ENG New site

Advanced search

[ New messages · Forum rules · Members ]
  • Page 1 of 1
  • 1
Texture and DEM for Vesta
JohnVVDate: Tuesday, 24.03.2015, 18:17 | Message # 1
Space Tourist
Group: Users
United States
Messages: 36
Status: Offline
xml errors on sending a PM

pm's are not working

----------------- quote ------------
i do not know if you got the long response , i got a XML error from the "send" button

the data i am using
see:
http://forum.celestialmatters.org/viewtopic.php?f=4&t=682
is the PDS data in
http://dawndata.igpp.ucla.edu/tw.jsp?....AMO_DTM
the ISIS3 cub map "global.cub " ( there is a PDS format image also "global.pds" )
and the texture from
http://dawndata.igpp.ucla.edu/tw.jsp?....aic_DLR
again -- the ISIS3 cub map "global.cub " ( there is a PDS format image also "global.pds" )
but there is also a png image

I use the blender ad on
http://wiki.blender.org/index.p....mporter
and a custom wrote headed ( for the add on )

this imports a certain type of DEM into blender as a 3d object

as to-----
"seam at 180/-180 latilude "

that looks like a null value pixel at 0 and at 360
----------------
"artifacts at pole (it was in the darkness, but this area could be smoothed) "

this is a common problem
use blenders "remesh" tool .This remeshes it using a QuadSphere mesh ( basically a cube map mesh)

"filled craters (maybe to hide artefacts)"
and
"clamped mountains (out of 16 bit range?) "

this is do to improper floating point to unsigned integer conversion
the data has values of radii in METERS
your conversion did not fit in the MAX 0 - 65536 for a 16 bit UNSIGNED integer

this is a BIG problem
" which is invisible on the map in GIMP, but clearly visible in SE: "

the gimp handles ONLY 8 bit data
not 16 bit nor 32 bit data

the gimp AUTO converts 16 bit unsigned into 8 bit unsigned

i use "nip2/ vips"
nip is the GUI to the vips image library
http://www.vips.ecs.soton.ac.uk/index.php?title=VIPS
it handles 16 and 32 bit images

that and i use ISIS3 on a redhat linux based system

you might want to post a question ( public) on
"The Celestial Matters Lounge"
http://forum.celestialmatters.org/viewforum.php?f=2

that way others can also LEARN something new

-- if needed i can add a tutorial

or

seeing as the forum looks to be working
i could post a tutorial here
BUT it will be for using a redhat type Linux Operating System
i do not own any Microsoft software and NEVER WILL !!!


Edited by JohnVV - Tuesday, 24.03.2015, 23:53
 
SpaceEngineerDate: Wednesday, 25.03.2015, 00:59 | Message # 2
Author of Space Engine
Group: Administrators
Russian Federation
Messages: 4800
Status: Offline
Quote JohnVV ()
this is a common problem
use blenders "remesh" tool .This remeshes it using a QuadSphere mesh ( basically a cube map mesh)

Hmm, interesting method. Is it possible to export back the elevation map? SE needed the map, not the mesh.

Quote JohnVV ()
this is do to improper floating point to unsigned integer conversion
the data has values of radii in METERS
your conversion did not fit in the MAX 0 - 65536 for a 16 bit UNSIGNED integer

I downloaded this map already with this error.

Quote JohnVV ()
this is a BIG problem
" which is invisible on the map in GIMP, but clearly visible in SE: "
the gimp handles ONLY 8 bit data
not 16 bit nor 32 bit data

Yeah, I probably tried to edit it in Photoshop, not in GIMP, few month ago, so forgot details.





 
HarbingerDawnDate: Wednesday, 25.03.2015, 01:05 | Message # 3
Cosmic Curator
Group: Administrators
United States
Messages: 8717
Status: Offline
Quote JohnVV ()
the gimp handles ONLY 8 bit data

GIMP 3.0 will fix this, fortunately.





