Fullness, what is this?

Discussion in 'Mapping Questions & Discussion' started by Riever, Dec 8, 2011.

  1. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    Hi,
    Could someone explain, or point me at an explanation on fullness?

    I am about 3/4 of the way through half of a symmetrical map and am now adding the lighting. All is OK (ish), however I have noticed that the LDR fullness is now at 60%. When I flip the map for the red side I'll end up with 120%, or more.

    Now there are a lot of displacements in the map (101 so far) as it's in a chasm, could this be the reason, or am I stuffing up on the lighting some how.

    Does it actually matter? Can't find anything about this other than people with 'very full' maps asking for help.

    Thanks in advance.
     
  2. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    Hmmm,

    Well I now have the map compiled, but am no nearer an answer as to Fullness. after mirroring the blue side to make the red (symmetrical map) I currently have 3 items that are listed 'very full'.

    What I cannot understand is why LDR leaf ambient is 152%, but the LDR light data is 0% and the LDR ambient table is only 30%.

    Also I only have 1 water brush, so why on Earth are the indices 98%?

    The map does run, but the performance on my 4 core PC is poor - so it'd tank on a single core one.

    Could someone please explain, or point me at a tutorial that shows how to reduce the LDR and water indices. I guess the only way to reduce the vertexes is to delete stuff from the map (?).

    Below is an edited compile list.

    Object names Objects/Maxobjs Memory / Maxmem Fullness
    ------------ --------------- --------------- --------
    models 153/1024 7344/49152 (14.9%)
    brushes 4236/8192 50832/98304 (51.7%)
    brushsides 34772/65536 278176/524288 (53.1%)
    planes 26714/65536 534280/1310720 (40.8%)
    vertexes 55484/65536 665808/786432 (84.7%) VERY FULL!
    ...
    faces 28791/65536 1612296/3670016 (43.9%)
    ...
    leaffaces 38143/65536 76286/131072 (58.2%)
    leafbrushes 15152/65536 30304/131072 (23.1%)
    areas 12/256 96/2048 ( 4.7%)
    surfedges 202063/512000 808252/2048000 (39.5%)
    edges 121016/256000 484064/1024000 (47.3%)
    LDR worldlights 261/8192 22968/720896 ( 3.2%)
    HDR worldlights 0/8192 0/720896 ( 0.0%)
    leafwaterdata 4/32768 48/393216 ( 0.0%)
    waterstrips 3877/32768 38770/327680 (11.8%)
    waterverts 0/65536 0/786432 ( 0.0%)
    waterindices 64776/65536 129552/131072 (98.8%) VERY FULL!
    ..
    LDR lightdata [variable] 26618432/0 ( 0.0%)
    HDR lightdata [variable] 0/0 ( 0.0%)
    visdata [variable] 0/16777216 ( 0.0%)
    entdata [variable] 249896/393216 (63.6%)
    LDR ambient table 19857/65536 79428/262144 (30.3%)
    HDR ambient table 19857/65536 79428/262144 (30.3%)
    LDR leaf ambient 99781/65536 2793868/1835008 (152.3%) VERY FULL!

    HDR leaf ambient 19857/65536 555996/1835008 (30.3%)
    ...
    physics [variable] 1684947/4194304 (40.2%)
    physics terrain [variable] 73486/1048576 ( 7.0%)
     
  3. Ravidge

    aa Ravidge Grand Vizier

    Messages:
    1,544
    Positive Ratings:
    2,495
    I don't know how you manage to get such values on Vertexes and LDR leaf ambient. It seems unnaturally high to me. Just for comparison I went and checked my compile log for the 3 stage plr map I made, fully detailed.

    While I get close to the limits for brushes, I'm nowhere near the vertex limit, which makes me believe you have some seriously complex brushwork? But your brushsides are relatively low, so I just don't know how that would work.
    models 168/1024 8064/49152 (16.4%)
    brushes 7852/8192 94224/98304 (95.8%) VERY FULL!
    brushsides 58650/65536 469200/524288 (89.5%) VERY FULL!
    planes 25094/65536 501880/1310720 (38.3%)
    vertexes 41328/65536 495936/786432 (63.1%)

    For LDR leaf ambient, I get around 60%
    And I honestly don't know what that means. And how you'd lower it.

    waterindices has nothing to do with water. It's a little more complicated, and there are threads for it on this forum if you search. Basically it's about how source engine will cut up brushes based on how they touch each other, and what entity types the brushes are. This value is the one most common to reach the limit on any detailed map.
     
  4. grazr

    aa grazr Old Man Mutant Ninja Turtle

    Messages:
    5,436
    Positive Ratings:
    3,573
    Perhaps he's using power 4 displacements and has a bunch of unseen displacement faces rendered.

    This is very curious. Even fully finished maps usually hover around the 50-60% margin. Have you used the carve tool?
     
  5. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    Hi,

    Many thanks for the reply, VRAD is producing a massive file currently so I'm guessing I have goofed somewhere in the lighting I have so far. Sadly I don't know where so it's very trial and error (or at present error and error). Could it be as I have some dense shadows?

    I've been compiling with fastvis on so am currently doing a test to see if it is even possible to compile with normal vis in a sensible time, just in case an optimised vis reduces these.

    I have a feeling not, VVis is taking a 1gb of ram at present and hasn't even reached 1. :blushing:
     
  6. Wander

    Wander L3: Member

    Messages:
    148
    Positive Ratings:
    42
    You can try using the cordon tool: :cordonenable:
    To compile parts of the map, and that way narrow down the problem
     
  7. re1wind

    aa re1wind

    Messages:
    644
    Positive Ratings:
    382
    For starters, linking to your complete compile log would help in helping to determine where the problem(s) is(are) (i.e. vbsp error/warning lines, or a redonculously long vvis compile time)

    judging by the above, it looks like you have too many leaves in your map, which exponentially increase vvis compile time, and can also explain the large amounts of waterindecies, as world geometry (aka brushes that are not tied to func_detail, or any other brush-based entity)will act as a solid for vvis, hence increasing the number of leaves.


    Edit: looking at the ratio of brushes to vertices:
    4236 brushes, 55484 vertices.

    thats an average of 13 vertices per brush, a wedge has 6, a cube has 8, an 'octagon' has 16. As most brushes will tend to be wedges or cubes, at least for world geometry (and most func_details ) i'd say part of your problem is a large amount of cylinder brushes.

    even if you func_detail them and snap them into the surface of a world geometry wall, it will generate a lot of vertices AND waterindices.
     
  8. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    Many thanks for the replies.

    I've checked and there are no 4 power displacements. That said displacements don't block vis and I have found large areas of 'map' which contain visleaves but which the players will never (or should never) see. This is due to the tunnels in the mine section of the map (basically 4 brushes as a square tube) and the fact that the vis leaves are generated on every brush(?) so there are vis leaves above and below the tunnel.

    I guess I can use the viscluster brush to speed the compile in these areas, but would this reduce the vis data, or increase it?

    Could I just extend the floors of these tunnels into the void to kill the visleaves?

    On the subject of hex/octagons etc, oops. I have columns which I have now switched to models.

    I did use carve, till I found what a complete pile of dud it is. Rebuilt those sections. Could it be that there is rubbish left over? If so how do I find it?

    Here's a complete compile - minus the 'Im running this stuff' (incidentally I have seen posts where logs have scoll bars in them - how can I do this?)


    ** Executing...
    ** Command:
    Valve Software - vbsp.exe (Oct 25 2011)
    4 threads

    Patching WVT material: maps/testmap212gr28pp28tfd/nature/blendcliffgrass001a_wvt_patch
    Patching WVT material: maps/testmap212gr28pp28tfd/nature/blendrockgroundwall003_wvt_patch
    Patching WVT material: maps/testmap212gr28pp28tfd/nature/blendrockground007_wvt_patch
    Patching WVT material: maps/testmap212gr28pp28tfd/nature/blendrockgroundwall001_wvt_patch
    Patching WVT material: maps/testmap212gr28pp28tfd/nature/blendrockgroundwall004_wvt_patch
    Patching WVT material: maps/testmap212gr28pp28tfd/brick/blendcobbletocobblesnow001_wvt_patch
    Patching WVT material: maps/testmap212gr28pp28tfd/nature/blendrockground001_wvt_patch
    Patching WVT material: maps/testmap212gr28pp28tfd/nature/blendrockground002_wvt_patch
    Patching WVT material: maps/testmap212gr28pp28tfd/cp_mountainlab/nature/blendrocktograss002_wvt_patch
    fixing up env_cubemap materials on brush sides...
    ProcessBlock_Thread: 0...1...2...3...4...5...6...7...8...9...10 (2)
    ProcessBlock_Thread: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Processing areas...done (0)
    Building Faces...done (1)
    Chop Details...done (0)
    Find Visible Detail Sides...
    Merged 1490 detail faces...done (0)
    Merging details...done (0)
    FixTjuncs...
    PruneNodes...
    WriteBSP...
    done (3)
    writing
    Building visibility clusters...
    done (1)
    *** Error: Skybox vtf files for skybox/sky_well_01 weren't compiled with the same size texture and/or same flags!
    Can't load skybox file skybox/sky_well_01 to build the default cubemap!
    *** Error: Skybox vtf files for skybox/sky_well_01 weren't compiled with the same size texture and/or same flags!
    Can't load skybox file skybox/sky_well_01 to build the default cubemap!
    Finding displacement neighbors...
    Finding lightmap sample positions...
    Displacement Alpha : 0...1...2...3...4...5...6...7...8...9...10
    Building Physics collision data...
    done (1) (1575181 bytes)
    Placing detail props : 0...1...2...3...4...5...6...7...8...9...10
    Compacting texture/material tables...
    Reduced 10360 texinfos to 5650
    Reduced 164 texdatas to 148 (4094 bytes to 3482)
    Writing 19 seconds elapsed

