Project Reality Forums


https://forum.realitymod.org/

[Request] Help us to Optimize PR! (LOD Switching & Cull Distances)

https://forum.realitymod.org/viewtopic.php?t=92073

Page 1 of 6

[Request] Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-12 10:28
by Rhino
Hey guys,

We've identified recently that very few of the vehicles in PR have there "LOD Switching Distances" and our "Cull Distances" are setup correctly and are just using the default LOD switching distances and random cull distances which means the LODs are switching much later than they should with also some objects drawing much further than they need to be etc and everyone's GFX cards are working much harder than they need to be as a result.

This is an opportunity where you can help since this is something that isn't that hard to do. All that is required is some very basic coding knowledge and some basic experience of the BF2 Editor, with also a "pr_edit" set up, all of which there are explained in detailed tutorials for so you can really start off from scratch and learn everything you need to do this just from these tutorials.
It is how ever pretty time consuming and there are a lot of vehicles and possibly after that other objects that need doing and we don't have anyone in the team with the time to do this right now so if anyone from the community can help us with this, it would be greatly appreciated :D

If you are interested in helping out, set up your BF2 Editor and read though the tuts listed and then post your cull and lod switching distances in this topic and any issues you find with any of the LODs and providing your settings are all good, we will implement them into the next release

Cheers! :D

Files You Need:
  1. The BF2 Editor v1.3
    Link #1: Realitymod Files - BF2 Editor v1.3
    Link #2: FilePlanet BF2 Editor v1.3
    (Vista/Win7 users, remember to run the editor & the installer in XP compatibility mode and as admin)
Tutorials you need to read:
  1. Setting up the BF2 Editor for Project Reality Mapping/Modding
  2. Cull & LOD Switching Distances (Everyone needs to read this one, even if you have experience with doing LOD & Cull distances)
List of Vehicles To Do:

Air

Code: Select all

[list=1]
[*]gb_jet_tornadogr4
[*]gb_the_lynx
[*]gb_the_merlin
[*]ger_ahe_tiger
[*]ger_jet_tornadogr4
[*]ger_the_uh1d
[*]idf_jet_f16
[*]idf_the_uh60
[*]mec_ahe_ec635
[*]mec_ahe_havoc
[*]mec_ahe_sa341h
[*]mec_jet_mig29
[*]mec_jet_su25a
[*]mec_the_mi17
[*]mec_the_sa341h
[*]ru_ahe_havoc
[*]ru_jet_mig29
[*]ru_jet_su25a
[*]ru_the_mi8
[*]us_ahe_ah1z
[*]us_ahe_ah6
[*]us_ahe_ah6a
[*]us_ahe_kiowa
[*]us_ahe_uh1nrockets
[*]us_jet_a10a
[*]us_jet_f15
[*]us_jet_f16
[*]us_jet_f18
[*]us_jet_f35b
[*]us_the_mh6
[*]us_the_uh1n
[*]us_the_uh60
[/list]
Land

Code: Select all

[list=1]
[*]gb_aav_stormer
[*]gb_apc_warrior
[*]gb_ifv_scimitar
[*]gb_jep_landrover
[*]gb_tnk_challenger
[*]gb_trk_logistics
[*]gb_trk_support
[*]ger_trk_logistics
[*]ger_trk_support
[*]idf_apc_namer
[*]idf_jep_zeev
[*]idf_tnk_merkava
[*]mec_aav_gopher
[*]mec_apc_btr60
[*]mec_apc_mtlb
[*]mec_apc_mtlb_30mm
[*]mec_apc_mtlb_hmg
[*]mec_ifv_bmp3
[*]mec_tnk_t72
[*]mec_tow_shturm
[*]mil_tnk_t62
[*]ru_aav_gopher
[*]ru_aav_tunguska
[*]ru_apc_btr60
[*]ru_apc_btr80
[*]ru_apc_btr80a
[*]ru_apc_btr90
[*]ru_apc_mtlb
[*]ru_apc_mtlb_hmg
[*]ru_ifv_bmp3
[*]ru_tnk_t90
[*]ru_tow_shturm
[*]us_aav_avenger
[*]us_apc_aavp7a1
[*]us_apc_lav25
[*]us_apc_stryker
[*]us_ifv_m2a2
[*]us_tnk_m1a1
[*]us_tnk_m1a2
[/list]
J.F.Leusch69's Tasks (not to be undertaken by anyone else)

Code: Select all

[list=1][*]cf_trk_logistics
[*]cf_trk_support
[*]ch_trk_logistics
[*]ch_trk_support
[*]idf_trk_logistics
[*]idf_trk_support
[*]mec_trk_logistics
[*]mec_trk_support
[*]mil_trk_logistics
[*]mil_trk_support
[*]ru_trk_logistics
[*]ru_trk_support
[*]us_trk_logistics
[*]us_trk_support
[*]ger_apc_puma
[*]ger_ifv_puma
[*]idf_jep_hmmwv
[*]us_jep_hmmwv
[*]us_tow_hmmwv[/list]

Completed Vehicles:

Air

Code: Select all

[code][b]cf_the_chinook, ch_the_chinook, us_the_chinook[/b]

ObjectTemplate.cullRadiusScale 5

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 125
GeometryTemplate.setSubGeometryLodDistance 1 3 250
GeometryTemplate.setSubGeometryLodDistance 2 0 75
GeometryTemplate.setSubGeometryLodDistance 2 1 125
GeometryTemplate.setSubGeometryLodDistance 2 2 250



[b]Issues:[/b]
[list=1]
[*]Geom_1, LOD4, Final Tri count above the limit at 467 trsi.
[/list]


[b]Coding:[/b] ZAP44
[b]Checked by:[/b] [R-DEV]Rhino
[b]Tweaked by:[/b] [R-DEV]Rhino
[b]Committed to REPO:[/b] [R-DEV]Rhino

Code: Select all

[b][u]ch_ahe_z10, ch_ahe_z10_sp[/u][/b] 

ObjectTemplate.cullRadiusScale  3.2

GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 90
GeometryTemplate.setSubGeometryLodDistance 1 2 130
GeometryTemplate.setSubGeometryLodDistance 2 0 40
GeometryTemplate.setSubGeometryLodDistance 2 1 100

[b]Issues:[/b] 
[list=1]
[*]Geom_1 LOD0 Noticeable size increase of fuselage when switching...less so if value is increased to 60 or 70.
[*]Geom_1 LOD3 Final Tri count above the limit at 477 Tris.
[/list] 


[b]Coding:[/b] ZAP44
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b]ch_ahe_wz11, ch_ahe_wz11_sp[/b]

ObjectTemplate.cullRadiusScale 3.5

GeometryTemplate.setSubGeometryLodDistance 1 0 20
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 20
GeometryTemplate.setSubGeometryLodDistance 2 1 80


[b]Issues:[/b]
[list=1]
[*][color=Red]Geom_1 LOD3 Final Tri count above the limit at 722 Tris.[/color]
[*][color=Red]Geom_2 LOD2 Final Tri count above the limit at 719 Tris.[/color]
[/list]
          

[b]Coding: ZAP44
Checked by: [R-DEV]Rhino
Tweaked by: [R-DEV]Rhino
Committed to REPO:[/b] [R-DEV]Rhino

Code: Select all

[b][u]ch_jet_fantan[/u][/b] 

ObjectTemplate.cullRadiusScale  3.5

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 40
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 25
GeometryTemplate.setSubGeometryLodDistance 2 1 35

[b]Issues:[/b] 
[list=1]
[*]Geom_1 LOD0 Change in geometry of engines very noticeable when switching.
[*]Geom_1 LOD3 Final Tri count above the limit at 917 Tris.
[*]Geom_2 LOD2 Final Tri count above the limit at 1482 Tris.
[/list] 


[b]Coding:[/b] ZAP44
[b]Checked by:[/b]  [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b]  [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]ch_jet_j10, ch_jet_j10_sp[/u][/b] 

ObjectTemplate.cullRadiusScale  3.5

GeometryTemplate.setSubGeometryLodDistance 1 0 45
GeometryTemplate.setSubGeometryLodDistance 1 0 90
GeometryTemplate.setSubGeometryLodDistance 1 0 130

[b]Issues:[/b] 
[list=1] Geom_2 Final Tri count above the limit at 743 Tris.

[/list] 

[b]Coding:[/b] ZAP44
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]ch_jet_SU30, ch_jet_SU30_sp[/u][/b] 

ObjectTemplate.cullRadiusScale  2.5

GeometryTemplate.setSubGeometryLodDistance 1 0 40
GeometryTemplate.setSubGeometryLodDistance 1 0 90
GeometryTemplate.setSubGeometryLodDistance 1 0 150

[b]Issues:[/b] 
[list=1]
[*]Geom_2 Final Tri count above the limit at 902 Tris.
[/list] 


[b]Coding:[/b] ZAP44
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]ch_the_z8, ch_the_z8_sp, chthe_z8[/u][/b] 

ObjectTemplate.cullRadiusScale  2.7

GeometryTemplate.setSubGeometryLodDistance 1 0 40
GeometryTemplate.setSubGeometryLodDistance 1 1 80
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 30
GeometryTemplate.setSubGeometryLodDistance 2 1 100

[b]Issues:[/b] 
[list=1] None
[/list] 


[b]Coding:[/b] ZAP44
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]gb_ahe_apache, gb_ahe_apache_sp, idf_ahe_apache, idf_ahe_apache_sp, us_ahe_apache, us_ahe_apache_sp[/u][/b] 

ObjectTemplate.cullRadiusScale  3.3

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 100
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 100

[b]Issues:[/b] 
[list=1]
[*]Geom_1 Final Tri count above the limit at 1355 Tris.
[*]Geom_2 Final Tri count above the limit at 1205 Tris.
[/list] 


[b]Coding:[/b] ZAP44
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]gb_jet_eurofighter, gb_jet_eurofighter_sp, ger_jet_eurofighter, ger_jet_eurofighter_sp[/u][/b] 

ObjectTemplate.cullRadiusScale  3.6

GeometryTemplate.setSubGeometryLodDistance 1 1 40
GeometryTemplate.setSubGeometryLodDistance 1 2 80
GeometryTemplate.setSubGeometryLodDistance 1 3 140
GeometryTemplate.setSubGeometryLodDistance 2 1 30
GeometryTemplate.setSubGeometryLodDistance 2 2 100

[b]Issues:[/b] 
[list=1]
[*]Geom_1 Final Tri count above the limit at 506 Tris.
[*]Geom_2 Final Tri count above the limit at 380 Tris.
[/list] 

[b]Coding:[/b] ZAP44 
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]gb_jet_harrier, us_jet_harrier[/u][/b] 

ObjectTemplate.cullRadiusScale  3.2

GeometryTemplate.setSubGeometryLodDistance 1 1 20
GeometryTemplate.setSubGeometryLodDistance 1 2 40
GeometryTemplate.setSubGeometryLodDistance 1 3 70
GeometryTemplate.setSubGeometryLodDistance 2 1 15
GeometryTemplate.setSubGeometryLodDistance 2 2 60

[b]Issues:[/b] 
[list=1]
[*]Geom_1 Final Tri count above the limit at 2037 Tris.
[*]Geom_2 Final Tri count above the limit at 1421 Tris.
[/list] 

[b]Coding:[/b] ZAP44 
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69



[/code]

Civilian

Code: Select all

[CODE][b][u]civ_bik_atv[/u][/b]

ObjectTemplate.cullRadiusScale 13

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 150

[b]Issues:[/b]
[list=1]
[*]Geom_2, LOD2: missing wheel geometry (maybe is to much difference compared to LOD1, or maybe not) [url=http://img189.imageshack.us/img189/3393/58009845.jpg]Screenshot [/url]
[*]Geom_1, LOD3: buggy wheels, has 8 wheels instead of 4...
[*]Geom_1, LOD3: Final LOD is 1,136 tris, should be around 300 tris, preferably 100 tris.
[/list]


[b]Coding:[/b] Vaiper
[b]Checked by:[/b] [R-DEV]Rhino
[b]Tweaks:[/b] [R-DEV]Rhino
[b]Committed to REPO:[/b] [R-DEV]Rhino

Code: Select all

[b][u]civ_bik_dirtbike[/u][/b]

ObjectTemplate.cullRadiusScale 10

GeometryTemplate.setSubGeometryLodDistance 1 0 20
GeometryTemplate.setSubGeometryLodDistance 1 1 80
GeometryTemplate.setSubGeometryLodDistance 2 0 30
GeometryTemplate.setSubGeometryLodDistance 2 1 60

[b]Issues:[/b]
[list=1]
[*]Geom_1, LOD2: Final LOD's tri count over the limit, is currently 615 tris.
[*]Geom_2, LOD2: Final LOD's tri count over the limit, is currently 554 tris.
[*]Geom_2, LODs 0, 1 & 2: No Alpha material applied to the spokes on the rear wheel: [url]http://img443.imageshack.us/img443/5238/12142010004404.jpg[/url]
[/list]


[b]Coding:[/b] Vaiper
[b]Checked by:[/b] [R-DEV]Rhino
[b]Tweaks:[/b] [R-DEV]Rhino
[b]Committed to REPO:[/b] [R-DEV]Rhino

Code: Select all


[b][u]civ_bik_hondacb500[/u][/b]

ObjectTemplate.cullRadiusScale 8.8

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 60
GeometryTemplate.setSubGeometryLodDistance 1 2 90
GeometryTemplate.setSubGeometryLodDistance 1 3 110
GeometryTemplate.setSubGeometryLodDistance 2 0 20
GeometryTemplate.setSubGeometryLodDistance 2 1 55
GeometryTemplate.setSubGeometryLodDistance 2 2 100
GeometryTemplate.setSubGeometryLodDistance 2 3 110


[b]Issues:[/b][list=1]
[*]Geom_1 & Geom_2, LOD_3 [b]&[/b] LOD_4 Tri/poly counts too high. LOD_3 [1168] should be halved to ~500 and LOD4 [769] can be easily reduced to ~250. The fuel tank and detailed seat can probably be removed at these ranges and replaced by rectangles to fill the space (so that it's not transparent and thus an obvious LOD switch.
[*]Geom_1 & Geom_2, LOD_2->LOD_3: ([url=http://i235.photobucket.com/albums/ee94/psyrus_uraya/civ_honda_spokes.jpg]Image[/url]) The contrast in the 'rim' or 'spokes' of the wheel going from pure black to transparent at almost any range is very obvious. Would suggest a simple black object (rectangle perhaps) to fill the space so that the LOD switch can be pulled in sooner. 
[*]Geom_1 & Geom_2: As it stands there is very little difference between LOD_3 and LOD_4, and I think LOD_4 could be heavily optimized and the [i]LOD 3->4 switch pushed out further to ~150m[/i]
[*][[i]Not sure if issue[/i]] Geom_1 & Geom_2, LOD_3 & LOD_4 : Tail light on both geom_1 and wreck (geom_2) is floating in mid air. Would not be visible at the LOD range (90 & 100m respectively) in my opinion.
[/list]


[b]Coding:[/b] Psyrus
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] n/a
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]civ_jep_car, civ_jep_car_black, civ_jep_car_blue, civ_jep_car_white, civ_jep_car_bomber, civ_jep_car_bomber_black, civ_jep_car_bomber_blue, civ_jep_car_bomber_white[/u][/b]

ObjectTemplate.cullRadiusScale 15

GeometryTemplate.setSubGeometryLodDistance 1 0 45
GeometryTemplate.setSubGeometryLodDistance 1 1 60
GeometryTemplate.setSubGeometryLodDistance 1 2 140
GeometryTemplate.setSubGeometryLodDistance 2 0 45
GeometryTemplate.setSubGeometryLodDistance 2 1 90

[b]Issues:[/b]
[list=1]
[*]geom_1 lod_0 - RHS front seat appears to be missing
[*]geom_1 lod_2 & lod_3 - front & rear wheels are misaligned
[*]geom_2 lod_2 - Same as lod_1; [b][Final LOD] Poly limit way above limit @ 1542[/b] 
[*]geom_2 lod switch distance will probably have to be edited when final LOD is fixed
[/list]

[b]Coding:[/b] Psyrus
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]civ_jep_forklift[/u][/b]