All forum users, please read this!
My SE mods and addons
Phenom II X6 1090T 3.2 GHz, 16 GB DDR3 RAM, GTX 970 3584 MB VRAM
 
JohnVVDate: Wednesday, 25.03.2015, 01:54 | Message # 4
Space Tourist
Group: Users
United States
Messages: 36
Status: Offline
the vesta cub file
http://dawndata.igpp.ucla.edu/tw.jsp?....AMO_DTM

can be easily converted into a 32 bit gtiff ( geo- tiff )
( this software ALSO works on Microsoft )
GDAL
there are a few versions including a perl wrapped version
the program "gdal_translate" defaults to a tiff , so if you leave out te output format a tiff is default
-- text only , terminal ONLY ( cmd.exe for win )
Code

gdal_translate Vesta.global.cub Vesta.global.tiff

you get a about 570 Meg tiff image in meters
Min value = 211946
Max value = 293010

Screenshot in Nip2 ( also runs on windows )


just resize this 17281 x 8641 32 bit float image to your needed size
and if needed move it into 16 bit unsigned space

now once there is a normal everyday format image
then things can be done

do you need a 3d MESH
OR
a normal map

very large and VERY big file size "NormalMaps" are rather EASY to make from a Height map like the above
however there is normally one pit fall and PROBLEM
THE POLES"
one of the best solutions is to map it to wrap around a sphere and NOT a flat plain

Photoshop maps to a flat plain and NOT!!! a sphere
use DR. Schrempp ( phd) tools here
http://forum.celestialmatters.org/viewtopic.php?f=3&t=18

he has a GREAT tutorial on using the nmtools ( maps to simple cylindrical )
for use here will need to be converted to a cube map
( i use perl scripts used with "huggin" this )

if a 3d mesh

i use the FREE GPL'ed program
"Blender "
meshes are HUGE!!!!!! in file size
a 16 bit 8ppd image( 2880x 1440 ) is 7.9 Meg
and a uv mapped mesh is 456 MEG

7.9 VS 456 meg

so you will really NEED a lot of RAM in the computer

getting it into blender is a bit of work


but seeing as you really do not need a mesh

the DEM above should not have much, if any, artifacts at the poles.
-- it is one of the OFFICIAL released scientifically accurate data sets


Edited by JohnVV - Wednesday, 25.03.2015, 02:51
 
SpaceEngineerDate: Wednesday, 25.03.2015, 09:51 | Message # 5
Author of Space Engine
Group: Administrators
Russian Federation
Messages: 4800
Status: Offline
JohnVV, thanks for tutorial.

Quote JohnVV ()
very large and VERY big file size "NormalMaps" are rather EASY to make from a Height map like the above

I don't needed normal map, SE generates it automatically.

Quote JohnVV ()
Photoshop maps to a flat plain and NOT!!! a sphere

Do you mean cylindrical projection? Photoshop can't change the map projection.

Quote JohnVV ()
for use here will need to be converted to a cube map
( i use perl scripts used with "huggin" this )

I have a tool to convert a cylindrical map of any size into cubemap and then split it into png/jpg tiles. I opened non-splitted cube faces in Photoshop and tried to edit them to fix artefacts, but the problem is what artefacts are so tiny in amplitude on the 16 bit map, so they cannot be seen on the screen.

Quote JohnVV ()
the DEM above should not have much, if any, artifacts at the poles.
-- it is one of the OFFICIAL released scientifically accurate data sets

I guess this is the same map I have used before, but I will try to re-convert it into 16 bit with right dynamic range (to fix clamped craters and mountains).

PS: thread renamed.





 
JohnVVDate: Wednesday, 25.03.2015, 17:30 | Message # 6
Space Tourist
Group: Users
United States
Messages: 36
Status: Offline
for " normalizing " a image
8 bit > 1 to 256
16 bit SIGNED > -32767 to +32768
16 bit UNSIGNED > 1 to 65536

there are many ways

the vesta image has a range
Min value = 211946
Max value = 293010
this is above the 16 bit max value of 65536

nip /vips has built in math functions and so dose "gmic "
nip is manual so
subtract the MIN value
divide by the new MAX
( this gives you a image that has values from 0 to 1 )
multiply by 65536

Gmic uses number pairs and internally calculates the min/max
Code

gmic Input.tiff -n 1,65536 -type ushort -o OutpuImage.16U.tiff  


( i use a older version if gmic , changes were made in newer versions that disrupt my current workflow)
 
SpaceEngineerDate: Thursday, 26.03.2015, 09:27 | Message # 7
Author of Space Engine
Group: Administrators
Russian Federation
Messages: 4800
Status: Offline
Quote JohnVV ()
subtract the MIN value
divide by the new MAX
( this gives you a image that has values from 0 to 1 )
multiply by 65536


You must subtract MIN and divide on (MAX-MIN). My utility can do this automatically.

I downloaded the original cub file. I curious, why it called cub. I guess not because it is in a cubemap projection (that would be great!).





 
JohnVVDate: Thursday, 09.04.2015, 00:45 | Message # 8
Space Tourist
Group: Users
United States
Messages: 36
Status: Offline
How is this going

i might have made things sound more complex that they are
 
  • Page 1 of 1
  • 1
Search: