Phobos impressions

Description: Tips for creating and manipulating planet textures for Celestia.

MiR
Topic author
MiR
Topic author
Posts: 247
With us: 7 years 8 months
Location: Germany

#1by MiR » 30.10.2010, 09:44

Just as I was leaving the dust-storming Mars I saw its moon Phobos and made this picture... :wink:

phobos016.jpg

Regards
Michael

Mynameislol
Mynameislol
Age: 23
Posts: 15
With us: 8 years 2 months
Location: Mercury

Re: Phobos impressions

#2by Mynameislol » 31.10.2010, 21:27

Wow, it looks finished already! Great work! :D

MiR
Topic author
MiR
Topic author
Posts: 247
With us: 7 years 8 months
Location: Germany

Re: Phobos impressions

#3by MiR » 02.11.2010, 13:40

Thanks, Mynameislol, you're very kind :D

I'm still working on it. Have to make the texture more accurate.
As soon as possible I will attach here a picture of the map.

Regards
Michael

Mynameislol
Mynameislol
Age: 23
Posts: 15
With us: 8 years 2 months
Location: Mercury

Re: Phobos impressions

#4by Mynameislol » 03.11.2010, 21:03

Your texture is accurate enough, it's the model that is inaccurate. Once again, great work! :wink:

t00fri
Avatar
t00fri
Age: 15
Posts: 8814
With us: 15 years 6 months
Location: Hamburg, Germany

Re: Phobos impressions

#5by t00fri » 03.11.2010, 22:10

Michael,

I suppose you are aware of the high-quality and scientifically accurate previous renderings of Phobos. What are the specific improvements over existing results that you are addressing in your ongoing project?

Fridger
Image

MiR
Topic author
MiR
Topic author
Posts: 247
With us: 7 years 8 months
Location: Germany

Re: Phobos impressions

#6by MiR » 04.11.2010, 19:25

Fridger,

t00fri wrote:I suppose you are aware of the high-quality and scientifically accurate previous renderings of Phobos.
Yes, I'm aware of this. Of course, I will keep this map scientifically...

t00fri wrote:What are the specific improvements over existing results that you are addressing in your ongoing project?
My Intention is to make the appearance of Phobos a bit more realistic. The current one (in Celestia) looks something strange; a bit like a drawing draft. And the other available map from DLR more as a "limit of knowledge"-map (imho).
I'm using both for this work, but there are some differences between these maps(?)

It's a lot of work. Because I try to realize this job as pedantically and meticulously as possible.
The project is still in a start-up phase. From time to time I will publish here the progress of the Phobos map.
And a supervising look on this work by an expert couldn't harm...

Your competent critics and suggestions are always welcome.

I appreciate your work here very much. :)

Regards
Michael

t00fri
Avatar
t00fri
Age: 15
Posts: 8814
With us: 15 years 6 months
Location: Hamburg, Germany

Re: Phobos impressions

#7by t00fri » 04.11.2010, 19:53

MiR wrote:Fridger,

t00fri wrote:I suppose you are aware of the high-quality and scientifically accurate previous renderings of Phobos.
Yes, I'm aware of this. Of course, I will keep this map scientifically...

t00fri wrote:What are the specific improvements over existing results that you are addressing in your ongoing project?
My Intention is to make the appearance of Phobos a bit more realistic. The current one (in Celestia) looks something strange; a bit like a drawing draft. And the other available map from DLR more as a "limit of knowledge"-map (imho).
I'm using both for this work, but there are some differences between these maps(?)

It's a lot of work. Because I try to realize this job as pedantically and meticulously as possible.
The project is still in a start-up phase. From time to time I will publish here the progress of the Phobos map.
And a supervising look on this work by an expert couldn't harm...

Your competent critics and suggestions are always welcome.

I appreciate your work here very much. :)

Regards
Michael


Michael,

I was referring to previous high-quality work about Phobos rendering outside the official Celestia distribution! There is notably what John Van Vliet and I did (independently). John used Blender a lot along with considerable "handicrafting" , while I used ONLY the published scientific vertex data for the (ascii) cmod model along with hires photographic imaging for the texture.

A fair amount of care went into the coloration by computer (GIMP) as usual. Here is the template I used which is part of the excellent HiRISE IRB color imaging of the MRO mission,

http://hirise.lpl.arizona.edu/phobos.php
Image

Another important focus in my work was to find a triangulation that entirely eliminates the infamous polar pinch effects! Note that one must always work simultaneously on modifications of the model AND the associated texture.

Let me refer you to some most detailed threads over at our CelestialMatters site as to my work on Phobos quite a while ago. John made also a number of contributions there.

1) Import Scientific Shape Models to CMOD without Blender!
http://forum.celestialmatters.org/viewtopic.php?t=337

In the second thread I have entirely solved the polar pinch problem!
2) Fighting the Polar Pinch (ascii CMOD shape models)
http://forum.celestialmatters.org/viewtopic.php?t=339

The third thread about Phobos is a "photoshooting" with some neat images based on my
Phobos rendering!
3) Fun with my 4k colored Phobos
http://forum.celestialmatters.org/viewtopic.php?t=333

+++++++++++++++++++++++++++++
My final result can also be downloaded from thread 2):
http://www.celestialmatters.org/users/t ... os_1.3.zip
As you can see, I have done a LOT of work about Phobos more than a year ago.

So let me ask again: What specifically are you trying to improve?
+++++++++++++++++++++++++++++

This archive contains everything, including the scientific vertex data, my perl script that does the optimized tesselations and my final 4k Phobos texture (phobos.jpg) together with the resulting ascii and binary CMOD shape models, where the binary version can just be plugged into Celestia (phobos.cmod).

My 4k texture has lots of detail and subtle colors while the corresponding CMOD model was directly constructed from scientific measurements:

Image

Image

Image

Here is a little video of my final Phobos rendering without any remaining polar pinch!

http://forum.celestialmatters.org/userp ... opinch.avi


Fridger

PS:
For completeness let me quote also the scientific ESA sites about Phobos, with their model reconstruction and a matching texture that resulted from 53 hires images!

http://www.esa.int/esaCP/SEMXE6RTKMF_Austria_1.html
http://www.geoinf.fu-berlin.de/eng/proj ... vation.php

PPS:
In case you are interested in high level texture work by myself, you might want to always have a look at our Celestial Matters site, since as a matter of principle I do NOT publish my work at the Motherlode! Some textures of mine are part of the official Celestia distribution, but there is lots of more recent stuff over at CM.
Image

MiR
Topic author
MiR
Topic author
Posts: 247
With us: 7 years 8 months
Location: Germany

Re: Phobos impressions

#8by MiR » 05.11.2010, 17:39

Mynameislol,
Mynameislol wrote:Your texture is accurate enough, it's the model that is inaccurate. Once again, great work! :wink:
A very well done model (low- and highpoly) by John Van Vliet is available here:
http://www.celestiamotherlode.net/catalog/marsmoons.php

I use the standard model (and a minimal configuration) for rendering tests. (Celestia starts faster)

Thanks again for your interest! :D

Regards
Michael

MiR
Topic author
MiR
Topic author
Posts: 247
With us: 7 years 8 months
Location: Germany

Re: Phobos impressions

#9by MiR » 05.11.2010, 19:09

Fridger,

your model of the Mars moon Phobos is simply fantastic 8O :-)

t00fri wrote:So let me ask again: What specifically are you trying to improve?
Probably you overestimate me. I'm not a very important person here.
My main motivation is only this: I wish to give something back of the joyfull hours I've spent (and still spending) with Celestia. That's all. But, of course, I am always trying to make a good job. Therefore: any help and criticism are very welcome.

For my renderings I use the same pages; images from Hirise, NASA, DLR, ESA...

Celestial Matters.org is well known to me. Very high-grade and scientific!

CelestialMatters could be (or is) a very good scientific (without Fiction) alternative to shatters.net. But it seem to me it is - unfortunately - fallen asleep a little. And as a newbie it was not quite clear to me - at least since I've read the "About Us" part - if people like me are welcome there.

t00fri wrote: My final result can also be downloaded from thread 2):
http://www.celestialmatters.org/users/t ... os_1.3.zip
I really do not know what's to do with the contents of this file? Is this Phobos? With texture?

t00fri wrote:Let me refer you to some most detailed threads over at our CelestialMatters site as to my work on Phobos quite a while ago. John made also a number of contributions there.

1) Import Scientific Shape Models to CMOD without Blender!
http://forum.celestialmatters.org/viewtopic.php?t=337

In the second thread I have entirely solved the polar pinch problem!
2) Fighting the Polar Pinch (ascii CMOD shape models)
http://forum.celestialmatters.org/viewtopic.php?t=339

The third thread about Phobos is a "photoshooting" with some neat images based on my
Phobos rendering!
3) Fun with my 4k colored Phobos
http://forum.celestialmatters.org/viewtopic.php?t=333
Thanks for the links. At the moment I make a short pause; I have to prepare a big presentation for Monday (Very important for my job).
I will read it.

I apologize for my stenographic writing style. I'm a little in time pressure...

Regards
Michael

t00fri
Avatar
t00fri
Age: 15
Posts: 8814
With us: 15 years 6 months
Location: Hamburg, Germany

Re: Phobos impressions

#10by t00fri » 05.11.2010, 19:58

Michael,
MiR wrote:Fridger,

your model of the Mars moon Phobos is simply fantastic 8O :-)

t00fri wrote:So let me ask again: What specifically are you trying to improve?
Probably you overestimate me. I'm not a very important person here.

No I don't ;-) . A sensible project requires first a solid study of what is known already and then a clearcut aim of what should be improved. Hence I was just inquiring about both these essential prerequisites.

And as a newbie it was not quite clear to me - at least since I've read the "About Us" part - if people like me are welcome there.

All interested CM users are welcome, of course. Unlike the Motherlode, we have however imposed serious restrictions about submitted contributions. All work including our own is subject to peer-reviewing by experts before publication in CM. That serves to retain highest possible standards and merely corresponds to good scientific practise. In the course of time, some particularly interesting contributors/authors have been invited to become CM members with personal web space etc. Their listing may be found under "Authors".

A good mastering of standard image manipulation techniques is of course required for publication of graphical work in CM. However, The scope of CM is way more general than Celestia. Hence people who can master the camera better than Gimp or Photoshop may well submit excellent photos (+ stories) to CM ...
Like e.g. these two ;-) :

http://forum.celestialmatters.org/userpix/4_4_1.jpg
(from http://www.carlosbcn.com/)

http://forum.celestialmatters.org/userp ... 04_4_1.jpg

t00fri wrote: My final result can also be downloaded from thread 2):
http://www.celestialmatters.org/users/t ... os_1.3.zip

I really do not know what's to do with the contents of this file? Is this Phobos? With texture?

Of course this is all one needs for displaying my 4k Phobos in Celestia: phobos.jpg is my final 4k texture. phobos.cmod is the corresponding binary cmod model. I suppose you know where to place textures and where to place models...

But I thought I've described this above? What in my view is much more important is to understand the underlying general procedure of tesselation that eliminates the polar pinch entirely AND directly works with the scientific vertex data as published! No fiddling in Blender etc.

Fridger
Last edited by t00fri on 05.11.2010, 23:55, edited 1 time in total.
Image

John Van Vliet
Avatar
John Van Vliet
Posts: 2651
With us: 15 years 1 month

Re: Phobos impressions

#11by John Van Vliet » 05.11.2010, 21:47

--- edit ---
Last edited by John Van Vliet on 19.10.2013, 08:34, edited 1 time in total.

danielj
danielj
Posts: 1481
With us: 14 years 2 months

Re: Phobos impressions

#12by danielj » 05.11.2010, 21:49

I downloaded Phobos 1.3.Put the texture in hi res folder and the model in models.Nothing happened.Then I put all the zip folder to Extras.The old texture remained.I renamed the "new" texture to phobos4K.jpg and changed the solarsys.ssc accordingly.NOTHING!What?s going on?

