Preview Phobos

[John Van Vliet]

--- edit ---

[W0RLDBUILDER]

Unfortunately, AFAIK, we have to stick to those preshaded/baked-shading maps.

it is NOT pre shaded just some of the shading in the orig photos is not removed.

with this detail on a small body ???
Celestia


Blender 2.5 alpha2


it is not needed that much

I know the details aren't pre-shaded, I just think that even with this detailed of a model, there still needs to be smaller details that could be easily created with a bumpmap. It just looks a bit too smooth. At least some LunarLambert would look good on this. <censored>, it looks good on all asteroids and irregular moons. An albedo map, too, would work wonders on this already awesome model.

[t00fri]

Here is my version of Phobos from our CelestialMatters site.
http://forum.celestialmatters.org/viewt ... 9&start=15
http://forum.celestialmatters.org/viewt ... 33&start=0

Shape model and texture were directly taken from scientific publications.
I.e. NO fiddling with Blender whatsoever! No remaining polar pinch after re-tesselation...

++++++++++++++++++++++++++++++
Have a look at this video:
http://forum.celestialmatters.org/userp ... opinch.avi
++++++++++++++++++++++++++++++

Coloration is computer-mapped using this true-color photo:











Fridger

[DataPacRat]

Here is my version of Phobos from our CelestialMatters site.
http://forum.celestialmatters.org/viewt ... 9&start=15
http://forum.celestialmatters.org/viewt ... 33&start=0

Shape model and texture were directly taken from scientific publications.
I.e. NO fiddling with Blender whatsoever! No remaining polar pinch after re-tesselation...

++++++++++++++++++++++++++++++
Have a look at this video:
http://forum.celestialmatters.org/userp ... opinch.avi
++++++++++++++++++++++++++++++

Coloration is computer-mapped using this true-color photo:

Fridger

Hello,

I haven't been able to find your updated Phobos model and texture on any Celestia site I know of; could you please tell me what it would take for me to produce similarly improved renderings of Phobos on my own computer?

[Fenerit]

Looking good, but it would be nice if Celestia supported bumpmapping/normalmapping on meshes. It would look even better then. Unfortunately, AFAIK, we have to stick to those preshaded/baked-shading maps.

Celestia support normalmap on meshes. Just for 3D CMOD meshes, though.

[t00fri]

Hello,

I haven't been able to find your updated Phobos model and texture on any Celestia site I know of; could you please tell me what it would take for me to produce similarly improved renderings of Phobos on my own computer?

For various reasons I never upload work of mine to the Motherlode. I am a Celestia developer and co-author, not an add-on creator. Hence my Celestia-related work mainly goes to the official Celestia distribution via the SVN archive or to our CelestialMatters site that hosts spacecraft models of superb quality, my texture tools for making highest quality virtual textures, and plenty of other stuff of interest.

I suggest you first start reading this thread about my Phobos approach:

http://forum.celestialmatters.org/viewt ... 37&start=0

There I explain in detail my Perl script for the CMOD conversion of the published data. The Perl script is also available there for download. Perl runs on EVERY operating system and is part of the system software in Linux and MAC OS X.

Thereafter read this thread

http://forum.celestialmatters.org/viewtopic.php?t=339

where I presented my results about an almost complete reduction of the familiar polar pinch effect at the level of (ascii) shape models from scientific publications! This just requires some re-tesselation near the poles, as is shown there. The final set of files is offered towards the end. Here is the link

http://www.celestialmatters.org/users/t ... os_1.3.zip

Fridger

[DataPacRat]

Hello,

I haven't been able to find your updated Phobos model and texture on any Celestia site I know of; could you please tell me what it would take for me to produce similarly improved renderings of Phobos on my own computer?

For various reasons I never upload work of mine to the Motherlode. I am a Celestia developer and co-author, not an add-on creator. Hence my Celestia-related work mainly goes to the official Celestia distribution via the SVN archive or to our CelestialMatters site that hosts spacecraft models of superb quality, my texture tools for making highest quality virtual textures, and plenty of other stuff of interest.

I suggest you first start reading this thread about my Phobos approach:

http://forum.celestialmatters.org/viewt ... 37&start=0

There I explain in detail my Perl script for the CMOD conversion of the published data. The Perl script is also available there for download. Perl runs on EVERY operating system and is part of the system software in Linux and MAC OS X.

Thereafter read this thread

http://forum.celestialmatters.org/viewtopic.php?t=339

where I presented my results about an almost complete reduction of the familiar polar pinch effect at the level of (ascii) shape models from scientific publications! This just requires some re-tesselation near the poles, as is shown there. The final set of files is offered towards the end. Here is the link

http://www.celestialmatters.org/users/t ... os_1.3.zip

Fridger

Thank you, Fridger, for your response. I've read the threads, and downloaded the zip file to try out the cmod file. While it certainly gets rid of the polar pinch, I'm a little confused, since it seems to be, overall, a less complicated texture than I was already using - smoother, with fewer lumps and bumps.

Let me see if I can show the difference...

The cmod I was already using:


Using your cmod instead:


Would applying your perl script to my existing, bumpier cmod remove the polar-pinch without smoothing the rest of the moon?


I've also noted in your posts at http://forum.celestialmatters.org/viewt ... 33&start=0 that you've mentioned having created a new 4k Phobos texture, color-matched to HiRSE; but I can't seem to find any such texture, not even at http://celestia.svn.sourceforge.net/vie ... /textures/ . Are you keeping it private for the moment, or have I simply not looked in the right place?

[John Van Vliet]

--- edit ---

[DataPacRat]

Would applying your perl script to my existing, bumpier cmod remove the polar-pinch without smoothing the rest of the moon?

No the script works from the *.tab data
m1phobos.tab > m1phobos.cmod
-- NOT --
old.cmod > new.cmod

Thank you for the clarification.

i have not looked for t00fri's map but as i recall it will be in the celestia-sci
as i recall it is posted on the celestiamatters forum , you might want to reread the CM forum .

I've just gone through every thread I could find there that seemed slightly relevant; and while I found a good number of screenshots, I wasn't able to find any actual textures/maps. The closest I could find was in http://forum.celestialmatters.org/viewtopic.php?t=333 , where Fridger provided two links to 5760x2880 textures he used.

t00fri and i differ on how the pole is fixed

i used Blender and did it by hand
t00fri wrote a script to have it "auto" done

both are good ways
his IS using newer data ( scientifically done)
mine is using "artistically" created data that matches with photos

I am going to admit that I don't have any particular concern about how this one problem is solved, as long as the solution works. I have an acquaintance who intends to illustrate a digest-comic with several pages set in and around Phobos; I'd like to be able to offer her the best reference images I can. The existing shape-model and texture I have are good, probably good enough - but if I can get her better ones, that would be, well, /better/.

[BobHegwood]

If I may here...

John has also uploaded both of these moons to the ML at the locations
listed below. These models are of the highest quality and have been
the result of MUCH work by John. Methinks that a little appreciation might be in order.

Deimos Very High Quality is HERE.
Phobos Very High Quality is HERE.

Thanks, Brain-Dead

[t00fri]

I am going to admit that I don't have any particular concern about how this one problem is solved, as long as the solution works. I have an acquaintance who intends to illustrate a digest-comic with several pages set in and around Phobos; I'd like to be able to offer her the best reference images I can. The existing shape-model and texture I have are good, probably good enough - but if I can get her better ones, that would be, well, /better/.

DataPacRat,

as a Celestia developer I am rather concerned about methods that allow a most direct transfer of scientifically validated data and imaging into Celestia and my forthcoming Celestia.Sci, respectively. Thus, the methods I am exploring have a way more general scope than just producing e.g. a texture & shape model for Phobos. The identical method can be and has been applied e.g to the asteroid Itokawa



So if you want the Phobos model with most bumps and holes, you should definitely download John VanVliet's add-on. Guided by subjective "aesthetic realism" , bumps and holes can easily be added with modelling software tools like Blender.

In contrast, a scientific approach only implements the 3d surface structures that have been clearly reconstructed in the scientific imaging process. Moreover, a given scientific texture must be perfectly matched to the shape model that comes with the texture. This is only guaranteed if one refers to the official scientific publications that release BOTH the shape model and the corresponding texture...