    ** Executing...
    ** Command:
    Valve Software - vvis.exe (Oct 25 2011)
    fastvis = true
    4 threads
    reading
    7066 portalclusters
    23782 numportals
    BasePortalVis: 0...1...2...3...4...5...6...7...8...9...10 (46)
    Optimized: 2899940 visible clusters (0.00%)
    Total clusters visible: 39437293
    Average clusters visible: 5581
    Building PAS...
    Average clusters audible: 6968
    visdatasize:12004636 compressed from 12549216
    writing
    1 minute, 1 second elapsed

    ** Executing...
    ** Command:
    Valve Software - vrad.exe SSE (Oct 25 2011)

    Valve Radiosity Simulator
    4 threads
    [Reading texlights from 'lights.rad']
    [34 texlights parsed from 'lights.rad']

    Loading
    Setting up ray-trace acceleration structure... Done (6.24 seconds)
    24299 faces
    25 degenerate faces
    3358097 square feet [483566080.00 square inches]
    206 Displacements
    751650 Square Feet [108237600.00 Square Inches]
    24274 patches before subdivision
    zero area child patch
    zero area child patch
    zero area child patch
    zero area child patch
    330136 patches after subdivision
    sun extent from map=0.008727
    261 direct lights
    BuildFacelights: 0...1...2...3...4...5...6...7...8...9...10 (94)
    BuildVisLeafs: 0...1...2...3...4...5...6...7...8...9...10 (355)
    transfers 32267905, max 1936
    transfer lists: 246.2 megs
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #1 added RGB(1340285, 1038629, 758482)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #2 added RGB(271547, 183788, 115546)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #3 added RGB(69366, 41099, 22053)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #4 added RGB(20440, 10285, 4615)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #5 added RGB(6788, 2852, 1044)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #6 added RGB(2471, 853, 251)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #7 added RGB(969, 273, 63)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #8 added RGB(402, 92, 17)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #9 added RGB(174, 33, 5)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #10 added RGB(78, 12, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #11 added RGB(36, 4, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #12 added RGB(17, 2, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #13 added RGB(8, 1, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #14 added RGB(4, 0, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #15 added RGB(2, 0, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #16 added RGB(1, 0, 0)
    Build Patch/Sample Hash Table(s).....Done<0.1832 sec>
    FinalLightFace: 0...1...2...3...4...5...6...7...8...9...10 (32)
    FinalLightFace Done
    0 of 0 (0% of) surface lights went in leaf ambient cubes.
    ThreadComputeLeafAmbient: 0...1...2...3...4...5...6...7...8...9...10 (46)
    Writing leaf ambient...done
    Ready to Finish