John Van Vliet
Avatar
John Van Vliet
Posts: 2651
With us: 15 years 1 month

Re: Phobos impressions

#13by John Van Vliet » 05.11.2010, 23:09

--- edit ---
Last edited by John Van Vliet on 19.10.2013, 08:33, edited 1 time in total.

t00fri
Avatar
t00fri
Age: 15
Posts: 8814
With us: 15 years 6 months
Location: Hamburg, Germany

Re: Phobos impressions

#14by t00fri » 05.11.2010, 23:47

:D

I always thought that meanwhile I have a reputation of

NOT MAKING ADD-ONS

Fridger

PS:
As John noted, originally, the archive did not contain my 4k Phobos texture in order to avoid people concluding that I have now started making add-ons ;-) . But since yesterday, the archive contains everything needed for a proper display of my Phobos rendering...and quite a bit more (for those with an urge to understand what I did)!
Image

Mynameislol
Mynameislol
Age: 23
Posts: 15
With us: 8 years 2 months
Location: Mercury

Re: Phobos impressions

#15by Mynameislol » 10.11.2010, 19:31

MiR wrote:Mynameislol,
Mynameislol wrote:Your texture is accurate enough, it's the model that is inaccurate. Once again, great work! :wink:
A very well done model (low- and highpoly) by John Van Vliet is available here:
http://www.celestiamotherlode.net/catalog/marsmoons.php

I use the standard model (and a minimal configuration) for rendering tests. (Celestia starts faster)

Thanks again for your interest! :D

Regards
Michael
I knew about John Van Vliet's phobos. But i noticed the model you where using was the standard, as you just told me.

John Van Vliet
Avatar
John Van Vliet
Posts: 2651
With us: 15 years 1 month

Re: Phobos impressions

#16by John Van Vliet » 10.11.2010, 19:47

--- edit ---
Last edited by John Van Vliet on 19.10.2013, 08:32, edited 1 time in total.

t00fri
Avatar
t00fri
Age: 15
Posts: 8814
With us: 15 years 6 months
Location: Hamburg, Germany

Re: Phobos impressions

#17by t00fri » 10.11.2010, 20:15

john Van Vliet wrote:
But i noticed the model you where using was the standard, as you just told me.
i do not think so
the default svn "phobos.cmob" is 145.2 KB
the mesh in the Phobos 1.3 zip is 738.2 KB

I never understand why people don't simply quote the references of the used data. I have given the reference for my input vertex data of Phobos in my respective CelestialMatters thread. Unfortuately, PDS changed the URL recently. So, here is a working reference again:
(I have also changed my original reference at CM)

http://sbn.psi.edu/pds/resource/oshape.html

The archive reads:
EAR-A-5-DDR-SHAPE-MODELS-V2.1/data/m1phobos.tab

NOTE: The scientific reference for Phobos' 3D shape only contains vertex data (16471), but NO (triangular) faces that are required for using ChrisL's cmodfix tool. The quoted Phobos shape model is by Peter Thomas/Cornell U., to which my new 4k texture was matched.

++++++++++++++++++++++
My Perl script that you find in my above Phobos_1.3.zip archive generates the missing triangulation in an optimized fashion, such that all traces of the Polar Pinch effect are eliminated. No fiddling with Peter Thomas' published data!
++++++++++++++++++++++
How I did this is explained in my corresponding CM thread:
http://forum.celestialmatters.org/viewtopic.php?t=339

As usual, the m1phobos.lbl file contains all the specs of the data set.

t00fri wrote:All interested CM users are welcome, of course. Unlike the Motherlode, we have however imposed serious restrictions about submitted contributions.

This is one of the reasons that i do not have all that many posts there

i tend to the "artistic" side of things

Then the Motherlode is just the right addess for your texture work.

Fridger
Image

Fenerit M
Avatar
Fenerit M
Posts: 1874
With us: 10 years 6 months
Location: Thyrrenian sea

Re: Phobos impressions

#18by Fenerit » 10.11.2010, 23:15

BTW, then I shut here, Fridger's Phobos works perfecly on my system even after having removed the texture's declaration from within the CMOD itself and having instead assigned to it its "phobos.jpg" texture through SSC, once moved inside the "../medres" folder (with normalmap too, made from it). Thus, Danielj:

- put "phobos_ascii.cmod" inside the main "../models" folder
- open it within an editor and do comment (#) the string "texture0"
- convert the "phobos_ascii.cmod" in binary through the CMODview (make the normals, first)
- rename the file as "phobos.cmod"
- put the texture "phobos.jpg" inside the main "..textures/medres" folder
- check that your Texture SSC's Phobos directive has the "jpg" as extension, otherwise do change it, together with the name.
- Start Celestia and go to Phobos.
Never at rest.
Massimo

MiR
Topic author
MiR
Topic author
Posts: 247
With us: 7 years 8 months
Location: Germany

Re: Phobos impressions

#19by MiR » 22.01.2011, 09:22

Here I go now...

