Rudolf Polzer
d5d6e98191
add option -nofastpoint; make old buggy behaviour default as it is typically a lot faster. To not use fast point lights, pass -nofastpoint but not -fast.
2012-02-09 11:28:49 +01:00
Rudolf Polzer
8dcd2d22e6
slow down non-fast operation but make it more accurate by not using falloff tolerance at all when -fast is not used
2012-02-09 10:51:09 +01:00
Rudolf Polzer
c30d367a12
FIXME comment because falloffTolerance is checked in non-fast too for point lights, should we really do that? Maybe add a -slow then...
2012-02-09 10:41:17 +01:00
Rudolf Polzer
c60d22dd26
swap values 0 and 1 for -exposure (now -exposure 0 does nothing, -exposure 1 did what 0 did before)
2012-02-09 10:40:48 +01:00
Rudolf Polzer
272a29bdd8
always use color normalization, vector normalization on colors is worng
2011-12-18 22:36:53 +01:00
Rudolf Polzer
bb15de2ddc
also support -sRGBcolor for floodlight
2011-12-14 09:51:15 +01:00
Rudolf Polzer
48a8cb1d67
mark two shader checks nullable
2011-12-09 17:24:33 +01:00
Rudolf Polzer
b104009ba6
no, we do NOT want to ignore caulk
2011-12-09 17:17:31 +01:00
Rudolf Polzer
35162b6b63
properly ignore caulk
2011-12-09 17:15:51 +01:00
Rudolf Polzer
aef478a99a
PROPERLY ignore sky brushes for minimap
2011-12-09 16:36:07 +01:00
Rudolf Polzer
e8f27310d6
also new option: -sRGBtex, and renamed -sRGB to -sRGBlight
2011-10-29 16:33:46 +02:00
Rudolf Polzer
5c97bccbf6
experimental q3map2 command line option: -sRGB (store lightmaps in sRGB)
2011-10-21 17:25:09 +02:00
Rudolf Polzer
c75d4746fd
some warning fixes
2011-06-05 19:32:39 +02:00
Rudolf Polzer
dbaf1a8f00
fix some more warnings
2011-06-05 19:21:31 +02:00
Rudolf Polzer
772c8fda0f
fix a scanf warning
2011-06-05 19:19:45 +02:00
Rudolf Polzer
6b48a837f9
forgot to check lightDeluxel for NULL here
2011-02-22 17:48:30 +01:00
Rudolf Polzer
62fdec79ee
local variables ALSO work better when declared
2010-10-30 18:10:23 +02:00
Rudolf Polzer
b7a62893a4
add a floodlightDirectionScale value to the _floodlight key too (whatever it does)
2010-10-30 18:07:37 +02:00
Rudolf Polzer
020a3febce
no need to artificially limit lightmap position for random supersampling
2010-10-30 16:37:04 +02:00
Rudolf Polzer
aa00101d5e
dirtmapping: PROPERLY prevent hitting sky
2010-10-30 16:36:53 +02:00
Rudolf Polzer
d1e3042389
support search box size in randomsamples too (although useless)
2010-10-22 09:35:41 +02:00
jal
8c7ca79ef5
Ignore transparent surfaces when floodlighting
2010-10-13 16:26:28 +02:00
Rudolf Polzer
18c11bf55e
YEAH! I think this fixes the lightmaps-of-patches bug!
2010-10-05 18:50:24 +02:00
rpolzer
5ec3a47e10
q3map2 is now waring free
2010-10-05 10:57:07 +02:00
rpolzer
5f1f257a1c
add a missing VectorClear
2010-10-04 10:48:22 +02:00
Rudolf Polzer
4c13f31535
use a Gaussian-like random distribution for random subsampling
2010-10-03 21:14:54 +02:00
Rudolf Polzer
02d17b50a7
also subsample the deluxemap (this was forgotten all the time!)
