Analysis of "harsh lighting."
Dennis E. Hamilton
dennis.hamilton at acm.org
Thu Mar 25 15:40:26 CET 2010
In researching Defect Report item SC34 N1309:03 on ODF 1.2 section 9.5.2 on
"harsh lighting" I have the following results of that investigation:
1. Harsh lighting is not a term of art for a special lighting mode in
computer graphics mode. The term is widely used in photography and it is
discussed in the context of computer graphics, but only ODF 1.0 seems to
have a harsh-lighting property for a light source.
2. In photography and computer graphics, harsh lighting is understood as
undiffused light from a point light source. An example is the light from
the mid-day sun on the Sahara Desert (or the open ocean or a mountain snow
field) in a cloudless, clear sky.
3. Most use of the term is around avoiding or devices for minimizing harsh
lighting, not selecting it. (Photoshop does not have a feature for
harshening the lighting in an image.) The same happens in theater and
cinematography. (Sometimes a light is described as too hot, and this seems
often related to harshness and intensity, not color temperature.)
APPLICATION IN ODF 1.0 AND PERHAPS OTHER DOCUMENT FORMATS
4. In computer graphics applications that provide an illumination source
other than ambient lighting and perhaps glowing (internally-lighted)
surfaces, such as that to which
draw:extrusion-first-(and-second)-light-harsh applies, THE LIGHTING IS HARSH
BY DEFAULT. IT IS WHAT YOU GET WITH SIMPLE RAY TRACING.
(draw:extrusion-first-light-harsh is also true by default.)
5. The reason that these OASIS discussions about Harsh Lighting are so high
in Web searches on harsh lighting is that TURNING ON HARSH LIGHTING IS
USUALLY NOT AN OPTION. What 3D Computer graphics users want to be able to
do is TURN IT OFF OR MITIGATE IT. No one else does anything active to turn
it on. It's what you get without some sort of intervention. Just like
photographing someone standing in bright sunlight.
6. These are, of course, my personal findings. I have no idea what the
drafters of ODF 1.0 had in mind for the behavior that draw:..-harsh="false"
would elicit in contrast with the default harsh=true case, since there is
apparently no simple true/false way to speak of light being "soft" or
7. I have added a comment to the ODF TC JIRA issue OFFICE-2200 that tracks
against the ODF TC disposition of SC34 N1309:03. You can see our full
issue-discussion record at
More information about the sc34wg6