Posted: Thu Nov 26, 2009 11:22 am
Just add an unique role to your ship and then add ships with that role.goran wrote:Don't know how to add specific ship (not ships by role).
Page 100 of 460
Posted: Thu Nov 26, 2009 12:53 pm
Thanks.Zieman wrote:Just add an unique role to your ship and then add ships with that role.goran wrote:Don't know how to add specific ship (not ships by role).
Tested and fps drops 2fps per ship added.
Posted: Thu Nov 26, 2009 4:15 pm
My like too. But the vertice count should come down a bit 
.
Perhaps a bit of smoothing and spiffy specular map to polish it?
I was also wondering if those many little engines could be simulated by solid cylinders with a strong indent in the end resulting from a normal map. It would be very fake, but it might not be visible from any reasonable distance. It may reduce the vertice count a lot because you do not need the insides of the engine tubes..
Maybe also an idea for other small hollow pipe bits on ships in general?
Best wishes,
Oscar
.Perhaps a bit of smoothing and spiffy specular map to polish it?
I was also wondering if those many little engines could be simulated by solid cylinders with a strong indent in the end resulting from a normal map. It would be very fake, but it might not be visible from any reasonable distance. It may reduce the vertice count a lot because you do not need the insides of the engine tubes..
Maybe also an idea for other small hollow pipe bits on ships in general?
Best wishes,
Oscar
Posted: Thu Nov 26, 2009 5:14 pm
Smoothing only ads more vertices, at least in Wings. Specular/normal maps and shading are still rocket science to me.ovvldc wrote:My like too. But the vertice count should come down a bit
Perhaps a bit of smoothing and spiffy specular map to polish it?
I managed to more than halve number of polys and vertices but engines doesn't look so good now.
I turned those 3d cylinders into curved planes but now oolite doesn't show the hole at the ends, displaying filled cylinder instead. My lack of knowledge is still massive.ovvldc wrote:I was also wondering if those many little engines could be simulated by solid cylinders with a strong indent in the end resulting from a normal map. It would be very fake, but it might not be visible from any reasonable distance. It may reduce the vertice count a lot because you do not need the insides of the engine tubes..
As this is a Screenshot thread, I apologize for discussing here.
Posted: Thu Nov 26, 2009 7:55 pm
It would be very visible, but you could reduce this using a parallax map. Conveniently, the standard shader supports a combined normal_and_parallax_map.ovvldc wrote:I was also wondering if those many little engines could be simulated by solid cylinders with a strong indent in the end resulting from a normal map. It would be very fake, but it might not be visible from any reasonable distance.
Smoothing with the smooth key in Oolite only manipulates vertex normals. However, mixing smooth surfaces and hard edges requires the use of smoothing groups, which is a bit of a black art. I can’t actually remember whether you can export them from Wings, but specifying hard edges in Wings should be a starting point. (Select edges in edge mode, choose “Hardness” from the context menu, then “Hard”.) Frame recently came across some problems with the Obj2DatTex tool and smooth groups. Frame?goran wrote:Smoothing only ads more vertices, at least in Wings.ovvldc wrote:Perhaps a bit of smoothing and spiffy specular map to polish it?
Posted: Fri Nov 27, 2009 11:56 am
Yes the way obj2dattex.py works in writing / determining smooth groups for Oolite Dat model files is faulty if..Ahruman wrote:Frame recently came across some problems with the Obj2DatTex tool and smooth groups. Frame?
the OBJ file in question has disorganised groups, meaning that the same group is several places in the obj file.. just representing different normals
this means you have to modify the obj2dattex file..
They way I have done it is to allow for explicit smoothing groups, meaning that you cannot combine the different groups as you normally can..
Which now that I think of it is an impossible thing to do in Oolite... Oolite only accepts explicit smoothing groups not mixed smoothing groups.. (unless that is what the last two tokens are supposed to be used for but never am in obj2dattex.py)
Code: Select all
FACES
128,127,127, 0.00000,1.00000,0.00000, 3, 2,1,0
.........X.....X..
Any way here is my alteration for the obj2dattex.file.
this allows for 17 explicit smoothing groups..
Backup your own copy..
My Modified Python Script...
Download here
http://www.box.net/shared/g2tsau77ck
or alter your own accordingly
Cheers Frame...
