GNOME.org
 

[gimp-help-2] Add new terms to glossary

commit 33968f105974f9fb23fc502e882adb314a8ef5f4
Author: Elle Stone <ellestone ninedegreesbelow com>
Date:   Sun Apr 2 09:00:52 2017 -0400

    Add new terms to glossary
    
    Add the following terms to the glossary:
    Channel encoding, including
        linear gamma/linear light/linear/gamma=1.0
        sRGB TRC/LAB TRC as almost/exactly perceptually uniform
        Linear vs Perceptually uniform
    Color
    Display-referred
    Display-referred white
    Display-referred black
    High Dynamic Range
    Scene-referred

 images/C/glossary/companding-curves-compared.png |  Bin 0 -> 221691 bytes
 src/glossary/glossary.xml                        |  293 ++++++++++++++++++++++
 2 files changed, 293 insertions(+), 0 deletions(-)
---
diff --git a/images/C/glossary/companding-curves-compared.png 
b/images/C/glossary/companding-curves-compared.png
new file mode 100644
index 0000000..580c157
Binary files /dev/null and b/images/C/glossary/companding-curves-compared.png differ
diff --git a/src/glossary/glossary.xml b/src/glossary/glossary.xml
index 6a3c9cf..9bbf7fb 100644
--- a/src/glossary/glossary.xml
+++ b/src/glossary/glossary.xml
@@ -231,6 +231,77 @@
     </glossdef>
   </glossentry>
 
+  <glossentry id="glossary-channel-encoding">
+    <glossterm>
+      <phrase>Channel encoding</phrase>
+    </glossterm>
+    <indexterm>
+      <primary>Channel encoding</primary>
+    </indexterm>
+    <glossdef>
+      <para>
+      Channel encoding refers to how fast the intensity (more technically
+      correct for grayscale and RGB images, the <ulink 
url="https://en.wikipedia.org/wiki/Relative_luminance";>relative
+      Luminance</ulink>) of a channel in a digital image progresses from dark
+      to light as the channel values progress from 0.0 to 1.0 floating point
+      (0 to 255 for 8-bit integer, 0 to 65535 for 16-bit integer).</para>
+      <para>Other ways of referring to "channel encoding" include
+      "companding curve", "gamma" (which is technically not correct unless the
+      channel encoding is an actual gamma curve), "tone reproduction curve"
+      ("TRC" for short), and "tone response curve" (also "TRC" for short).
+      </para>
+      <para>The linear light channel encoding reflects the way lightwaves
+      combine there in the real world. The linear light channel encoding is
+      also referred to as "gamma=1.0", "linear gamma" or simply "linear".</para>
+      <para>Perceptually uniform channel encodings reflects the way our eyes
+      respond to changes in luminance.</para>
+      <para>In ICC profile color managed workflows, the following channel
+      encodings are commonly used:</para>
+      <orderedlist>
+        <listitem><para>The LAB companding curve, which is exactly
+        perceptually uniform.</para></listitem>
+        <listitem><para>The linear light channel encoding, which of course is
+        exactly linear.</para></listitem>
+        <listitem><para>The sRGB channel encoding and the "gamma=2.2"
+        channel encoding, which are both approximately perceptually uniform
+        and approximately equal to each other.</para></listitem>
+        <listitem><para>The "gamma=1.8" channel encoding, which is neither
+        linear nor approximately perceptually uniform, though it's closer to
+        being perceptually uniform than it is to being linear.</para></listitem>
+      </orderedlist>
+
+      <mediaobject>
+        <imageobject>
+          <imagedata fileref="images/glossary/companding-curves-compared.png" format="PNG"/>
+          </imageobject>
+        <caption>
+          <para>The Linear light, sRGB, and LAB channel encodings compared.</para>
+        </caption>
+      </mediaobject>
+
+      <para>Looking at the above image:</para>
+      <orderedlist>
+      <listitem><para>The Linear light channel encoding (top row) represents
+      how lightwaves combine out there in the real world.</para></listitem>
+        <listitem><para>The sRGB channel encoding (middle row) is almost
+        perceptually uniform.</para></listitem>
+        <listitem><para>The LAB channel encoding (bottom row) is exactly
+        perceptually uniform, which means it represents how our eyes respond to
+        changes in luminance.</para></listitem>
+      </orderedlist>
+      <para>In GIMP 2.10 two different channel encodings are used internally
+      for various editing operations, these being "Linear light" and
+      "Perceptually uniform (sRGB)".</para>
+      <para>The companding-curves-compared.png shown above is a slightly
+      modified version of an image from <ulink 
url="http://ninedegreesbelow.com/photography/xyz-rgb.html#Color";>
+      Completely Painless Programmer's Guide to XYZ, RGB, ICC, xyY, and TRCs
+      </ulink>, which is licensed as
+      <ulink url="http://creativecommons.org/licenses/by-sa/3.0/deed.en_US";>
+      Creative Commons Attribution-ShareAlike 3.0 Unported License</ulink>.
+      </para>
+    </glossdef>
+  </glossentry>
+
   <glossentry id="glossary-channels">
     <glossterm>
       <phrase>Channel</phrase>
@@ -297,6 +368,57 @@
     </glossdef>
   </glossentry>
 
