Bug report #4917: when setting contrast enanchement to someting else than "no stretch", standard deviation is automatically set to "2" - QGIS Application - QGIS Issue Tracking

Bug report #4917

when setting contrast enanchement to someting else than "no stretch", standard deviation is automatically set to "2"

Added by Giovanni Manghi about 13 years ago. Updated almost 13 years ago.

Status:	Closed
Priority:	Normal
Assignee:	Alexander Bruy
Category:	Rasters
Affected QGIS version:	master	Regression?:	No
Operating System:		Easy fix?:	No
Pull Request or Patch supplied:	No	Resolution:	fixed
Crashes QGIS or corrupts data:	No	Copied to github as #:	14723

Description

Subject says almost all.

When setting the contrast enhancement to something else then "no stretch", then QGIS also sets the standard deviation to "2", automatically.

Use of SD should be just a user choice.

This makes certain rasters to load with "strange" colormaps when stretching is used.

Tested on qgis-master.

Associated revisions

Revision 67b698e2
Added by Alexander Bruy almost 13 years ago

use standard deviations only when requested (fix #4917)

History

#1 Updated by Maxim Dubinin about 13 years ago

Giovanni,

I believe, change back to 0 is not warranted. STD stretches are primarily used for remote sensing data where initial automatic 2 std stretch is common throughout different software packages. Automatic cutting off of the 5% of tails of the histogram is a good and rather conservative choice.

Can you show which example shows strange results?

Maxim

#2 Updated by Giovanni Manghi about 13 years ago

Hi Maxim!

I believe, change back to 0 is not warranted. STD stretches are primarily used for remote sensing data where initial automatic 2 std stretch is common throughout different software packages. Automatic cutting off of the 5% of tails of the histogram is a good and rather conservative choice.

good to know :)

Can you show which example shows strange results?

truemarble seems to me to load with a "strange" colormap when using stretching (and so the STD to 2)

http://download.gfoss.it/TrueMarble/TrueMarble-2km.sqlite

I have also a bunch of coloured military maps that looks strange with STD to 2.

Nevertheless with a STD 1.7 they both looks ok.

In any case there is something weird/wrong with STD as I can see clearly that the STD configuration "comes and go" when opening the raster properties.

I also notice that if I choose another STD value in the raster properties, then I click "apply", then nothing happens. If I save the new value and then reload the raster then I see the expected colormap changes.