2010-10-03 20:50:09 +02:00
Rudolf Polzer
61d0c06cc6
option -randomsamples: makes -samples use adaptive random subsampling (only subsamples near edges, uses random distribution for luxels then)
2010-10-03 20:23:14 +02:00
Rudolf Polzer
c13784bc3a
force subsampling only where REALLY needed
2010-09-22 08:49:43 +02:00
Rudolf Polzer
382e2da9a0
fix compile error
2010-09-22 08:28:00 +02:00
Rudolf Polzer
d14718daef
force going all the way down when subsampling grates
2010-09-22 08:26:54 +02:00
Rudolf Polzer
fd6d38c979
force the first stage of subsampling on luxels that are hit through an alphashadow surface
2010-09-22 08:22:26 +02:00
Rudolf Polzer
4243255b4a
new funny option: -samplessearchboxsize, tunes the size of the search box of -samples (default 1 = "correct" size, can be up to 4)
2010-09-22 06:38:02 +02:00
Rudolf Polzer
5522a08f6c
fix the maths REALLY
2010-09-22 06:23:27 +02:00
Rudolf Polzer
07cf00e0a2
fix a small math error in -samples causing it to only ever consider 2/3 of the length of each luxel
2010-09-21 22:04:55 +02:00
Rudolf Polzer
e6ca10bec7
crlf fixes
2010-07-18 16:32:03 +02:00
jal
1e42a4f095
Fix vertexlighting and gridlighting being too dark when using floodlight and bounce (stop floodlighting being re-applied to lightmaps with each bounce stage)
2010-06-14 19:50:45 +02:00
jal
ff8feb8f31
Apply floodlight to vertex lighting (there seems to remain some problem with
...
vertexlighting not receiving enough bounced light)
2010-06-12 13:36:06 +02:00
jal
778af4dde0
add shader key q3map_noDirty, which disables the dirty pass on the surface. It's handy for lightmapped terrain models where dirty produces odd darkness on mesh splits.
2010-05-29 12:22:02 +02:00
jal
be4e243888
Increase the influence of directional lights on the deluxemap by a)Ignoring angle attenuation for directional contribution and b)Scaling down by 1/4 the contribution of radiosity (deeper explanation in a comment at LightContributionToSample).
2010-04-30 11:32:11 +02:00
jal
36e9024e86
fix deluxemapping when using floodlight. Also, add to floodlight the same trick as it's done to ambient to have a minimum normal assigned.
2010-04-22 12:07:00 +02:00
Rudolf Polzer
b8839b8771
small fix by jal regarding normal vectors from lighting
2010-02-14 20:09:15 +01:00
divverent
d909e5b8c7
patches by jal: fix some warnings, and spawnflags 64 on misc_model to not smooth the normals
...
git-svn-id: svn://svn.icculus.org/netradiant/trunk@324 61c419a2-8eb2-4b30-bcec-8cead039b335
2009-04-23 16:32:45 +00:00
divverent
f560e2703b
apply all of VorteX's changes except deviance (that one sucks :P)
...
git-svn-id: svn://svn.icculus.org/netradiant/trunk@237 61c419a2-8eb2-4b30-bcec-8cead039b335
2009-03-29 18:08:19 +00:00
rpolzer
aed6728816
dirt: don't treat skybox as solid
...
git-svn-id: svn://svn.icculus.org/netradiant/trunk@133 61c419a2-8eb2-4b30-bcec-8cead039b335
2008-11-28 10:15:18 +00:00
rpolzer
0178f5e7c6
fix crash with non-deluxe
...
git-svn-id: svn://svn.icculus.org/netradiant/trunk@119 61c419a2-8eb2-4b30-bcec-8cead039b335
2008-10-16 15:46:12 +00:00
rpolzer
f6b04e0860
deluxe: always use a MINIMUM light contribution from ambient
...
git-svn-id: svn://svn.icculus.org/netradiant/trunk@117 61c419a2-8eb2-4b30-bcec-8cead039b335
2008-10-12 13:44:03 +00:00
rpolzer
ef9e16cdb9
fix deluxemap when lit from the wrong side as good as possible (by ignoring the deluxemap contribution). now the invariant "deluxemap vector points away from the surface" should hold.
...
cull none + deluxemapping still is broken by design, but less broken now.
git-svn-id: svn://svn.icculus.org/netradiant/trunk@111 61c419a2-8eb2-4b30-bcec-8cead039b335
2008-09-30 13:29:31 +00:00
rpolzer
42d241347b
fix two deluxemap bugs
...
git-svn-id: svn://svn.icculus.org/netradiant/trunk@110 61c419a2-8eb2-4b30-bcec-8cead039b335
2008-09-30 13:07:10 +00:00
rpolzer
3b3f1bd8d1
remove not-normalized deluxemap support again - it did not look any better, still had enough of the usual deluxe drawbacks
...
git-svn-id: svn://svn.icculus.org/netradiant/trunk@109 61c419a2-8eb2-4b30-bcec-8cead039b335
2008-09-30 12:20:22 +00:00