I am still unsure which one of the available Phobos-map is the best...? :?

Because they are all different...

:?: :idea:

t00fri
Avatar
t00fri
Age: 15
Posts: 8814
With us: 15 years 6 months
Location: Hamburg, Germany

Re: Phobos impressions

#20by t00fri » 22.01.2011, 10:57

MiR wrote:Here I go now...

I am still unsure which one of the available Phobos-map is the best...? :?

Because they are all different...

:?: :idea:

It's obvious why they are different: only one used exclusively the measured and published scientific modeling data and a computer coloration from a truecolor photographic template.

Note the file phobos_ascii.cmod just contains the vertices of the Scientific Phobos shape model by Peter Thomas/Cornell University,

http://sbn.psi.edu/pds/asteroid/EAR_A_5 ... phobos.tab
-------------------------m1phobos.lbl -----------------------------

Code: Select all

PDS_VERSION_ID        = PDS3                                                 
                                                                             
RECORD_TYPE           = "FIXED_LENGTH"                                       
RECORD_BYTES          = 32                                                   
FILE_RECORDS          = 16471                                                 
                                                                             
^TABLE                = "M1PHOBOS.TAB"                                       
                                                                             
DATA_SET_ID           = "EAR-A-5-DDR-SHAPE-MODELS-V2.1"                       
PRODUCT_NAME          = "M1 PHOBOS SHAPE MODEL"                               
PRODUCT_ID            = "SHAPE-M1PHOBOS-200202"                               
INSTRUMENT_HOST_NAME  = {"VIKING ORBITER 1", "VIKING ORBITER 2"}             
INSTRUMENT_NAME       = {"VISUAL IMAGING SUBSYSTEM CAMERA A",                 
                         "VISUAL IMAGING SUBSYSTEM CAMERA B"}                 
TARGET_NAME           = "M1 PHOBOS"                                           
START_TIME            = 1976-06-22  /* Viking Orbiter Mission Dates */       
STOP_TIME             = 1980-07-30                                           
PRODUCT_CREATION_TIME = 2000-06-01  /* Peer Review Date */                   
RECORD_FORMAT         = "3(2X,F8.4)"                                         
                                                                             
OBJECT     = TABLE                                                           
  ROWS               = 16471                                                 
  ROW_BYTES          = 32                                                     
  INTERCHANGE_FORMAT = "ASCII"                                               
  COLUMNS            = 3                                                     
  DESCRIPTION        = "Numerical shape model of M1 Phobos, derived           
    from Viking orbiter images with some checks with Mariner 9 and           
    Phobos 2 data.  The model has uncertainties of a few hundred meters       
    in places, especially between 35 and 190 degrees longitude.  This         
    shape model consists of a grid with points every 2 degrees in             
    latitude and longitude, with a radius in km associated with each         
    point.  The model is written with both 0 and 360 longitude entries.       
    Coordinates are planetocentric.  Generation is described in Thomas       
    (1993).                                                                   
                                                                             
    Phobos rotates synchronously with its orbital period, with libration     
    of about 1 degree.  Spin pole: RA=317.68, Dec= 52.90  J2000 in degrees.   
    Prime meridian W = 35.06 + 1128.8445850d   in degrees, where d is the     
    number of days from the standard epoch (JD 2451545.0).                   
                                                                             
    See Davies, et al. (1996) for slight variations with precession;         
    Duxbury and Callahan (1989) for libration."                               
 ...


I tesselated these vertices with triangles such that there is NO polar pinch effect remaining. There was NO handicrafting involved whatsoever.

Fridger
Image


Return to “Textures”

Who is online (over the past 5 minutes)

Users browsing this forum: 1 guest