

In article , Eric Stevens
wrote:


the dehaze filter mathematically models haze, something that is
*impossible* to do by manually adjusting sliders.

What kind of haze? What is the source of the haze? Or are you going to
argue (of course you are - silly question) that a sulphate haze from
coal burning power stations is identical to the haze caused by forest
fires or aerosols naturally created by a forest or the haze of nitrous
oxides and aerosols associated with a large city? I could go one but
the point is that there is no such thing as an ISO standard
atmospheric haze.

ask simon chen, who wrote it.

it might look ok, but it won't ever be as good.

http://forum.luminous-landscape.com/index.php?topic=101244.40
"The Dehaze technology is based on a physical model of how light is
transmitted, and it tries to estimate light that is lost due to
absorption and scattering through the atmosphere"

Conceptually, Dehaze estimates two quantities from the original photo
1 a constant haze color (AKA the airlight color
2 a light transmission map which acts as a per-pixel blending
parameter between the true image (without haze, not observed) and the
estimated haze color that leads to the observed image (with haze).

So the software makes estimates and adjusts accordingly.
So why can't I? You could be right and my adjustment may not be as
good but in difficult cases I prefer my own estimates.

you can, but you aren't going to model haze in your head.

you might like the results but that's a separate issue entirely.

With these estimates, one can run the process of removing haze and
adding haze.

and for something more technical:
http://students.cec.wustl.edu/~jwaldron/559/project_final/assets/defog_fattal.pdf

read that before commenting further.

On Fri, 20 May 2016 22:12:44 -0400, nospam
wrote:

In article , Eric Stevens
wrote:


the dehaze filter mathematically models haze, something that is
*impossible* to do by manually adjusting sliders.

What kind of haze? What is the source of the haze? Or are you going to
argue (of course you are - silly question) that a sulphate haze from
coal burning power stations is identical to the haze caused by forest
fires or aerosols naturally created by a forest or the haze of nitrous
oxides and aerosols associated with a large city? I could go one but
the point is that there is no such thing as an ISO standard
atmospheric haze.

ask simon chen, who wrote it.

it might look ok, but it won't ever be as good.

http://forum.luminous-landscape.com/index.php?topic=101244.40
"The Dehaze technology is based on a physical model of how light is
transmitted, and it tries to estimate light that is lost due to
absorption and scattering through the atmosphere"

Conceptually, Dehaze estimates two quantities from the original photo
1 a constant haze color (AKA the airlight color
2 a light transmission map which acts as a per-pixel blending
parameter between the true image (without haze, not observed) and the
estimated haze color that leads to the observed image (with haze).

So the software makes estimates and adjusts accordingly.
So why can't I? You could be right and my adjustment may not be as
good but in difficult cases I prefer my own estimates.

you can, but you aren't going to model haze in your head.

you might like the results but that's a separate issue entirely.

To me, that is *the_issue*.

With these estimates, one can run the process of removing haze and
adding haze.

and for something more technical:
http://students.cec.wustl.edu/~jwaldron/559/project_final/assets/defog_fattal.pdf

read that before commenting further.
--

Regards,

Eric Stevens