From: nicolas <nicolas@git.imagemagick.org>
Date: Tue, 11 Sep 2012 12:09:19 +0000 (+0000)
Subject: (no commit message)
X-Git-Tag: 7.0.1-0~4966
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=9192ad37fe6e02cc8eaf87663b61efec7f75fa2d;p=imagemagick

---

diff --git a/MagickCore/enhance.c b/MagickCore/enhance.c
index 853505162..67bb1ca47 100644
--- a/MagickCore/enhance.c
+++ b/MagickCore/enhance.c
@@ -3356,10 +3356,11 @@ MagickExport MagickBooleanType SigmoidalContrastImage(Image *image,
   (Sigmoidal((a),(b),(x))-Sigmoidal((a),(b),0.0)) / \
   (Sigmoidal((a),(b),1.0)-Sigmoidal((a),(b),0.0)) )
   /*
-    Inverse of ScaledSigmoidal, used for +sigmoidal-contrast. In
-    HDRI, the argument of the hyperbolic tangent (resp. logistic
-    sigmoidal) may be outside of the interval (-1,1) (resp.
-    (0,1)), hence the branching.
+    Inverse of ScaledSigmoidal, used for +sigmoidal-contrast.
+    Because b may be 0 or 1, the argument of the hyperbolic tangent
+    (resp. logistic sigmoidal) may be outside of the interval (-1,1)
+    (resp. (0,1)), even when creating a LUT, hence the branching.
+    In addition, HDRI may have out of gamut values.
     InverseScaledSigmoidal is not a two-side inverse of
     ScaledSigmoidal: It is only a right inverse. This is unavoidable.
   */