Code: Select all
#!/usr/bin/python
# EXTENSIONS : "obj" "OBJ" # Accepted file extentions
# OSTYPES : "****" # Accepted file types
# ROLE : Editor # Role (Editor, Viewer, None)
# SERVICEMENU : Obj2DatTex/Convert to .dat # Name of Service menu item
"""
This script takes a Wavefront .obj file
and exports a .dat file containing the same trimesh.
Colour for the faces is set to flat grey (127,127,127)
and surface normals calculated for each triangle.
"""
import sys, string, math
def vertex_reference(n, nv):
if (n < 0):
return n + nv
return n - 1
inputfilenames = sys.argv[1:]
print "converting...this"
print inputfilenames
for inputfilename in inputfilenames:
outputfilename = inputfilename.lower().replace(".obj", ".dat")
if (outputfilename == inputfilename):
outputfilename = outputfilename,append(".1")
print inputfilename+"->"+outputfilename
inputfile = open( inputfilename, "r")
lines = inputfile.read().splitlines(0)
outputfile = open( outputfilename, "w")
mode = 'SKIP'
vertex_lines_out = ['VERTEX\n']
faces_lines_out = ['FACES\n']
n_verts = 0
n_faces = 0
skips = 0
vertex=[]
uv=[]
face=[]
texture=[]
uvForVertex=[]
uvsForTexture={}
textureForFace=[]
uvsForFace=[]
textureCounter = 0
interpretTexture = 0
materials = {}
max_v = [0.0, 0.0, 0.0]
min_v = [0.0, 0.0, 0.0]
# find materials from mtllib
for line in lines:
tokens = string.split(line)
#print "line :"
#print line
#print "tokens :"
#print tokens
if (tokens != []):
if (tokens[0] == 'mtllib'):
path = string.split(inputfilename, '/')
path[-1] = tokens[1]
materialfilename = string.join(path,'/')
print "going to open material library file: %s" % materialfilename
infile = open( materialfilename, "r")
mlines = infile.read().splitlines(0)
newMaterial = 0
for mline in mlines:
tokens1 = string.split(mline)
if (tokens1 != []):
if (tokens1[0] == 'newmtl'):
newMaterialName = tokens1[1]
newMaterial = 1
if (tokens1[0] == 'map_Kd'):
if (newMaterial):
materials[newMaterialName] = tokens1[1]
print "Material %s -> %s" % (newMaterialName, tokens1[1])
newMaterial = 0
#print "materials :"
#print materials
# find geometry vertices first
for line in lines:
tokens = string.split(line)
if (tokens != []):
if (tokens[0] == 'v'):
n_verts = n_verts + 1
# negate x value for vertex to allow correct texturing...
x = -float(tokens[1])
y = float(tokens[2])
z = float(tokens[3])
vertex.append( ( x, y, z) )
vertex_lines_out.append('%.5f, %.5f, %.5f\n' % ( x, y, z))
if (x > max_v[0]):
max_v[0] = x
if (y > max_v[1]):
max_v[1] = y
if (z > max_v[2]):
max_v[2] = z
if (x < min_v[0]):
min_v[0] = x
if (y < min_v[1]):
min_v[1] = y
if (z < min_v[2]):
min_v[2] = z
#print "vertex:"
#print vertex, len(vertex), n_verts
#print "\n"
# find texture coordinates next
for line in lines:
tokens = string.split(line)
if (tokens != []):
if (tokens[0] == 'vt'):
uv.append( ( float(tokens[1]), 1.0 - float(tokens[2])) )
#print "uv:"
#print uv, len(uv), n_verts
#print "\n"
# find faces next
# use red colour to show smoothing groups
"""
smoothing_group = 127
this makes very little sense since the groups in the obj files
do not come in any organised manner...
Frame 20-11-2009 16:27
"""
smoothing_group = 127;
group_token = 0;
# the following is a better way of organising our smoothing
smoothing_id = [] ;
smoothing_group_for_id = [];
for line in lines:
tokens = string.split(line)
if (tokens != []):
if (tokens[0] == 's'):
# we check the group number if it's zero this is a non-smoothed group
group_token = int(tokens[1])
if (group_token > 0):
# outcommented as this assumes organised non segmented smoothing
# smoothing_group = smoothing_group +1
# if (smoothing_group > 255):
# smoothing_group = 0
# Frame 27-11-2009 12:42
if(group_token == 1):
smoothing_group = 127;
if(group_token == 2):
smoothing_group = 128;
if(group_token == 4):
smoothing_group = 129;
if(group_token == 8):
smoothing_group = 130;
if(group_token == 16):
smoothing_group = 131;
if(group_token == 32):
smoothing_group = 132;
if(group_token == 64):
smoothing_group = 133;
if(group_token == 128):
smoothing_group = 134;
if(group_token == 256):
smoothing_group = 135;
if(group_token == 512):
smoothing_group = 136;
if(group_token == 1024):
smoothing_group = 137;
if(group_token == 2048):
smoothing_group = 138;
if(group_token == 4096):
smoothing_group = 139;
if(group_token == 8192):
smoothing_group = 140;
if(group_token == 16384):
smoothing_group = 141;
if(group_token == 32768):
smoothing_group = 142;
if(group_token == 65536):
smoothing_group = 143;
if(group_token == 131072):
smoothing_group = 144;
"""
if you need more than 17 smoothing groups, just continue like above
only group_token # 18 would be 131072*2 = 262144 and smoothing_group would be Equal to 145
Frame...
"""
if (tokens[0] == 'usemtl'):
textureName = tokens[1]
if (materials.has_key(textureName)):
textureName = materials[textureName]
interpretTexture = 1
texture.append(textureName)
uvsForTexture[textureName] = n_verts * [[]]
if (tokens[0] == 'f'):
#print "line: %s" % line
while (len(tokens) >=4):
bits = string.split(tokens[1], '/')
v1 = vertex_reference(int(bits[0]), n_verts)
if (bits[1] > ''):
vt1 = vertex_reference(int(bits[1]), n_verts)
bits = string.split(tokens[2], '/')
v2 = vertex_reference(int(bits[0]), n_verts)
if (bits[1] > ''):
vt2 = vertex_reference(int(bits[1]), n_verts)
bits = string.split(tokens[3], '/')
v3 = vertex_reference(int(bits[0]), n_verts)
if (bits[1] > ''):
vt3 = vertex_reference(int(bits[1]), n_verts)
else:
interpretTexture = 0
#print "face (geometry): %d %d %d" % (v1, v2, v3)
#print "face (textures): %d %d %d\n" % (vt1, vt2, vt3)
d0 = (vertex[v2][0]-vertex[v1][0], vertex[v2][1]-vertex[v1][1], vertex[v2][2]-vertex[v1][2])
d1 = (vertex[v3][0]-vertex[v2][0], vertex[v3][1]-vertex[v2][1], vertex[v3][2]-vertex[v2][2])
xp = (d0[1]*d1[2]-d0[2]*d1[1], d0[2]*d1[0]-d0[0]*d1[2], d0[0]*d1[1]-d0[1]*d1[0])
det = math.sqrt(xp[0]*xp[0] + xp[1]*xp[1] + xp[2]*xp[2])
if (det > 0):
n_faces = n_faces + 1
# norm = (xp[0]/det, xp[1]/det, xp[2]/det)
# negate the normal to allow correct texturing...
norm = ( -xp[0]/det, -xp[1]/det, -xp[2]/det)
face.append((v1,v2,v3))
faces_lines_out.append('%d,127,127,\t%.5f,%.5f,%.5f,\t3,\t%d,%d,%d\n' % (smoothing_group,norm[0],norm[1],norm[2],v1,v2,v3))
#
# check if we're in a non-smoothed group - if so keep incrementing the 'red' smoothing_group value...
#
if (group_token == 0):
# smoothing_group = smoothing_group + 1
if (smoothing_group > 255):
smoothing_group = 0
if (interpretTexture):
textureForFace.append(textureName)
uvsForTexture[textureName][v1] = uv[vt1]
uvsForTexture[textureName][v2] = uv[vt2]
uvsForTexture[textureName][v3] = uv[vt3]
uvsForFace.append([ uv[vt1], uv[vt2], uv[vt3]])
tokens = tokens[:2]+tokens[3:]