ObjectTemplate.cullRadiusScale 1

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 60
GeometryTemplate.setSubGeometryLodDistance 1 2 95
GeometryTemplate.setSubGeometryLodDistance 2 0 30
GeometryTemplate.setSubGeometryLodDistance 2 1 70

[b]Issues:[/b]
[list=1]
[*]geom_1 LOD_3 - [b]Final LOD poly count above limit @ 590[/b]
[*]geom_2 LOD_2 - [b]Final LOD poly count above limit @ 839[/b]
[/list]

[b]Coding:[/b] Psyrus
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]civ_jep_support, civ_jep_support_militia[/u][/b]

ObjectTemplate.cullRadiusScale 4.5

GeometryTemplate.setSubGeometryLodDistance [b]N/A[/b]

[b]Issues:[/b]
[list=1]
[*]geom_1 & geom_2 - There is only one LOD for each geom!!
[*]geom_1 & geom_2 - [b]Only LOD is way above final LOD poly limit @ 3688[/b]
[/list]

[b]Coding:[/b] Psyrus
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]civ_jep_technical, civ_jep_technical_black, civ_jep_technical_blue, civ_jep_technical_militia, civ_jep_technical_red, civ_tow_technical, civ_tow_technical_militia[/u][/b]

ObjectTemplate.cullRadiusScale 9.7

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 135
GeometryTemplate.setSubGeometryLodDistance 2 0 40
GeometryTemplate.setSubGeometryLodDistance 2 1 70

[b]Issues:[/b][list=1]
[*]Geom_1, LOD3: front texture is plain white, which leads to it being quite hard to inconspicuously LOD switch from the front. If a texturer could place a bit of black smudging where the headlights are ([url=http://i235.photobucket.com/albums/ee94/psyrus_uraya/techie_bad_lod3_tex.jpg]see picture for difference[/url]) you could LOD switch earlier, at 100-110m imo.
[*]Geom_2, LOD1: The stark colour difference on the right hand side of the techie wreck LOD 0vs1 means that switching has to be pushed out, whereas I would've gone with 30m otherwise. Even @ 40m the difference is reasonably noticeable imo. 
[*]Geom_1 & 2, LOD3: It is also my humble opinion that the LOD3 wheels could be even more simple without much/any aesthetic loss.
[*][color=Red]Geom_2, LOD2: Final LOD's tri count way over the limit, tri count is over 1k tris.[/color]
[/list]


[b]Coding:[/b] Psyrus
[b]Checked by:[/b] [R-DEV]Rhino
[b]Tweaks:[/b] n/a
[b]Committed to REPO:[/b] [R-DEV]Rhino

Code: Select all

[b][u]civ_jep_technical2[/u][/b]

ObjectTemplate.cullRadiusScale 6

GeometryTemplate.setSubGeometryLodDistance 1 0 35
GeometryTemplate.setSubGeometryLodDistance 1 1 60
GeometryTemplate.setSubGeometryLodDistance 1 2 130
GeometryTemplate.setSubGeometryLodDistance 2 0 60
GeometryTemplate.setSubGeometryLodDistance 2 1 90
GeometryTemplate.setSubGeometryLodDistance 2 2 135

[b]Issues:[/b]
[list=1]
[*]geom_2 - lod_1, 2, 3 have deleted faces (transparent) and floating objects, not to mention disjointed faces. [url=http://i235.photobucket.com/albums/ee94/psyrus_uraya/technical_2_failLODs.jpg]See image[/url]. 
[*]geom_2 lod_3 - poly count is only 120 but there are many floating faces and the front RHS wheel is a mess.
[/list]

[b]Coding:[/b] Psyrus
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]civ_trk_dumpster, civ_trk_dumpster_bomber[/u][/b]

ObjectTemplate.cullRadiusScale 10

GeometryTemplate.setSubGeometryLodDistance [b][color=Red]N/A[/color][/b]

[b]Issues:[/b]
[list=1]
[*]geom_1, geom_2 - There is [b]NO[/b] LOD switching as it only has LOD_0 for each
[*]geom_1, geom_2 - Final (and only) LOD is [i]way[/i] over polygon limit @ 4318
[*]geom_1 lod_0 - Back of steering wheel appears white in three spots (no texture?) [url=http://i235.photobucket.com/albums/ee94/psyrus_uraya/gary_steering_white.jpg]See image[/url]. 
[/list]


[b]Coding:[/b] Psyrus
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]civ_trk_semi, civ_trk_semi_bomber[/u][/b]

ObjectTemplate.cullRadiusScale 7

GeometryTemplate.setSubGeometryLodDistance 1 0 40
GeometryTemplate.setSubGeometryLodDistance 1 1 80
GeometryTemplate.setSubGeometryLodDistance 1 2 135
GeometryTemplate.setSubGeometryLodDistance 2 0 95
GeometryTemplate.setSubGeometryLodDistance 2 1 135


[b]Issues:[/b]
[list=1]
[*]geom_0 - Seemingly not used for anything (?)
[*]geom_1 lod_0 -> lod_1 - Front grill texture is the only thing stopping the LOD switch occurring between 25-30m rather than 40. The size change of the grill makes switching earlier look bad. 
[*]geom_1 lod_1 -> lod_2 - RHS of truck colour change is quite obvious from dark to lighter. 
[*]geom_2 lod_0 -> lod_1 - Front fog lights are not present on LOD1 which means that the lod switch from the front is very obvious until the range is far enough to not distinguish the fog lights. If white blotches were added to the bumper texture or foglamps removed from lod_0 the switch could occur much earlier in the 25-40m range. [url=http://i235.photobucket.com/albums/ee94/psyrus_uraya/bigred_foglamps.jpg]See image[/url]
[*]geom_2 lod_2 - Blacked out windows mean that from any side angles the lod switch is very obvious, thus why it had to be kept so far @ 135m.
[/list]


[b]Coding:[/b] Psyrus
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69



[/code]

Land

Code: Select all

[code][b][u]cf_apc_lav3[/u][/b]

ObjectTemplate.cullRadiusScale 12

GeometryTemplate.setSubGeometryLodDistance 1 0 40
GeometryTemplate.setSubGeometryLodDistance 1 1 70
GeometryTemplate.setSubGeometryLodDistance 1 2 250
GeometryTemplate.setSubGeometryLodDistance 2 0 30
GeometryTemplate.setSubGeometryLodDistance 2 1 150

[b]Issues:[/b]
Geom_1, LOD1: Floating geometry above the turret: [url]http://realitymodfiles.com/rhino/tuts/lod-cull_distances/tut_lods-culls_distances07.jpg[/url]

[b]Coding:[/b] [R-DEV]Rhino
[b]Checked by:[/b] n/a
[b]Committed to REPO:[/b] Yes, [R-DEV]Rhino

Code: Select all

[b][u]cf_jep_gwagon[/u] [u]cf_jep_gwagon_support[/u] [u]ger_jep_gwagon_support[/u] [u]ger_jep_support_alt[/u][/b] 

ObjectTemplate.cullRadiusScale 10

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 80
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 30
GeometryTemplate.setSubGeometryLodDistance 0 1 100

[b]Issues:[/b] 
[list=1] cf_jep_gwagon
[*] geom_1 LOD3 polys 668
[/list] 

[b]Issues:[/b] 
[list=1] cf_jep_gwagon_support
[*] geom_2 LOD2 license plate visable in the right front wheel well
[/list]

[b]Issues:[/b] 
[list=1] ger_jep_gwagon_support
[*] geom_2 LOD2 license plate visable in the right front wheel well
[/list]

[b]Issues:[/b] 
[list=1] ger_jep_gwagon_support_alt
[*] geom 1 LOD3 poly count 1156
[*] geom 2 LOD1 bit of floating texture in front of left headlight
[*] geom 2 LOD2 license plate visable in the right front wheel well
[/list]


[b]Coding:[/b] SnipeHunt
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]cf_tnk_leo2a6[/u] [u]ger_tnk_leo2a6[/u][/b] 

ObjectTemplate.cullRadiusScale 4.8
  
GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 120
GeometryTemplate.setSubGeometryLodDistance 1 2 175

[b]Issues:[/b] 
[list=1] cf_tnk_leo2a6
[*] geom_2 LOD0 poly count 6279
[*] geom_2 has no LOD1 or LOD2
[/list] 


[b]Coding:[/b] SnipeHunt
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]ch_aav_type95[/u][/b] 

ObjectTemplate.cullRadiusScale 5.2

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 70
GeometryTemplate.setSubGeometryLodDistance 1 2 180
GeometryTemplate.setSubGeometryLodDistance 2 0 40
GeometryTemplate.setSubGeometryLodDistance 2 1 150

[b]Issues: None[/b] 

 

[b]Coding:[/b] SnipeHunt 
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b]ch_apc_wz551, ch_apc_wz551_bf2, ch_apc_wz551a[/b]

ObjectTemplate.cullRadiusScale 8

GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 100
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 100


[b]Issues:[/b]
[list=1]
[*]Geom_2, LOD1: propeller missing from lod0, minor issue.
[/list]
          

[b]Coding: [R-DEV]Rhino
Checked by: [R-DEV]Rhino
Tweaked by: [R-DEV]Rhino
Committed to REPO: [R-DEV]Rhino[/b]

Code: Select all

[b][u]ch_jep_paratrooper[/u][/b] 

ObjectTemplate.cullRadiusScale 9

GeometryTemplate.setSubGeometryLodDistance 1 0 35
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 40
GeometryTemplate.setSubGeometryLodDistance 2 1 90

[b]Issues: None[/b] 



[b]Coding:[/b] SnipeHunt 
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]ch_jep_uaz[/u] [u]mec_jep_uaz[/u][/b] 

ObjectTemplate.cullRadiusScale 7

GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 125
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 75

[b]Issues: [/b]
[list=1]
[*] geom_1 LOD 1 door handles removed from left side, still on the right side of model
[*] geom_1 LOD 2 ^^
[*] geom_1 LOD 3 poly count 1797
[*] geom_2 LOD 2 poly count 977
[/list] 



[b]Coding:[/b] SnipeHunt 
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]ch_jep_vn3[/u][/b] 

ObjectTemplate.cullRadiusScale 10.8

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 110
GeometryTemplate.setSubGeometryLodDistance 1 3 150
GeometryTemplate.setSubGeometryLodDistance 2 0 20
GeometryTemplate.setSubGeometryLodDistance 2 1 50
GeometryTemplate.setSubGeometryLodDistance 2 2 125
GeometryTemplate.setSubGeometryLodDistance 2 3 175

[b]Issues:[/b] 
[list=1]
[*] geom_1 LOD4 poly count 870
[*] geom_2 does not look wrecked at all.
[*] geom_2 LOD4 poly count 990
[/list] 


[b]Coding:[/b] SnipeHunt 
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]ch_tnk_type98[/u][/b]

ObjectTemplate.cullRadiusScale 5.5

GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 95
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 125

[b]Issues:[/b] 
[list=1]
[*]n/a
[/list]

[b]Coding:[/b] SnipeHunt
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]mec_aav_gaskin[/u][/b]

ObjectTemplate.cullRadiusScale 6

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 90
GeometryTemplate.setSubGeometryLodDistance 2 0 30
GeometryTemplate.setSubGeometryLodDistance 2 1 60

[b]Issues:[/b] 
[list=1]
[*] geom1 lod1 @6500polys
[*] geom1 lod2 @5000polys
[*] geom1 lod3 @4000polys
[*] geom2 lod1 @3400polys
[*] geom2 lod2 @2400polys 
[/list]

[b]Coding:[/b] Mats391
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]mec_jep_brdm2[/u] [u]mec_jep_brdm2_support[/u] [u]ru_jep_brdm2[/u] [u]ru_jep_brdm2_alt[/u] [u]ru_jep_brdm2_support[/u][/b] 

ObjectTemplate.cullRadiusScale  11

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 55
GeometryTemplate.setSubGeometryLodDistance 1 2 90
GeometryTemplate.setSubGeometryLodDistance 2 0 25
GeometryTemplate.setSubGeometryLodDistance 2 1 55

[b]Issues:[/b] 
[list=1] 
[*] geom1 lod1 @5500polys
[*] geom1 lod2 @4500polys
[*] geom1 lod3 @3500polys
[*] geom2 lod1 @4500polys
[*] geom2 lod2 @3300polys 
[/list] 

[b]Coding:[/b] Mats391
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]mec_tow_spandrel[/u] [u]ru_tow_spandrel[/u][/b]

ObjectTemplate.cullRadiusScale 11

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 55
GeometryTemplate.setSubGeometryLodDistance 1 2 90
GeometryTemplate.setSubGeometryLodDistance 2 0 35
GeometryTemplate.setSubGeometryLodDistance 2 1 65

[b]Issues:[/b] 
[list=1] 
[*] geom1 lod1 @5000polys
[*] geom1 lod2 @4000polys
[*] geom1 lod3 @3000polys
[*] geom2 lod1 @3500polys
[*] geom2 lod2 @1500polys 
[/list] 

[b]Coding:[/b] Mats391
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]idf_apc_m113[/u][/b]

ObjectTemplate.cullRadiusScale 9

GeometryTemplate.setSubGeometryLodDistance 1 0 40
GeometryTemplate.setSubGeometryLodDistance 1 1 60
GeometryTemplate.setSubGeometryLodDistance 2 0 35
GeometryTemplate.setSubGeometryLodDistance 2 1 55

[b]Issues:[/b]
[list=1]
[*]geom1 lod1 @5400polys
[*]geom1 lod2 @4500polys
[*]geom2 lod1 @4200polys
[*]geom1 lod2 @3200polys
[/list]

[b]Coding:[/b] Mats391
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69

Code: Select all

[b][u]idf_apc_m113_logistics[/u][/b]

ObjectTemplate.cullRadiusScale 9

GeometryTemplate.setSubGeometryLodDistance 1 0 40
GeometryTemplate.setSubGeometryLodDistance 1 1 60
GeometryTemplate.setSubGeometryLodDistance 2 0 35
GeometryTemplate.setSubGeometryLodDistance 2 1 65

[b]Issues:[/b]
[list=1]
[*]geom1 lod1 @4800polys
[*]geom1 lod2 @4000polys
[*]geom2 lod1 @3900polys
[*]geom1 lod2 @2200polys
[/list]

[b]Coding:[/b] Mats391
[b]Checked by:[/b] [R-DEV]J.F.Leusch69
[b]Tweaked by:[/b] [R-DEV]J.F.Leusch69
[b]Committed to REPO:[/b] [R-DEV]J.F.Leusch69



[/code]

Sea

Code: Select all

[code][b][u]boat_rib[/u][/b] 

ObjectTemplate.cullRadiusScale 4.5  

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 100
GeometryTemplate.setSubGeometryLodDistance 2 0 25
GeometryTemplate.setSubGeometryLodDistance 2 1 90

[b]Issues:[/b]
[list=1]
[*]n/a
[/list] 


[b]Coding:[/b]  SnipeHunt 
[b]Checked by:[/b] [R-DEV]Rhino
[b]Tweaked by:[/b] [R-DEV]Rhino
[b]Committed to REPO:[/b] [R-DEV]Rhino

[/code]






Template:
[php]

Code: Select all

[b][u]object_name[/u][/b]

ObjectTemplate.cullRadiusScale 

GeometryTemplate.setSubGeometryLodDistance 

[b]Issues:[/b]

[list=1]

[*]

[*]

[*]

[*]

[*]

[*]

[*]

[*]

[/list]


[b]Coding:[/b] *put your name here*
[b]Checked by:[/b] 
[b]Tweaked by:[/b] 
[b]Committed to REPO:[/b]
[/php]

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-14 00:52
by Rhino
ohhh nice, was begging to think that no one was going to help out with this hehe :mrgreen:
Vaiper wrote:Some thinks i have notice:

About geometry templates im not entirely sure, in editor without aa and windowed can look one way, while in game it could look different. Maybe LodDistance still should be higher, between 1-1 and 1-2...
I'm not entirely sure what you mean by this but like I said in my tut, you need to add a little extra distance above what you would think in the editor to compensate for high screen resolutions etc, the lack of AA wont change much in how an object looks at a distance.

Vaiper wrote:I checked with most of insurgent maps MaximumLevelViewDistance, in albasrash its 350, karbala, ramiel, kokan - 500, however in laskar valley its 800, so because of what most of civilians vehicles should be render 800 meter at least (set cullratio 8 ).

However im not sure about distace from UAV, for excample if UAV is floating like 500 over terrain, and vehicle is another 700 meter from uav, the overall distance would be ~860 meters, and its probably not rendered to comanders eye.

So i assume cullratio should be at least 10 for most of the vehicles.
As explained in my tut here: https://www.realitymod.com/forum/f189-m ... ances.html

Map View distance isn't a factor when doing LOD or cull distances as a new map could come out in the future with much higher view distances than any of the current maps and as such all objects need to be "future proofed". Having there cull distances go beyond a map's view distance isn't a problem since anything past a maps view distance dose not get drawn.
[R-DEV]Rhino wrote:For this weapon, I found a CullRadiusScale of 12 seems about right, where it then draws for 1,288 meters before it gets culled and at that point, all it is is a little pixel on the screen :)
Of course your probably thinking that having the object draw over 1km is pointless because we dont have any maps with a view distance over 1km, but its still a good idea to future proof everything just encase we do at some point get a map with a view distance of 2km or something :)

Now as for your tweaks, let me see.

Code: Select all

[b][u]civ_bik_atv[/u][/b] 

ObjectTemplate.cullRadiusScale  10

GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 80
GeometryTemplate.setSubGeometryLodDistance 1 2 250
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 150

[b]Issues:[/b] Geom_2, LOD2: missing wheel geometry (maybe is to much difference compared to LOD1, or maybe not) [url=http://img189.imageshack.us/img189/3393/58009845.jpg]Screenshot [/url]


[b]Coding:[/b]  
[b]Checked by:[/b]  
[b]Committed to REPO:[/b]
With your current cullRadiusScale of 10 the ATV only draws for 424m which isn't really enough since its quite big still.
http://img24.imageshack.us/img24/5350/1 ... 235544.jpg
http://img220.imageshack.us/img220/5303 ... 235600.jpg

A cullRadiusScale of 13 seems to work better, drawing it upto 576m which by then its only a couple of pixels :)
http://img254.imageshack.us/img254/8712 ... 235729.jpg
http://img88.imageshack.us/img88/2815/1 ... 235733.jpg

For the LOD switching distances, I found these better values, although your wreck lod distances where just right I found so nice work there, just your normal 3p lod distances needed tweaking a little :)

GeometryTemplate.setSubGeometryLodDistance 1 0 25 (was 50 before)
GeometryTemplate.setSubGeometryLodDistance 1 1 50 (was 80 before)
GeometryTemplate.setSubGeometryLodDistance 1 2 150 (was 250 before)
GeometryTemplate.setSubGeometryLodDistance 2 0 50 (no change)
GeometryTemplate.setSubGeometryLodDistance 2 1 150 (no change)

As for the wheel missing on the final wreck lod you spotted, nice catch there that is a bit of a problem, but with it showing up at only 150m it isn't noticeable, but if it didn't have that bug you could draw that lod much, much sooner :(

You also missed a pretty big issue here, with the final 3p LOD of the ATV having buggy wheels, with it having 8 wheels for some odd reason, and its final poly/tri count being way too high at 1136 tris, should be around 300 tris, really good final lods should be under 100 tris as stated in my tut :)
http://img199.imageshack.us/img199/3038 ... 000753.jpg

But other than that, nice work! :D

Code: Select all

[b][u]civ_bik_dirtbike[/u][/b] 

ObjectTemplate.cullRadiusScale  10

GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 70
GeometryTemplate.setSubGeometryLodDistance 1 2 250
GeometryTemplate.setSubGeometryLodDistance 2 0 30
GeometryTemplate.setSubGeometryLodDistance 2 1 150

[b]Issues:[/b] none


[b]Coding:[/b]  
[b]Checked by:[/b]  
[b]Committed to REPO:[/b]
Your cull distance seems about right :)

Right umm, the first major problem with your lod distances is that this model only has 3 lods (final lod being lod2) where you have defined distances for 4 lods. There is no "GeometryTemplate.setSubGeometryLodDistance 1 2" (lod3) lod for this model. Make sure that you have all the lods set up correctly in the code and all the lods are accounted for correctly, otherwise things can screw up bit time.

For the LOD switching distances, I found these better values.

GeometryTemplate.setSubGeometryLodDistance 1 0 20 (was 50 before)
GeometryTemplate.setSubGeometryLodDistance 1 1 80 (was 70 before)
- Removed: GeometryTemplate.setSubGeometryLodDistance 1 2 250 (not needed since this lod doesn't exist)
GeometryTemplate.setSubGeometryLodDistance 2 0 30 (no change)
GeometryTemplate.setSubGeometryLodDistance 2 1 60 (was 150 before)


As for Issues with the LODs, you missed one with the final 3p LOD not being around 300 tris, as also said above, a good final lod should be under 100tris where this one is over 600. http://img21.imageshack.us/img21/6772/1 ... 001812.jpg
Also the same issue with the wreck final lod, with its final lod being 554 tris. http://img443.imageshack.us/img443/5830 ... 001926.jpg

Also another issue with this bike is that its rear wheel on all its wreck LODs hasn't got a transparent alpha material set to it and as such, the rear wheel isn't see though. http://img443.imageshack.us/img443/5238 ... 004404.jpg
it should be like this (note this pic is of geom_1, 3p meshes): http://img406.imageshack.us/img406/7996 ... 004503.jpg



Apart from that it looks good :)
Please don't take this feedback the wrong way since it takes a bit of practice to get the values just right so after you have done a few more models with any luck I will not need to check them any more hehe :)

Committed the above values with fixes :)

Cheers! :mrgreen:

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-14 02:59
by ZAP44
What you seem to need doesn’t appear that demanding, more tedious than anything. I have tons of free time so am willing to have a go at it. I’m at the point of having unzipped the objects and was trying to interact with the object like making it move back to the point it disappears. No luck. I can make it do everything but that. Is there a listing of key commands for use with the editor somewhere? The help supplied with the editor is missing most of its content.

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-14 03:18
by Rhino
ZAP44 wrote:What you seem to need doesn?t appear that demanding, more tedious than anything. I have tons of free time so am willing to have a go at it.
Indeed, with you guys being able to do this leaves me and the other devs to work on optimizing other areas of the mod that requite a bigger skill set, like optimizing the models themselves :)

ZAP44 wrote:I?m at the point of having unzipped the objects and was trying to interact with the object like making it move back to the point it disappears. No luck. I can make it do everything but that. Is there a listing of key commands for use with the editor somewhere? The help supplied with the editor is missing most of its content.
Hummm I dont think there is a tutorial on this anywhere that I can think of, if anyone knows of one let me know was soo long ago since when I first tried using the editor I forgot what I learnt the basic commands with hehe.

But basically to interact with your view in the BF2 editor, what you need to do is hover your mouse over the 3D centre area and then press the space bar, or hold down the middle mouse button and then you can use your WASD keys to move about and use your mouse to look in a direction to look ingame. To exit movement mode, if you pressed space bar either press the "esc" key or space bar again and if you gone into by holding down the MMB, just let go of it and you will be back into normal mouse control :)

Any other questions let me know :)

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-14 04:06
by ZAP44
WASD keys...why didn't I think of that?? :smile: It still would be nice to have a comprehensive list of those key commands though if anyone has them. I see a post someone made in 2005 complaining about the unfinished BF2 Editor manual...you have to love EA's unfinished work. I'll start with Air objects beings Vaiper is working with land. I'm sure there will be questions. I'll try not to make it a habit. I've seen posts of ppl offering to help...here's a good opportunity.

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-14 16:21
by karambaitos
ill be sure to help out as soon as i read the tuts :)

also do the lods on the kit geometry not work as well???