Since my new textures are supposed to be part of Celestia.Sci, it may be that I did not yet offer my latest Phobos update for download. Unfortunately, right now I have many of my textures packed in a 75GB (!) ZIP-archive that I needed to do because of a recent upgrade and hard-disk replacement. Unfortunately the unpacking takes hours

So just take John's Phobos add-on and you will be fine.

Fridger

[DataPacRat]

I do appreciate all the work that has been done on Phobos models and textures, and thank those who have done that work. John's model-and-texture add-on should do quite well enough as a reference for my artist friend, especially if I also show her the color-references Fridger posted.

(Nearly completely unrelated to the above; if I wanted to be silly, this would be the point where I linked to and suggested someone more skilled than I am at tweaking Celestia textures might enjoy figuring out the scale, and where on Phobos that could be dropped... )

[John Van Vliet]

--- edit ---

[Fenerit]

I do appreciate all the work that has been done on Phobos models and textures, and thank those who have done that work. John's model-and-texture add-on should do quite well enough as a reference for my artist friend, especially if I also show her the color-references Fridger posted.

(Nearly completely unrelated to the above; if I wanted to be silly, this would be the point where I linked to and suggested someone more skilled than I am at tweaking Celestia textures might enjoy figuring out the scale, and where on Phobos that could be dropped... )

Without tweaking textures, such view can be achieved with pure 3D models. See on Motherlode e.g:

[BobHegwood]

bob "These models are of the highest quality and have been "
this is only because i go back an air brush things
and in the process add data that was not there or remove data that was there.

Yes, I DO understand that John, but your models and textures are still of the
highest quality available to the Brain-Dead. All I was saying is THANKS.

[DataPacRat]

(Nearly completely unrelated to the above; if I wanted to be silly, this would be the point where I linked to and suggested someone more skilled than I am at tweaking Celestia textures might enjoy figuring out the scale, and where on Phobos that could be dropped... )

Without tweaking textures, such view can be achieved with pure 3D models. See on Motherlode e.g:

Hm... the AcrossToMars models are good enough for my purposes - with the minor exception that the AcrossToMars package's default installation puts them on Mars, rather than Phobos. I've been fiddling with my Celestia installation's .ssc files, and I'm able to kind of get the models in the general vicinity of Phobos... but floating over or under the surface, and rotated at quite weird angles.

Does anybody here know how to get, say, even just the AcrossToMars dome (from AcrossToMars's e_mars_landing.ssc file, the domefix.cmod mesh) placed somewhere on the surface of John van Vliet's high-poly Phobos, and positioned with its base in the right direction? If I can see the numbers for placing just one such model, I'm sure I'll be able to adapt the rest of the AcrossToMars models into a quick little Phobos base to help give my artist friend a sense of scale for Phobos...

[selden]

Specifying an object's orientation and long/lat on a spherical body are "easy" if you use one of v1.6.0's new directives.
Getting the altitude right is going to take trial-end-error, though, since Phobos is an irregular 3D model, not a sphere.

Here's the simplest SSC entry for placing an object on the surface of a sphere:

Code: Select all

SurfaceObject "Dome" "Sol/Mars/Phobos" {
       Mesh "domefix.cmod"
       Radius 0.01
       FixedPosition { Planetographic [ 90 45 0 ] }
}


For information about the new v1.6.0 SurfaceObject specification, see the Celestia WikiBook entry at
http://en.wikibooks.org/wiki/Celestia/1 ... _Placement

Unfortunately, the dome model in _Across to Mars_ is not designed in a way which would make placement and orientation particularly easy in 1.6.0. You can see how far off it is by selecting the object in Celestia and then using the right-mouse-button popup menu item ReferenceMarks / Show Planetographic Grid. You should be able to use "FixedRotation" to twist it around to an appropriate orientation. A TwoVector operation could be used, too.

If you view the planetographic grid for Phobos, you'll see just how irregular the moon is.

[DataPacRat]

Specifying an object's orientation and long/lat on a spherical body are "easy" if you use one of v1.6.0's new directives.
Getting the altitude right is going to take trial-end-error, though, since Phobos is an irregular 3D model, not a sphere.

