rthorvald wrote:t00fri wrote:The reason is quite simple and we already had a glimpse of it: smells of quarrel about whose images will be used after all. This is bound to happen if the hires package will be done as a community effort.
I don??t quite understand this. Surely, a *real* hi-res package should use the same source for graphics as the existing lores one? Or maybe this information don??t exist as hi-res data...
I think this should be turned upside down: the official textures should be produced in as high a resolution as possible, then scaled down. This will also give us a resource base to work with for the future. Doing it the other way will soon become difficult to keep track of.
Of course, this means a lot of work will have to be re-done. But what is the hurry... Not everything needs to be replaced at once. I??d be happy if 1.6 featured just *one* hi-res map. Besides, many objects do not have any hi-res info to work with at all today, so this package would take many years to complete anyway.
This means there will NOT be any discussion as to whose textures to use. It will just be a question of who has the time...
- rthorvald
Runar,
at present the maximum size of the default textures is limited to ONLY 2k! So even 4k textures would count as hires! There is virtually no major moon in our system where we don't have >= 4k imaging available.
As to the quarrels, let me just remind you of the recent one associated with the discussion about the surface color of Venus! When I am insisting I usually have some physics arguments in my mind. See e.g. a later summary here:
http://www.celestiaproject.net/forum/viewtopic.php?t=10611
I have a long recorded list of such disputes, notably also related to the treatments of LOK=limit of knowledge.
As a further example, take the texture of the Saturn rings.
A number of people have tried to do "better" than the Celestia default. There were hefty disputes, since some texture creators simply did not want to realize that there are two totally disjunct data sets that
need to be implemented: the ring
transparency measured from Earth via a star's transit across the rings
http://pds-rings.arc.nasa.gov/ringocc/ringocc.html
and the ring
albedo that can be reconstructed from Cassini imaging. With lots of effort Grant Hutchison and I have reconstructed the rings from
both these data sources .
Of course it is tempting to "create" larger sized ring textures from the more recent Cassini imaging alone. Yet the corresponding transparency is lacking and hence imagination has to replace the lacking data...
As concerns myself, I am not willing to make any compromises ending in the type of chunk textures found in dozens on the Motherlode...Since I am aware that many users wouldn't care for this amount of rigor, I keep going on in 'private' since a long time.
Actually, I personally use few planetary textures (beyond the default ones) that I have not made myself for Celestia according to strict criteria. Only some have been uploaded into the community. The reasons should be obvious: My textures are usually large, I deeply dislike the Motherlode concept and there is no other convenient storage so far.
Bye Fridger
(too bad I don't have such config... (cannot manipulate easely more than 8k maps))