Regression on integration test 094 after ashot to later D65 about darktable HOT 5 CLOSED

@jenshannoschwalm : This is for you :)

from darktable.

A side question, is it expected that almost all tests are slightly changed in the regression testsuite? I mean the new mode is the default but not for old edits, right? For example for recent test Test 0158-coloreq-no-guided-filter we had yesterday:

Test 0158-coloreq-no-guided-filter
      Image mire1.cr2
      Expected CPU vs. current CPU report :
      ----------------------------------
      Max dE                   : 0.00000
      Avg dE                   : 0.00000
      Std dE                   : 0.00000
      ----------------------------------
      Pixels below avg + 0 std : 100.00 %
      Pixels below avg + 1 std : 100.00 %
      Pixels below avg + 3 std : 100.00 %
      Pixels below avg + 6 std : 100.00 %
      Pixels below avg + 9 std : 100.00 %
      ----------------------------------
      Pixels above tolerance   : 0.00 %
 
  OK

So fully clean, 100%... and now:

Test 0158-coloreq-no-guided-filter
      Image mire1.cr2
      Expected CPU vs. current CPU report :
      ----------------------------------
      Max dE                   : 0.96704
      Avg dE                   : 0.00004
      Std dE                   : 0.00474
      ----------------------------------
      Pixels below avg + 0 std : 99.99 %
      Pixels below avg + 1 std : 99.99 %
      Pixels below avg + 3 std : 99.99 %
      Pixels below avg + 6 std : 99.99 %
      Pixels below avg + 9 std : 99.99 %
      ----------------------------------
      Pixels above tolerance   : 0.00 %
 
  OK

from darktable.

The PR #15602 did introduce a regression on this test.

The diff:

The expected output:

The new output:

See the green part in the skip, less redish seems good but it doesn't cover all the area...

So maybe we need:

* to update the integration test

* fix segmentation based parameters to cope with PR

I will check if there is a missing part in correction in segmentation stuff ...

from darktable.

A side question, is it expected that almost all tests are slightly changed in the regression testsuite? I mean the new mode is the default but not for old edits, right? For example for recent test Test 0158-coloreq-no-guided-filter we had yesterday:

I think we both agree on this shouldn't be the case :-) (Old edits should have the same output...)

from darktable.

It seems that there is multiple issue.

For example reverting:

fba96f2 Use vectorized functions

Makes test 0158-coloreq-no-guided-filter to be 100% clean again.

But the test in the initial message is still failing.

from darktable.