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-14 19:12
by Psyrus

Code: Select all

[b][u]civ_jep_technical[/u][/b]

ObjectTemplate.cullRadiusScale 9.7

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 135
GeometryTemplate.setSubGeometryLodDistance 2 0 40
GeometryTemplate.setSubGeometryLodDistance 2 1 70

[b]Issues:[/b][list=1]
[*]Technical LOD3 front texture is plain white, which leads to it being quite hard to inconspicuously LOD switch from the front. If a texturer could place a bit of black smudging where the headlights are ([url=http://i235.photobucket.com/albums/ee94/psyrus_uraya/techie_bad_lod3_tex.jpg]see picture for difference[/url]) you could LOD switch earlier, at 100-110m imo.
[*]The stark colour difference on the right hand side of the techie wreck LOD 0vs1 means that switching has to be pushed out, whereas I would've gone with 30m otherwise. Even @ 40m the difference is reasonably noticeable imo. 
[*]It is also my humble opinion that the LOD3 wheels could be even more simple without much/any aesthetic loss. 
[*]Between 800-1000m the model switches between 1-3 pixels on my screen, but since you approved the cull distance of ~10 for the dirtbike, and the techie is bigger, I figured I shouldn't go less than 9.[/list]
[b]Coding:[/b] 
[b]Checked by:[/b] 
[b]Tweaks:[/b] 
[b]Committed to REPO:[/b]
Those are my proposed changes. I'm not sure if I really "get it", nor do I have a particularly good eye for these things... but since about '08 I've been trying to figure out if I can contribute in some way and this seems like something menial enough that it'd suit me :)

Edit: The darker coloured technicals (red, black, blue) can have an earlier LOD switch from 2->3 due to the front headlight issue being less obvious. The only issue with this is the windscreen, which is blacked out in LOD3. I would suggest the texture be lightened on LOD3 around the window parts so that the black->transparent switch is less obvious (although barely noticeable @ 135m imo).

If I got it reasonably right I'll knock a couple of these out per day when I'm not studying until I get sick of it :)

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-15 02:41
by ZAP44
Reviewed your sample LAV and the two objects done by Vaiper. As pointed by Psyrus I may not have an eye for this stuff either. You can view a LOD and fly out till you think the switching distance is perfect then turn the object a bit and something appears and you have to move out some more till that disappears. Anyway this is what I came up with for the Chinook. Cull of 7 is 1000.69 meters

Code: Select all

[b]cf_the_chinook, ch_the_chinook, us_the_chinook
[/b]

ObjectTemplate.cullRadiusScale 7

GeometryTemplate.setSubGeometryLodDistance 1 0 134
GeometryTemplate.setSubGeometryLodDistance 1 1 179
GeometryTemplate.setSubGeometryLodDistance 1 2 202
GeometryTemplate.setSubGeometryLodDistance 1 3 458
GeometryTemplate.setSubGeometryLodDistance 2 0 109
GeometryTemplate.setSubGeometryLodDistance 2 1 157 
GeometryTemplate.setSubGeometryLodDistance 2 2 275



Issues: Final Tri 3P is 467, wreck is 220


Coding: 
Checked by: 
Committed to REPO:
Be interesting to see how badly I've done here...seabag is packed for re-training command :(

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-15 05:12
by Rhino
Psyrus wrote:

Code: Select all

[b][u]civ_jep_technical, civ_jep_technical_black, civ_jep_technical_blue, civ_jep_technical_militia, civ_jep_technical_red, civ_tow_technical, civ_tow_technical_militia[/u][/b]

ObjectTemplate.cullRadiusScale 9.7

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 135
GeometryTemplate.setSubGeometryLodDistance 2 0 40
GeometryTemplate.setSubGeometryLodDistance 2 1 70

[b]Issues:[/b][list=1]
[*]Technical LOD3 front texture is plain white, which leads to it being quite hard to inconspicuously LOD switch from the front. If a texturer could place a bit of black smudging where the headlights are ([url=http://i235.photobucket.com/albums/ee94/psyrus_uraya/techie_bad_lod3_tex.jpg]see picture for difference[/url]) you could LOD switch earlier, at 100-110m imo.
[*]The stark colour difference on the right hand side of the techie wreck LOD 0vs1 means that switching has to be pushed out, whereas I would've gone with 30m otherwise. Even @ 40m the difference is reasonably noticeable imo. 
[*]It is also my humble opinion that the LOD3 wheels could be even more simple without much/any aesthetic loss. 
[*]Between 800-1000m the model switches between 1-3 pixels on my screen, but since you approved the cull distance of ~10 for the dirtbike, and the techie is bigger, I figured I shouldn't go less than 9.[/list]
[b]Coding:[/b] 
[b]Checked by:[/b] 
[b]Tweaks:[/b] 
[b]Committed to REPO:[/b]
Those are my proposed changes. I'm not sure if I really "get it", nor do I have a particularly good eye for these things... but since about '08 I've been trying to figure out if I can contribute in some way and this seems like something menial enough that it'd suit me :)

Edit: The darker coloured technicals (red, black, blue) can have an earlier LOD switch from 2->3 due to the front headlight issue being less obvious. The only issue with this is the windscreen, which is blacked out in LOD3. I would suggest the texture be lightened on LOD3 around the window parts so that the black->transparent switch is less obvious (although barely noticeable @ 135m imo).

If I got it reasonably right I'll knock a couple of these out per day when I'm not studying until I get sick of it :)
That looks all good to me, how ever when stating issues, can you state at the very start what geom and lod the issue is with, ie "1. Geom_1, LOD2: bla bla" :)
Also you missed out quite a critical error with the wrecks lods, with the wrecks final lod being over 1k tris which is waaayy over the limit :p

Also if future can you put your code in "[ code ]" tags as in the template I've edited your post to put the code in code tags, it just makes it easier for me if its in them :)

Also when there are multiples of the same vehicle that all use the same mesh, if you can include all of the names of all the vehicles in the title it would be good like I've done in your edit.

Other than that, looks good, nice work :D

ZAP44 wrote:Reviewed your sample LAV and the two objects done by Vaiper. As pointed by Psyrus I may not have an eye for this stuff either. You can view a LOD and fly out till you think the switching distance is perfect then turn the object a bit and something appears and you have to move out some more till that disappears. Anyway this is what I came up with for the Chinook. Cull of 7 is 1000.69 meters

Code: Select all

[b]cf_the_chinook, ch_the_chinook, us_the_chinook[/b]

ObjectTemplate.cullRadiusScale 7

GeometryTemplate.setSubGeometryLodDistance 1 0 134
GeometryTemplate.setSubGeometryLodDistance 1 1 179
GeometryTemplate.setSubGeometryLodDistance 1 2 202
GeometryTemplate.setSubGeometryLodDistance 1 3 458
GeometryTemplate.setSubGeometryLodDistance 2 0 109
GeometryTemplate.setSubGeometryLodDistance 2 1 157 
GeometryTemplate.setSubGeometryLodDistance 2 2 275



Issues: Final Tri 3P is 467, wreck is 220


[b]Coding:[/b] 
[b]Checked by:[/b] 
[b]Committed to REPO:[/b]
Be interesting to see how badly I've done here...seabag is packed for re-training command :(
Well first of all, a cullradiusscale of 7 is way too high for this, 5 seems a better value where it draws for 2kms with that and only a few pixels big before it disappears.

As for your LOD distances, they are totally off...

These are some better values:

Code: Select all

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 125
GeometryTemplate.setSubGeometryLodDistance 1 3 250
GeometryTemplate.setSubGeometryLodDistance 2 0 75
GeometryTemplate.setSubGeometryLodDistance 2 1 125
GeometryTemplate.setSubGeometryLodDistance 2 2 250

Also like Psyrus, if future can you put your code in "[ code ]" tags as in the template I've edited your post to put the code in code tags, it just makes it easier for me if its in them and when there are multiples of the same vehicle that all use the same mesh, if you can include all of the names of all the vehicles in the title it would be good like I've done in your edit. :)

Cheers :)

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-15 05:53
by Acemantura
Oh GOD, my brand spanking new laptop cant play a damn thing anymore because it is stuck at 15 fps.

I'll do at least one of these bad boys after finals.

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-15 07:09
by SnipeHunt
I seemed to have started with an easy one. I made just some minor changes to LOD and cull. If I am on the right track I'll continue the work.