+  <glossentry id="glossary-color">
+    <glossterm>
+      <phrase>Color</phrase>
+    </glossterm>
+    <indexterm>
+      <primary>Color</primary>
+    </indexterm>
+    <glossdef>
+      <para>On the one hand, <ulink url="http://en.wikipedia.org/wiki/Light";>
+      light</ulink> comes from the sun or other radiant sources, and is
+      <ulink url="https://en.wikipedia.org/wiki/Atmospheric_refraction";>
+      refracted</ulink>by mediums (water, the atmosphere, glass) and
+      <ulink url="https://en.wikipedia.org/wiki/Diffuse_reflection";>diffusely
+      </ulink> or
+      <ulink url="https://en.wikipedia.org/wiki/Specular_reflection";>
+      specularly</ulink> reflected by surfaces.</para>
+
+      <para>On the other hand,
+      <ulink url="http://en.wikipedia.org/wiki/Color";>color</ulink> isn't out
+      there in the world in the same tangible way that light is. Rather color
+      is part of how we sense the world around us. Light enters the eyes, is
+      processed by light receptors
+      (<ulink url="http://en.wikipedia.org/wiki/Cone_cell";>cones</ulink> and
+      <ulink url="http://en.wikipedia.org/wiki/Rod_cell";>rods</ulink>), and
+      sent via the optic nerves to the brain for further processing and
+      interpretation.</para>
+
+      <para>Light varies in
+      <ulink url="http://en.wikipedia.org/wiki/Wavelength";>wavelengths</ulink>,
+      which our eyes and brain interpret as varying hues (reds, blues, greens,
+      and so on), and also in <ulink url="http://en.wikipedia.org/wiki/Luminance";>intensity (aka 
"luminance")</ulink>. So our
+      <ulink url="http://en.wikipedia.org/wiki/Color_vision";>perception of
+      color</ulink> is composed of both intensity ("luminance") information and
+      chromaticity information.</para>
+
+      <para>The <ulink url="http://www1.icsi.berkeley.edu/wcs/";>naming of colors
+      </ulink> carries one out of the narrow realm of color perception, and
+      into the larger realm of cultural and linguistic interpretation and
+      classification of color, and thence into even larger philosophical,
+      aesthetic, theological, and metaphysical considerations.</para>
+
+      <para>The above explanation of Color is a slightly modified excerpt from
+      the <ulink url="http://ninedegreesbelow.com/photography/xyz-rgb.html#Color";>
+      Completely Painless Programmer's Guide to XYZ, RGB, ICC, xyY, and TRCs
+      </ulink>, which is licensed as
+      <ulink url="http://creativecommons.org/licenses/by-sa/3.0/deed.en_US";>
+      Creative Commons Attribution-ShareAlike 3.0 Unported License</ulink>.
+      </para>
+    </glossdef>
+  </glossentry>
+
   <glossentry id="glossary-cmyk">
     <glossterm>
       <phrase>CMY, CMYK</phrase>
@@ -447,6 +569,94 @@
     </glossdef>
   </glossentry>
 