    Object names Objects/Maxobjs Memory / Maxmem Fullness
    ------------ --------------- --------------- --------
    models 145/1024 6960/49152 (14.2%)
    brushes 4090/8192 49080/98304 (49.9%)
    brushsides 33114/65536 264912/524288 (50.5%)
    planes 24968/65536 499360/1310720 (38.1%)
    vertexes 46793/65536 561516/786432 (71.4%)
    nodes 15111/65536 483552/2097152 (23.1%)
    texinfos 5650/12288 406800/884736 (46.0%)
    texdata 148/2048 4736/65536 ( 7.2%)
    dispinfos 206/0 36256/0 ( 0.0%)
    disp_verts 16238/0 324760/0 ( 0.0%)
    disp_tris 25600/0 51200/0 ( 0.0%)
    disp_lmsamples 990712/0 990712/0 ( 0.0%)
    faces 24299/65536 1360744/3670016 (37.1%)
    hdr faces 0/65536 0/3670016 ( 0.0%)
    origfaces 13622/65536 762832/3670016 (20.8%)
    leaves 15257/65536 488224/2097152 (23.3%)
    leaffaces 32555/65536 65110/131072 (49.7%)
    leafbrushes 14322/65536 28644/131072 (21.9%)
    areas 9/256 72/2048 ( 3.5%)
    surfedges 176741/512000 706964/2048000 (34.5%)
    edges 106480/256000 425920/1024000 (41.6%)
    LDR worldlights 261/8192 22968/720896 ( 3.2%)
    HDR worldlights 0/8192 0/720896 ( 0.0%)
    leafwaterdata 2/32768 24/393216 ( 0.0%)
    waterstrips 3483/32768 34830/327680 (10.6%)
    waterverts 0/65536 0/786432 ( 0.0%)
    waterindices 58725/65536 117450/131072 (89.6%) VERY FULL!
    cubemapsamples 0/1024 0/16384 ( 0.0%)
    overlays 22/512 7744/180224 ( 4.3%)
    LDR lightdata [variable] 22303044/0 ( 0.0%)
    HDR lightdata [variable] 0/0 ( 0.0%)
    visdata [variable] 12004636/16777216 (71.6%)
    entdata [variable] 240953/393216 (61.3%)
    LDR ambient table 15257/65536 61028/262144 (23.3%)
    HDR ambient table 15257/65536 61028/262144 (23.3%)
    LDR leaf ambient 72364/65536 2026192/1835008 (110.4%) VERY FULL!
    HDR leaf ambient 15257/65536 427196/1835008 (23.3%)
    occluders 0/0 0/0 ( 0.0%)
    occluder polygons 0/0 0/0 ( 0.0%)
    occluder vert ind 0/0 0/0 ( 0.0%)
    detail props [variable] 1/12 ( 8.3%)
    static props [variable] 1/63190 ( 0.0%)
    pakfile [variable] 5961/0 ( 0.0%)
    physics [variable] 1575181/4194304 (37.6%)
    physics terrain [variable] 73528/1048576 ( 7.0%)

    Level flags = 0

    Total triangle count: 72862
    Writing
     
    Last edited: Dec 14, 2011
  9. re1wind

    aa re1wind

    Messages:
    644
    Positive Ratings:
    382
    edit 0:

    sounds fishy, but http://www.interlopers.net/index.php?page=errors says that it can be safely ignored. I'm inclined to disagree considering you have such a crazy ldr leaf ambient and visdata values.


    Edit2: Judging from this thread, it seems like you have a similar problem, leafvis, not entities. mat_leafvis and r_visambient seem to indicate how many light samples per leaf you have, which can help you find why you ldr leaf ambient value is so high.

    I'm beginning to think that some rogue lights and/or broken brushes are cause the leaf ambient problem, while the broken brushes (from the carve) are causing the high vertexes number along with waterindicies.

    edit3: after testing r_visambient a bit, and checking that you have a reasonable amount of leaves (check), in addition to your very high visdata value and high ldr leaf ambient, i'm not entirely sure whats causing this. It could be a very bright light near the area your carved, improper/broken lightmap on a brush (aka broken brush), a combination of these, a lot of func_details, or someting else.

    Right now I strongly suggest you use the above-mentioned solution of checking you map with the cordon tool, compiling small parts of your map at a time, multiple times with varying visibility groups (world geometry only, add func detail, add everything else) for your whole map. Tedious, but you've learnt not to use the carve tool, so consider this punishment for never using the carve tool again. :p
     
    • Thanks Thanks x 1
    Last edited: Dec 16, 2011
  10. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    Hi,

    Many thanks for the reply.
    I found dead space under several tunnels that were adding to the vis. In rebuilding to remove this I also managed to get rid of the Zero are child patches, and reduce the number of degenerate faces - not sure if the last is relevant.

    Anyway. My stats are more sensible.

    Unfortunately I have also managed to totally stuff something as none of the lights are now illuminating world brushes. They light models but not brushes. Result is that all the rooms are now pitch black.