Code: Select all

[b][u]boat_rib[/u][/b] 

ObjectTemplate.cullRadiusScale 4.5  

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 100
GeometryTemplate.setSubGeometryLodDistance 2 0 25
GeometryTemplate.setSubGeometryLodDistance 2 1 90  

[b]Issues:[/b][list=1]Final Tri 3P is 183, wreck is 152
[*]Final LOD 3P and wreck has .50cal showing as a floating gun 
removal should lower tris without harm to model[*][*][*][*][*][*][*][/list] 
[b]Coding:Snipe[/b]  
[b]Checked by:[/b]  
[b]Tweaks:[/b]  
[b]Committed to REPO:[/b]

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-15 07:45
by Psyrus
[quote=""'[R-DEV"]Rhino;1512140']That looks all good to me, how ever when stating issues, can you state at the very start what geom and lod the issue is with, ie "1. Geom_1, LOD2: bla bla" :)
Also you missed out quite a critical error with the wrecks lods, with the wrecks final lod being over 1k tris which is waaayy over the limit :p

Also if future can you put your code in "[ code ]" tags as in the template I've edited your post to put the code in code tags, it just makes it easier for me if its in them :) [/quote]

Thanks for the feedback! All taken on board, although it seems when you added it to your first post you forgot to list the additional issue "wreck LODs poly count too high (>1000)" :)

Will get stuck into another one/two later tonight if time permits[quote="acemantura""]Oh GOD, my brand spanking new laptop cant play a damn thing anymore because it is stuck at 15 fps.

I'll do at least one of these bad boys after finals.[/quote]Well historically unless you're paying in excess of 3k for a laptop, it probably won't be a gaming beast, and further to that point 'when' you buy a computer has very little bearing on gaming performance, like how you can buy a 2011 corolla and can't be miffed that it doesn't even have 1991 ferrari performance ;)

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-15 08:58
by Rhino
SnipeHunt wrote:I seemed to have started with an easy one. I made just some minor changes to LOD and cull. If I am on the right track I'll continue the work.

Code: Select all

[b][u]boat_rib[/u][/b] 

ObjectTemplate.cullRadiusScale 4.5  

GeometryTemplate.setSubGeometryLodDistance 1 0 30
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 100
GeometryTemplate.setSubGeometryLodDistance 2 0 25
GeometryTemplate.setSubGeometryLodDistance 2 1 90  

[b]Issues:[/b][list=1]Final Tri 3P is 183, wreck is 152
[*]Final LOD 3P and wreck has .50cal showing as a floating gun 
removal should lower tris without harm to model[*][*][*][*][*][*][*][/list] 
[b]Coding: SnipeHunt[/b]  
[b]Checked by:[/b]  
[b]Tweaks:[/b]  
[b]Committed to REPO:[/b]
well tbh you have started with a model that already had pretty good cull and lod distances hehe :)

But your cull distance is a slight improvement over the old and your wreck lod2 distance is better than the old too. Other than that thou, the old distances where better but you where not far off them :p

Here's what I found worked well for the LOD distances, so ye only geom0, lod1 distance did I change down a bit from yours, other others where either the old settings still or your tweaks :)
GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 50
GeometryTemplate.setSubGeometryLodDistance 1 2 100
GeometryTemplate.setSubGeometryLodDistance 2 0 25
GeometryTemplate.setSubGeometryLodDistance 2 1 90

as for the issues, the final tri counts are fine, no need to note them, providing they are under 300tris its not worth changing/noting :)
As for the gun, that's fine its a separate mesh and works on its own system :)

Cheers :)
Psyrus wrote:Thanks for the feedback! All taken on board, although it seems when you added it to your first post you forgot to list the additional issue "wreck LODs poly count too high (>1000)" :)
hehe cheers, forgot about that one :)

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-15 15:52
by ZAP44
Ouch! OK...well not ready to quit just yet.

Code: Select all

 tags...my bad I knew it and didn't do it.  And multiple names I was wondering about.

I'm not moving on until I understand my errors with the Chinook.  Cull I had as 7 and you say 5 and go on to say it draws out to 2kms.  Agree...don't know why 7 seemed right last night.  However how do you get 2kms?  At 5 I'm getting only a couple pixels and the Object Info Tool Window gives me the Distance as 1002.37?  

Next the LODS.  Felt the values were way too high and think my assumptions on when to switch LODs is probably wrong.  A couple examples of how I came up with those values.  LOD 0 to LOD 1 you loose visibility of the hoist on the right side.  Setting the value to 25 causes this to disappear too early I thought as it stays visible out to 134.  This was the same reasoning I used for all the values.  LOD 3 a few pixels of the left wheel are still visible out to 458.  Think maybe my problem is I'm being too nitpicky.

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-16 01:23
by ZAP44
Had time so tried another. ch_ahe_wz11 also has ch_ahe_wz11_sp not sure I should have listed both. This model goes to only a few pixels at 1000 with a cull of 10. Big difference in what I'd establish LOD values at for this model compared to Chinook as it doesn't have things like hoists and wheels protruding that show much change between LODs.

Code: Select all

[b]ch_ahe_wz11, ch_ahe_wz11_sp[/b]

ObjectTemplate.cullRadiusScale 10

GeometryTemplate.setSubGeometryLodDistance 1 0 20
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 100
GeometryTemplate.setSubGeometryLodDistance 2 0 20
GeometryTemplate.setSubGeometryLodDistance 2 1 80


[b]Issues:[/b]
          1. Geom_1 LOD3 Final Tri count above the limit at 722 Tris.
          2. Geom_2 LOD2 Final Tri count above the limit at 719 Tris.
          3. Geom_2, LOD1 and 2, Rear rotor hanging in mid-air from missing after part of wreck.
          

[b]Coding:
Checked by:
Tweaked by:
Committed to REPO: [/b]

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-16 03:32
by Rhino
ZAP44 wrote:Had time so tried another. ch_ahe_wz11 also has ch_ahe_wz11_sp not sure I should have listed both. This model goes to only a few pixels at 1000 with a cull of 10. Big difference in what I'd establish LOD values at for this model compared to Chinook as it doesn't have things like hoists and wheels protruding that show much change between LODs.

Code: Select all

[b]ch_ahe_wz11, ch_ahe_wz11_sp[/b]

ObjectTemplate.cullRadiusScale 10

GeometryTemplate.setSubGeometryLodDistance 1 0 20
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 100
GeometryTemplate.setSubGeometryLodDistance 2 0 20
GeometryTemplate.setSubGeometryLodDistance 2 1 80


[b]Issues:[/b]
          1. Geom_1 LOD3 Final Tri count above the limit at 722 Tris.
          2. Geom_2 LOD2 Final Tri count above the limit at 719 Tris.
          3. Geom_2, LOD1 and 2, Rear rotor hanging in mid-air from missing after part of wreck.
          

[b]Coding:
Checked by:
Tweaked by:
Committed to REPO: [/b]
Right first of all, your cull distance is still way too high. A tiny scout chopper doesn't need to draw fro 2kms :p
A cullRadiusScale of 3.5 is much more appropriate with having it draw upto 800m.

As for your lod distances, they are much better although I extended the distance on geom0, lod3 to 150m as the blacking out of the windows was a little bit too obvious at 100m.

GeometryTemplate.setSubGeometryLodDistance 1 0 20
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 20
GeometryTemplate.setSubGeometryLodDistance 2 1 80

Issues 1 and 2 are good catches but issue 3 isn't an issue since when the vehicle is destroyed the rotors are removed as they are separate objects. If you want to see what the wreck is like once the vehicle has been properly destroyed, switch to Ammunition mode in the Object editor (button in the top with a bullet going though a wall icon) and then in the Editor bar on the right, scroll all the way down to the bottom and you should see an explode button, hit that and your vehicle will explode and you should see the wreck and the rotors and other child objects should now be removed :)
http://img63.imageshack.us/img63/8920/1 ... 040340.jpg
http://img824.imageshack.us/img824/1927 ... 040526.jpg
http://img573.imageshack.us/img573/6731 ... 040559.jpg

Also if you can list your issues inside the "[ List ]" tags it as per the template would be good :)

Cheers! :D

These comments where in relation to the WZ551 APC, not the WZ-11 Scout chopper :p

Right for starters its looking better, thou the cull radius scale I would recommend to be 8 with it drawing for 1.2km

Your LOD distances where better this time, some where just right but most needed a little tweaking. Here is what I found worked better :)

GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 75
GeometryTemplate.setSubGeometryLodDistance 1 2 100
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 100

Ummm also on your listed issues, where are you getting your tri/poly count from? Your listed tri counts are incorrect, Geom_1, LOD3's tri/poly count is 271 tris and Geom_2, LOD2's is 157 tris?

