|
Under the assumption that was considered to be clear enough for ODF 1.2 is also clear enough for ODF 1.1, I'm setting this issue to resolved.
**** A BIGGER ISSUE ***
There seems to be enough information to treat the default point sources provided for in 9.5.2. What is not specified at all is the consequence of specifying that one or both of the sources are NOT HARSH. There is no account for what is expected in those cases. Harsh lighting is the easier case and it is the default. Knowing what NOT HARSH involves is a bigger issue. I can't argue with the fact that current ODF 1.2 committee drafts also expect that "harsh light" is understood and don't bother to assert that it is as if from a point source (which if so well-defined, we should have been able to simply say). The more-serious omission with regard to what it means to be non-harsh would seem to apply there too. **** MORE DETAILS **** It appears to be understood that that harsh light is that from an undiffused point light source. For example, direct sunlight on the Sahara at 11am under clear skies is considered pretty harsh. The same is true on ice, snow, and at sea under bright daylight conditions. At solar distance, it might as well be a point light source. Most of the material I find is not about creating harsh lighting, but about diminishing its effects in photography and in digital photograph manipulation. Causing the appearance of harsh lighting by applying digital-darkroom techniques to a softly-lit image is actually quite involved. If you look at standard search results such as http://www.bing.com/search?q=harsh+lighting+computer+graphics&mkt=en-us you'll see that most use of the term is in how to avoid or overcome harsh lighting. It is understood what it is, but there doesn't seem to be much on how one would instruct a computer-graphic system to cause it. In fact, the flaw of most computer graphic simulation of lighting, historically, was that harsh lighting was the only lighting available. It is the easy default (as apparently for ODF 1.2 custom graphics too). Digging deeper, I find that the challenge in computer-graphics is avoiding harsh lighting, just as in photography. From those sources, one can find consensus that harsh lighting results from individual point light sources. To know its effect, apparently one must ascertain the intensity, the distance, and the assumed line of sight between the point source (if outside of the scene) and the illuminated scene. The effect has to be integrated with the effect of other light sources on the scene, transluscence in illuminated objects, etc. Reviewing section 9.5 and 9.5.2 in particular, there appears to be considerable information on position, intensity, and so on. There is not a lot on how this interacts with provisions of 9.4 around lighting (especially ambient lighting), if at all Apparently harsh light is not so much a well-defined term in computer graphics as much as it is the common default. Non-harsh lighting is the tricky part. The hole is that harsh="false" is completely undefined.
MY PROPOSED RESOLUTION:
1. Define draw:extrusion-first-light-harsh="true" and draw:extrusion-second-light-harsh="true", the defaults, to signify default harsh lighting, where harsh lighting is defined as undiffused point-source lighting impinging on the scene/shape as established by the other attributes governing those light sources. 2. Define the interpretation of draw:extrusion-first-light-harsh="false" and draw:extrusion-second-light-harsh="false" to each be implementation-dependent. See also Specific proposal text (consistent with the proposal for ODF 1.2 in
In section 9.5.2 subheading "Extrusion First Light Harsh" Add the sentences """ When the attribute value is true, the light is harsh. The light is an undiffused point source. When the attribute value is false, the light is not harsh. The means by which the light is softened is implementation-dependent. """ In section 9.5.2 subheading "Extrusion Second Light Harsh" Add the sentences """ When the attribute value is true, the light is harsh. The light is an undiffused point source. When the attribute value is false, the light is not harsh. The means by which the light is softened is implementation-dependent. """ The Errata item for 9.5.2 "Extrusion First Light Harsh" does not include instructions on what to do with the text that is there.
The equivalent Errata item for 9.5.2 "Extrusion Second Light Harsh" is missing from Errata CD04-rev03. Errata 02 - ODF 1.2 Reconciliation:
The ODF 1.2 sections corresponding to the ODF 1.0 9.5.2 subsections reflect the improvements of Errata 02 and no further reconciliation is required. |
|||||||||||||||||||||||||||||||||||||||||
 |
|
Bug
Applied
Major
draw:extrusion-first-light-harsh [N 1309]
OFFICE-1475. For ODF 1.2, the current specification was considered to be sufficient. The description was not changed.