+  <glossentry id="glossary-display-referred">
+    <glossterm>
+      <phrase>Display-referred</phrase>
+    </glossterm>
+    <indexterm significance="normal">
+      <primary>Display-referred</primary>
+    </indexterm>
+    <glossdef>
+      <para>The phrase "display-referred" refers to images that can be
+      displayed (either directly or by means of ICC profile color management)
+      on devices. The displaying device might be a monitor, or an image printed
+      on paper, or some other display technology.</para>
+
+      <para>Regardless of the technology, when you display an image on a
+      device, that device has a maximum and minimum brightness. The maximum
+      and minimum brightnesses are referred to as
+      <link linkend="glossary-display-referred-white">display-referred white</link>
+      and
+      <link linkend="glossary-display-referred-black">display-referred black</link>.
+      </para>
+
+      <para>The above explanation is a slightly modified excerpt from
+      <ulink url="http://ninedegreesbelow.com/photography/display-referred-scene-referred.html";>
+      Models for image editing: Display-referred and scene-referred</ulink>.
+      The modified excerpt was written and quoted by permission of the
+      author, who has licensed the modified excerpt under the
+      <ulink url="http://creativecommons.org/licenses/by-sa/3.0/deed.en_US";>
+      Creative Commons Attribution-ShareAlike 3.0 Unported License</ulink>.
+      </para>
+    </glossdef>
+  </glossentry>
+
+  <glossentry id="glossary-display-referred-white">
+    <glossterm>
+      <phrase>Display-referred white</phrase>
+    </glossterm>
+    <indexterm significance="normal">
+      <primary>Display-referred white</primary>
+    </indexterm>
+    <glossdef>
+      <para>"Display-referred white" (or for simplicity, "white")
+      means the floating point RGB color (1.0, 1.0, 1.0) and the integer
+      equivalents (255,255,255),(65535,65535,65535), etc, for 8-bit integer,
+      16-bit integer, etc.</para>
+
+      <para>"Display-referred white" has the very special significance that
+      in display-referred editing there's no such thing as
+      "brighter than white". So in display-referred image editing, all RGB
+      channel values are less than or equal to 1.0 and no color is brighter
+      than "white", (1.0, 1.0, 1.0).</para>
+
+      <para>The above explanation is a slightly modified excerpt from
+      <ulink url="http://ninedegreesbelow.com/photography/display-referred-scene-referred.html";>
+      Models for image editing: Display-referred and scene-referred</ulink>.
+      The modified excerpt was written and quoted by permission of the
+      author, who has licensed the modified excerpt under the
+      <ulink url="http://creativecommons.org/licenses/by-sa/3.0/deed.en_US";>
+      Creative Commons Attribution-ShareAlike 3.0 Unported License</ulink>.
+      </para>
+    </glossdef>
+
+  </glossentry>
+    <glossentry id="glossary-display-referred-black">
+    <glossterm>
+      <phrase>Display-referred black</phrase>
+    </glossterm>
+    <indexterm significance="normal">
+      <primary>Display-referred black </primary>
+    </indexterm>
+    <glossdef>
+     <para>"Display-referred black" (or for simplicity, "black") means the
+     floating point RGB color (0.0, 0.0, 0.0) and its integer equivalents. This
+     color has the very special significance that there's no such thing as
+     "less bright than black". So in display-referred image editing, all RGB
+     channel values are greater than or equal to 0.0 and no color is less
+     bright than "black", (0.0, 0.0, 0.0).</para>
+
+      <para>The above explanation is a slightly modified excerpt from
+      <ulink url="http://ninedegreesbelow.com/photography/display-referred-scene-referred.html";>
+      Models for image editing: Display-referred and scene-referred</ulink>.
+      The modified excerpt was written and quoted by permission of the
+      author, who has licensed the modified excerpt under the
+      <ulink url="http://creativecommons.org/licenses/by-sa/3.0/deed.en_US";>
+      Creative Commons Attribution-ShareAlike 3.0 Unported License</ulink>.
+      </para>
+    </glossdef>
+  </glossentry>
+
   <glossentry id="glossary-dithering">
     <glossterm>
       <phrase>Dithering</phrase>
@@ -864,6 +1074,64 @@
     </glossdef>
   </glossentry>
 