# begin final output...
outputfile.write('// output from Obj2DatTex.py Wavefront text file conversion script\n')
outputfile.write('// (c) 2005 By Giles Williams\n')
outputfile.write('// \n')
outputfile.write('// original file: "%s"\n' % inputfilename)
outputfile.write('// \n')
outputfile.write('// model size: %.3f x %.3f x %.3f\n' % ( max_v[0]-min_v[0], max_v[1]-min_v[1], max_v[2]-min_v[2]))
outputfile.write('// \n')
outputfile.write('// textures used: %s\n' % uvsForTexture.keys())
outputfile.write('// \n')
outputfile.write('NVERTS %d\n' % n_verts)
outputfile.write('NFACES %d\n' % n_faces)
outputfile.write('\n')
outputfile.writelines(vertex_lines_out)
outputfile.write('\n')
outputfile.writelines(faces_lines_out)
outputfile.write('\n')
# check that we have textures for every vertex...
okayToWriteTexture = 1
#print "uvsForTexture :"
#print uvsForTexture
#print "uvsForFace :"
#print uvsForFace
if (len(textureForFace) != len(face)):
okayToWriteTexture = 0
if (len(uvsForFace) != len(face)):
okayToWriteTexture = 0
for texture in textureForFace:
if (texture == ''):
okayToWriteTexture = 0
# if we're all clear then write out the texture uv coordinates
if (okayToWriteTexture):
outputfile.write('TEXTURES\n')
for i in range(0, len(face)):
facet = face[i]
texture = textureForFace[i]
uvForVertex = uvsForTexture[texture]
# outputfile.write('%s\t1.0 1.0\t%.5f %.5f\t%.5f %.5f\t%.5f %.5f\n' % (texture, uvsForFace[i][0][0], uvsForFace[i][0][1], uvsForFace[i][1][0], uvsForFace[i][1][1], uvsForFace[i][2][0], uvsForFace[i][2][1]))
mmod = 1
mdi= 1
mmdi = 1
outputfile.write('%s\t %.5f %.5f\t%.5f %.5f\t%.5f %.5f\t%.5f %.5f\n' % (texture, mmod, mmod, (uvsForFace[i][0][0]*mmdi), (uvsForFace[i][0][1]*mmdi), (uvsForFace[i][1][0]*mmdi), (uvsForFace[i][1][1]*mmdi), (uvsForFace[i][2][0]*mmdi), (uvsForFace[i][2][1]*mmdi)))
outputfile.write('\n')
outputfile.write('END\n')
outputfile.close();
print "done"
print ""
#
# end
#
Posted: Fri Nov 27, 2009 12:34 pm
shaders on Ahrumans new test planet object - this is just me mucking about not a preview of anything in the pipeline for Oolite's future just in case any devs get a heart attack, you can just about see them switching the street lights on in S America - the textures here are really low rez, 512px x 256px maps wrapped around a whole world and the normal map may actually be inside out - do the continents lower than the sea to any one else as well?
Posted: Fri Nov 27, 2009 12:37 pm
Its so purty...
Posted: Fri Nov 27, 2009 12:37 pm
OMG!Posted: Fri Nov 27, 2009 12:52 pm
The unused values really are unused. Originally the three first values for a face were its colour, but that isn’t used at all now.Frame wrote:(unless that is what the last two tokens are supposed to be used for but never am in obj2dattex.py)
Posted: Fri Nov 27, 2009 3:27 pm
that looks cracking Griff. the lower ones DO look recessed to me, but it's not like it ruins the effect.
how's the shader work w/ the planet rotation then, do you only see the lights when that side's in shadow?
how's the shader work w/ the planet rotation then, do you only see the lights when that side's in shadow?
Posted: Fri Nov 27, 2009 3:32 pm
Drool!
Griff, any chance of packaging that planet in a griff-earth oxp? I'd love to have a look at that in-game!
Griff, any chance of packaging that planet in a griff-earth oxp? I'd love to have a look at that in-game!
Posted: Fri Nov 27, 2009 3:33 pm
I think the east/west component of the normal map is OK, but the north/south component needs to be inverted.Griff wrote:shaders on Ahrumans new test planet object - this is just me mucking about not a preview of anything in the pipeline for Oolite's future just in case any devs get a heart attack, you can just about see them switching the street lights on in S America - the textures here are really low rez, 512px x 256px maps wrapped around a whole world and the normal map may actually be inside out - do the continents lower than the sea to any one else as well?
Posted: Fri Nov 27, 2009 4:05 pm
yes of course, it's hereKaks wrote:Drool!
Griff, any chance of packaging that planet in a griff-earth oxp? I'd love to have a look at that in-game!
http://www.box.net/shared/aapxtxx5a0
Posted: Fri Nov 27, 2009 5:22 pm
*Checks out Griff's planet in-game*
*Faints*
*Faints*