You should be getting your tri count off the object tool bar at the bottom where it says "Poly Count:", the very bottom arrow in this pic here: http://realitymodfiles.com/rhino/tuts/l ... nces13.jpg

Issue 3 is an issue thou, but not a big one but still worth listing.

Also when listing issues please use the [ List ] tags as provided in the template :)

Cheers! :D |

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-16 08:03
by Psyrus

Code: Select all

[b][u]civ_bik_hondacb500[/u][/b]

ObjectTemplate.cullRadiusScale 14 [color=Red]or 8.8*[/color]

GeometryTemplate.setSubGeometryLodDistance 1 0 25
GeometryTemplate.setSubGeometryLodDistance 1 1 60
GeometryTemplate.setSubGeometryLodDistance 1 2 90
GeometryTemplate.setSubGeometryLodDistance 1 3 110
GeometryTemplate.setSubGeometryLodDistance 2 0 20
GeometryTemplate.setSubGeometryLodDistance 2 1 55
GeometryTemplate.setSubGeometryLodDistance 2 2 100
GeometryTemplate.setSubGeometryLodDistance 2 3 110


[b]Issues:[/b]
[list=1]
[*]Geom_1 & Geom_2, LOD_3 [b]&[/b] LOD_4 Tri/poly counts too high. LOD_3 [1168] should be halved to ~500 and LOD4 [769] can be easily reduced to ~250. The fuel tank and detailed seat can probably be removed at these ranges and replaced by rectangles to fill the space (so that it's not transparent and thus an obvious LOD switch.
[*]Geom_1 & Geom_2, LOD_2->LOD_3: ([url=http://i235.photobucket.com/albums/ee94/psyrus_uraya/LODs/civ_honda_spokes.jpg]Image[/url]) The contrast in the 'rim' or 'spokes' of the wheel going from pure black to transparent at almost any range is very obvious. Would suggest a simple black object (rectangle perhaps) to fill the space so that the LOD switch can be pulled in sooner. 
[*]Geom_1 & Geom_2: As it stands there is very little difference between LOD_3 and LOD_4, and I think LOD_4 could be heavily optimized and the [i]LOD 3->4 switch pushed out further to ~150m[/i]
[*][[i]Not sure if issue[/i]] Geom_1 & Geom_2, LOD_3 & LOD_4 : Tail light on both geom_1 and wreck (geom_2) is floating in mid air. Would not be visible at the LOD range (90 & 100m respectively) in my opinion. 
[*]
[*]
[*]
[*]
[/list]

[b]Coding:[/b] Psyrus
[b]Checked by:[/b] 
[b]Tweaked by:[/b] 
[b]Committed to REPO:[/b]

* The reason I list 2 possible cull distances is because I'm not entirely sure what I'm going for here. On the LAV3 tutorial it seems to be when the LAV gets so small (like 1 pixel), that's when we cull it... but subsequently I've been reading you'd like to be able to accommodate ~1km+ view distances, so does this mean you'd like us to set the cull distances on even small objects close to 1km? From ~450m distance the bike flickers between 0-3 pixels, so the cull distance of 8.8 renders it out to 500m with 1 pixel, or the cull distance of 14 renders it to ~800m with 1 pixel... not sure which is "better"? :confused: :confused:

This one was definitely harder than the techie due to the transparent wheels on the distant LODs, my values might need to be pushed out a little if you think the switch is too obvious @ 90/100m (although tbh I think it looked obvious all the way out to ~175m and I've never seen the bike ingame so I can't really visualize it)

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-16 08:03
by SnipeHunt
Got another easy one :)

Code: Select all

[b][u]cf_jep_gwagon[/u] [u]cf_jep_gwagon_support[/u] [u]ger_jep_gwagon_support[/u] [u]ger_jep_support_alt[/u][/b] 

ObjectTemplate.cullRadiusScale 6.1  

GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 100
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 30
GeometryTemplate.setSubGeometryLodDistance 0 1 100  

[b]Issues:[/b] 
[list=1] cf_jep_gwagon
[*] geom_1 LOD3 polys 668
[/list] 

[b]Issues:[/b] 
[list=1] cf_jep_gwagon_support
[*] geom_2 LOD2 license plate visable in the right front wheel well
[/list]

[b]Issues:[/b] 
[list=1] ger_jep_gwagon_support
[*] geom_2 LOD2 license plate visable in the right front wheel well
[/list]

[b]Issues:[/b] 
[list=1] ger_jep_gwagon_support_alt
[*] geom 1 LOD3 poly count 1156
[*] geom 2 LOD1 bit of floating texture in front of left headlight
[*] geom 2 LOD2 license plate visable in the right front wheel well
[/list]

[b]Coding:[/b] *SnipeHunt* 
[b]Checked by:[/b]  
[b]Tweaked by:[/b]  
[b]Committed to REPO:[/b]

Code: Select all

[b][u]cf_tnk_leo2a6[/u] [u]ger_tnk_leo2a6[/u][/b] 

ObjectTemplate.cullRadiusScale 3.8  
  
GeometryTemplate.setSubGeometryLodDistance 1 0 50
GeometryTemplate.setSubGeometryLodDistance 1 1 100
GeometryTemplate.setSubGeometryLodDistance 1 2 150

[b]Issues:[/b] 
[list=1] cf_tnk_leo2a6
[*] geom_2 LOD0 poly count 6279
[*] geom_2 has no LOD1 or LOD2
[/list] 

[b]Issues:[/b] 
[list=1] ger_tnk_leo2a6
[*] geom_2 LOD0 poly count 6279
[*] geom_2 has no LOD1 or LOD2
[/list] 


[b]Coding:[/b] *SnipeHunt* 
[b]Checked by:[/b]  
[b]Tweaked by:[/b]  
[b]Committed to REPO:[/b]

Code: Select all

[b][u]cf_trk_logistics[/u] [u]idf_trk_logistics[/u][/b] 

ObjectTemplate.cullRadiusScale  2.9

GeometryTemplate.setSubGeometryLodDistance 1 0 50   
GeometryTemplate.setSubGeometryLodDistance 1 1 100
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 100


[b]Issues:[/b] 
[list=1]both models 
[*] geom_1 LOD3 poly count 1322
[*] geom_2 LOD2 poly count 1162
[*] The cull radius seems low at 2.9, but draws to a fine pixel at 1000m
 I am guessing the large poly count causes that?
[/list] 

[b]Coding:[/b] *SnipeHunt* 
[b]Checked by:[/b]  
[b]Tweaked by:[/b]  
[b]Committed to REPO:[/b]

Code: Select all

[b][u]ch_trk_logistics[/u] [u]mil_trk_logistics[/u] [u]ru_trk_logistics[/u] [u]us_trk_logistics[/u][/b] 

ObjectTemplate.cullRadiusScale  2.9

GeometryTemplate.setSubGeometryLodDistance 1 0 50   
GeometryTemplate.setSubGeometryLodDistance 1 1 100
GeometryTemplate.setSubGeometryLodDistance 1 2 150
GeometryTemplate.setSubGeometryLodDistance 2 0 50
GeometryTemplate.setSubGeometryLodDistance 2 1 100

[b]Issues:[/b] 
[list=1] all 4 models
[*] geom_1 LOD3 poly count 1078
[*] geom_1 LOD2 poly count 1037
[*] the rear mudflaps are bright like a light is shining on the side by the tire. 
It should be shaded since it is under the truck
[*] The cull radius seems low at 2.9, but draws to a fine pixel at 1000m
I am guessing the large poly count causes that?

[/list] 

[b]Coding:[/b] *SnipeHunt* 
[b]Checked by:[/b]  
[b]Tweaked by:[/b]  
[b]Committed to REPO:[/b]

Re: Help us to Optimize PR! (LOD Switching & Cull Distances)

Posted: 2010-12-16 08:23
by Amok@ndy
its CF for Canadian Forces and not CEF :) :razz:
All times are UTC
Page 1 of 6

Powered by phpBB® Forum Software © phpBB Limited