+  <glossentry id="glossary-high-dynamic-range">
+    <glossterm>
+      <phrase>High Dynamic Range</phrase>
+    </glossterm>
+    <indexterm significance="normal">
+      <primary>High Dynamic Range</primary>
+    </indexterm>
+    <glossdef>
+      <para>With
+      <link linkend="glossary-display-referred">display-referred</link> data
+      you have roughly two and half stops of head room above middle gray and
+      maybe six and a half useable stops below middle gray, at which point the
+      data is too densely packed into too few tonal steps to accurately display
+      differences between solid black and "just barely gray". So at best you
+      have 9 stops of dynamic range, compared to the 20 or more stops of
+      dynamic range you might find in some (certainly not all!) real world
+      scenes.</para>
+
+      <para>The usual solution to the dynamic range limitations of
+      display-referred data is to allow channel values to be however high as is
+      needed to encode the scene data. This means allowing channel values that
+      are above display-referred white.</para>
+
+      <para>Several file formats currently supported by GIMP 2.10 can be used
+      to import and export high dynamic range images, including floating point
+      tiffs, OpenEXR, and FITS.</para>
+
+      <para>When working with high dynamic range data in GIMP 2.10, the
+      <link linkend="glossary-channel-encoding">channel encoding</link> does
+      need to be linear to avoid gamma artifacts.</para>
+
+      <para>Editing high dynamic range data requires that there isn't any
+      clamping code in editing operations and blend modes. At floating point
+      precision:</para>
+      <orderedlist>
+        <listitem><para>Many (but not all) GIMP 2.10 blend modes are unclamped,
+        including Normal, Addition, Subtract, Multiply, Lighten Only,
+        Darken Only, Difference, and the LCH and Luminance blend modes.
+        Blend modes such as Screen, Soft Light, and Overlay are not unclamped
+        as these operations are designed to work with display-referred data.
+        </para></listitem>
+        <listitem><para>Many (too many to list but certainly not all, as some
+        editing operations are designed to work with display-referred data)
+        GIMP 2.10 editing operations also are unclamped, including Levels,
+        Exposure, transforms such as scaling and rotating, and various filter
+        operations such as Gaussian blur.</para></listitem>
+      </orderedlist>
+
+      <para>Portions of the above explanation of "high dynamic range" are
+      slightly modified excerpts from the <ulink 
url="http://ninedegreesbelow.com/photography/display-referred-scene-referred.html#scene-referred";>Models for 
image editing: Display-referred and scene-referred</ulink>.
+      These excerpts are quoted by permission and the modified excerpts are
+      licensed as
+      <ulink url="http://creativecommons.org/licenses/by-sa/3.0/deed.en_US";>
+        Creative Commons Attribution-ShareAlike 3.0 Unported License</ulink>.
+      </para>
+    </glossdef>
+  </glossentry>
+
   <glossentry id="glossary-histogram">
     <glossterm>
       <phrase>Histogram</phrase>
@@ -1803,6 +2071,31 @@
     </glossdef>
   </glossentry>
 
+  <glossentry id="glossary-scene-referred">
+    <glossterm>
+      <phrase>Scene-referred</phrase>
+    </glossterm>
+    <indexterm significance="normal">
+      <primary>Scene-referred</primary>
+    </indexterm>
+    <glossdef>
+      <para>When speaking of images captured by a camera, scene-referred means
+      that the intensities in the image RGB channels are proportional to the
+      intensities in the scene that was photographed.</para>
+
+      <para>"Scene-referred" is not the same as
+      <link linkend="glossary-high-dynamic-range">high dynamic range</link>,
+      as the camera might have been aimed at a low dynamic range scene such as
+      a foggy early morning view. However, adding a light source to the
+      captured frame (eg the moon breaking through the clouds or a street lamp)
+      will turn even a foggy morning into a high dynamic range scene.</para>
+
+      <para>As lightwaves do combine linearly, by definition a scene-referred
+      image (whether real or imaginary) must be encoded linearly to preserve
+      the scene-referred nature of the data.</para>
+    </glossdef>
+  </glossentry>
+
   <glossentry id="glossary-supersampling">
     <glossterm>
       <phrase>Supersampling</phrase>
[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]