[Bobboau]

and while were nit picking all our opinions of PCS, I'd also like to have that anoying "are you sure you want load that model" mesage removed

------------------
Bobboau, bringing you products that work.............. in theory

[Nico]

 

Originally posted by Bobboau:
and while were nit picking all our opinions of PCS, I'd also like to have that anoying "are you sure you want load that model" mesage removed

yeah, I already asked for that   it asks it even at the start, when there's no pof loaded  

[IceFire]

Looks great!

Venom: Very awesome planet models.  I also saw some new animations perhaps?

[Bobboau]

ya, that's how the Sharlin is is suposed to look

might there be a pach for TBP

------------------
Bobboau, bringing you products that work.............. in theory

[Nico]

 

Originally posted by IceFire:
Looks great!

Venom: Very awesome planet models.  I also saw some new animations perhaps?

thanx, but the ring is awfull, to be honest. it needs a better map.
the new anim? yup, a plasma kind of explosion.

for the sharlin, I suppose they'll just have the new models in the next release, you can't come close enough to them to see if they have shading or not in the demo anyway  

[jonskowitz]

 

Originally posted by Kazan:
it'll be smooth shading or not

it creates it's own shading data because TS calculates it dynamically and doesn't save it to the file

I'd ask Gary about one.  Using his COB2FS2 utility the model retains the same smoothing data it had in TS so it must be in the *.COB file somewhere.

------------------
I told you that It would be done by November, well, mostly anyway...

[Kazan]

 

Originally posted by venom2506:
for my part, i'd like PCS to save all the modification at the first time (it's not a big pb, just need to save several times, but wometimes, I just forget :þ). I didn't have thios pb with the last build, but woomeister told me he had, so...

im going to *****slap what ever control's not saving it the first time!

[Kazan]

 

Originally posted by jonskowitz:
I'd ask Gary about one.  Using his COB2FS2 utility the model retains the same smoothing data it had in TS so it must be in the *.COB file somewhere.

that's because you can calulate the smoothing according to what TS saves

generic part:

Code: [Select]

2 bytes short Material number
1 byte char Shader type   ('f'-flat, 'p'-phong, 'm'-metal)
1 byte char Facet type      ('f'-faceted, 'a'-autofacet, 's'-smooth)
1 byte char Autofacet angle   (0 - 179 degrees)
4 bytes float Red component of color   (0.0 - 1.0)
4 bytes float Green component of color   (0.0 - 1.0)
4 bytes float Blue component of color   (0.0 - 1.0)
4 bytes float Opacity   (0.0 - 1.0)
4 bytes float Ambient coefficient   (0.0 - 1.0)
4 bytes float Specular coefficient   (0.0 - 1.0)
4 bytes float Hilight size coefficient   (0.0 - 1.0)
4 bytes float Index of refraction


[Bobboau]

so is that how it's going to work when we next get to use it.

------------------
Bobboau, bringing you products that work.............. in theory

[Kazan]

here's why normal calculation fails sometimes
 http://astronomy.swin.edu.au/pbourke/geometry/insidepoly/

it's more of them failing to serve their function due to special circumstances breaking the equasions

[jonskowitz]

I see, thx for pointing it out.  Dumb question but am I right in assuming that's how your newest build of your utility (the one supporting smooth shading) does it.

[Kazan]

not yet.. im trying to find how out to calculate faceting from the angle

[Kazan]

i think autofacet works via a sort of 'weighted average' between the actual point normal and the average normal

120 degree is full smoothing
0 degree is faceted


trying to think what kind of function would produce that

perhaps

N = point normal
AN = Average Point Normal
M = new point normal
F = AutoFacet Angle

M = unitvector((((120-F)/120)N + (F/120)AN)/2)



[This message has been edited by Kazan (edited 11-24-2001).]

[Kazan]

it looks that as far as FS2 is concerned: you smooth it OR NOT

[Setekh]

Oooh, that smoothing is friggin' sweet  

As for FS2 smoothing/or not - I guess it's better than nothing...  

[Bobboau]

well just try to get it working so we can play with it.

------------------
Bobboau, bringing you products that work.............. in theory

[jonskowitz]

I know it can be done...
Heres a model in FS that was not smoothed...
   

...and here's one that was...
   

Both models exist in the same vp file and both pics are in-game (tech room shots).

I was thinking, it might help to look at ModelView32's source code and see if there's something in there that might help.  I think it's available at the descent developer's network site.

[This message has been edited by jonskowitz (edited 11-24-2001).]

[Kazan]

no model view's source is closed

[Setekh]

 

Originally posted by jonskowitz:
I know it can be done...
Heres a model in FS that was not smoothed...
     

...and here's one that was...
     

Both models exist in the same vp file and both pics are in-game (tech room shots).

I was thinking, it might help to look at ModelView32's source code and see if there's something in there that might help.  I think it's available at the descent developer's network site.

[This message has been edited by jonskowitz (edited 11-24-2001).]

Oh... damn...    

[jonskowitz]

I (obviously) don't know much about this stuff, but I went' through Knudd's COB2FS source code until I found something that looked relavent.  Hope this helps you out...

Code: [Select]

void calcNormal(float v[3][3], float out[3])
   {
   double v1[3],v2[3];
   static const int x = 0;
   static const int y = 1;
   static const int z = 2;

   v1[x] = v[0][x] - v[1][x];
v1[y] = v[0][y] - v[1][y];
   v1[z] = v[0][z] - v[1][z];

   v2[x] = v[1][x] - v[2][x];
   v2[y] = v[1][y] - v[2][y];
v2[z] = v[1][z] - v[2][z];

   out[x] = v1[y]*v2[z] - v1[z]*v2[y];
out[y] = v1[z]*v2[x] - v1[x]*v2[z];
out[z] = v1[x]*v2[y] - v1[y]*v2[x];

   ReduceToUnit(out);
   }