Here's the simplest SSC entry for placing an object on the surface of a sphere:

Code: Select all

SurfaceObject "Dome" "Sol/Mars/Phobos" {
       Mesh "domefix.cmod"
       Radius 0.01
       FixedPosition { Planetographic [ 90 45 0 ] }
}


For information about the new v1.6.0 SurfaceObject specification, see the Celestia WikiBook entry at
http://en.wikibooks.org/wiki/Celestia/1 ... _Placement

Unfortunately, the dome model in _Across to Mars_ is not designed in a way which would make placement and orientation particularly easy in 1.6.0. You can see how far off it is by selecting the object in Celestia and then using the right-mouse-button popup menu item ReferenceMarks / Show Planetographic Grid. You should be able to use "FixedRotation" to twist it around to an appropriate orientation. A TwoVector operation could be used, too.

If you view the planetographic grid for Phobos, you'll see just how irregular the moon is.

I haven't figured out how to adjust altitude with the necessary level of precision yet, but with a bit of trial-and-error, I've been able to pick a latitude-and-longitude that intersects the dome with Phobos' surface, and a reasonably reasonable rotation. There's still a bit of oddness, especially in trying to get decent camera angles that aren't looking up from underneath Phobos' surface, but I think I'll be able to fake up the rest of a base from the pieces from AcrossToMars.

In case anyone's curious, I'm currently fiddling with these settings, though I'll probably end up looking for a hill near a pole with a better view of Mars. (I aimed for this point when I noticed that, at this time of Mars's year, this hillside is in near-constant sunlight.)

SurfaceObject "Dome" "Sol/Mars/Phobos"
{
Mesh "domefix.cmod"

Radius 0.006

FixedPosition { Planetographic [ 107.5 45.5 -1.385 ] }
FixedRotation {
Inclination 90
AscendingNode 180
MeridianAngle 0
}
}

[DataPacRat]

I think I've got a reasonably decent size-scale base set up, for my artist friend to zoom in and out and swing the camera around and so on. In case anyone's curious, here's the ssc info, and a quickie screenshot.

Code: Select all

SurfaceObject "Dome" "Sol/Mars/Phobos"
{
   Mesh "domefix.cmod"
   Radius 0.006
   FixedPosition { Planetographic [ 94.880 50.100 -1.230 ] }
   FixedRotation {
      Inclination 90
      AscendingNode 0
      MeridianAngle 0
}
}

SurfaceObject "Dude" "Sol/Mars/Phobos"
{
   Mesh "blue_spacemanfix.cmod"
   Radius 0.000914
   FixedPosition { Planetographic [ 94.800 50.050 -1.234 ] }
   FixedRotation {
      Inclination 90
      AscendingNode 60
      MeridianAngle 0
}
}

SurfaceObject "Cat" "Sol/Mars/Phobos"
{
   Mesh "cat_drivenfix.cmod"
   Radius 0.003
   FixedPosition { Planetographic [ 94.760 50.100 -1.2365 ] }
   FixedRotation {
      Inclination 90
      AscendingNode -35
      MeridianAngle 20
}
}

SurfaceObject "Tail" "Sol/Mars/Phobos"
{
   Mesh "trailerfix.cmod"
   Radius 0.003
   FixedPosition { Planetographic [ 94.720 50.120 -1.239 ] }
   FixedRotation {
      Inclination 90
      AscendingNode -35
      MeridianAngle 20
}
}


[Fenerit]

Hi DataPacRat, you are on the right path in these matters! Anyhow, for a best result and placement is supposed one did model its features, since they would starts already "best fitted" within the 3D modeler's workplace before to be posited above a planetary surface.

I haven't figured out how to adjust altitude with the necessary level of precision yet, but with a bit of trial-and-error, I've been able to pick a latitude-and-longitude that intersects the dome with Phobos' surface, and a reasonably reasonable rotation. There's still a bit of oddness, especially in trying to get decent camera angles that aren't looking up from underneath Phobos' surface, but I think I'll be able to fake up the rest of a base from the pieces from AcrossToMars.

If you are a luatools/luaplugins user, for precise picking of planetary lat/long see this thread:
http://www.shatters.net/forum/viewtopic.php?f=6&t=15864
in which there is a script that can help your business and:

http://www.shatters.net/forum/viewtopic.php?f=23&t=15891
in which the discussion is yielded under the form of a more extensive Lua plug-in