    Also the sun has 'gone out' so the outside part of my map is pitch black as well.

    Since I haven't added or changed any light I'm baffled.

    There are no leaks but there are 100 lines referring to gather lights in the compile log. (like this -- GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #1 added RGB(1340285, 1038629, 758482) )

    I had 16, so what are these and could they be the reason for my lights suddenly not working?
     
    Last edited: Dec 17, 2011
  11. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    ???????????????!!!!!!!!!!!!!!!!!!!!!???????????????????????????

    OK Looking into the 'suddenly all textures are black in game, but models aren't' problem has shown that stone/stonewall2a is working! That is to say it is lit correctly. :blink:

    No other textures are! Vrad for some mad reason is not lighting any of them but it is rendering all the models correctly. :angry:

    If I load a previous map, from before I got this problem then it is fine all the textures are rendered correctly.

    Help?
     
    Last edited: Dec 17, 2011
  12. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    My apologies to anyone who has a black textures problem as the previous post and has stumbled upon this thread in the hope of a solution. I have none. I 'fixed' it by going back to a working version of the map and then redoing the bit that 'broke' it. I guess I did something slightly different but have no idea what as I no-longer have the black textures, or 100 lines of bounce lights in the compile.

    No idea if this will help anyone with a similar problem, but I found that lights that had been named still worked. All others became bugged and the textures around them black. Wish I knew why.
     
  13. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    Sigh....

    Typed too soon. The black texture bug is back. I just added a single light and bang every texture is black - unless it is lit by a bulb that is named.

    [​IMG]

    Anyone got a clue? There are no compile errors and no invalid solids. But I have the 100 gather lights again in the log ...
    ....

    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #98 added RGB(1, 1, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #99 added RGB(1, 1, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #100 added RGB(1, 1, 1)

    :confused:
     
  14. grazr

    aa grazr Old Man Mutant Ninja Turtle

    Messages:
    5,436
    Positive Ratings:
    3,573
    Do you have any texture lights?

    I encountered this once and it was because i had a large number of tex lights where the lightmaps were so small it caused the light count to go up crazy amounts. Each lightmap square is a luxel that emits light so if you have large blocks of texture lights, go through and increase the lightmap scale (from 4/8 or what ever to something higher like the default - 16) to reduce the amount of luxels that emit light. This'll bring your light count in line with engine limits and allow you to complete vrad.
     
    • Thanks Thanks x 1
  15. LeSwordfish

    aa LeSwordfish semi-trained quasi-professional

    Messages:
    4,113
    Positive Ratings:
    6,084
    You're naming your lights.

    You shouldn't name lights. Naming lights is pretty bad, since the compiler presumes named lights have to be turned on and off, and so vrad runs for both on and off for every texture they touch. If two named lights see a texture, it runs four times (OnOff, OnOn, OffOn, OffOff), if three see, it runs eight times (OnOffOn, OnOnOn, OffOnOn, OffOffOn, OnOffOff, OnOnOff, OffOnOff, OffOffOff) and so on, on an exponential scale. If six named lights are in a room, it runs thirty-two times what it has to for that room.

    If Vrad isn't finishing, that could be why. I'm no expert on the workings of the compiler, but i'd bet this is not helping. To unname all your lights at once, use View>Entity Report, then filter by class "light", select all of them, and open "Properties".
     
    • Thanks Thanks x 1
  16. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    Hi,

    Many thanks for the advice. I haven't adjusted the light maps yet so they are default 16 and I confess I don't know what a texlight is and so probably (?) haven't used them. I am lighting with light and light_spot.

    I think that Vrad does complete (or at least didn't say it didn't).

    The light names is a goof on my part - I got carried away naming things and am now un-naming them. There are 5 left to do. Though it is odd that the ones I did name are still working correctly while all the unnamed ones are failing.

    Curiously while looking for a solution I found a screenshot on the steam forums with the same problem. They had water in their 3d sky box and fixed the fault by removing the water.

    I don't have a 3d sky box but removing the water from my map has also 'fixed' the problem.

    What I don't yet know is if putting the water back, after I have unnamed the the lights, will bring the problem back. Will let you know.
     
    Last edited: Dec 22, 2011
  17. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    :mad:
    OK it looks like Vrad is having a fit about water on the map. Not, as far as I can tell any water. A small brush is ok. However a large one, like a river or lake results in the black texture bug. Without water the map runs fine, all textures lit correctly.

    So I am now officially stuck. I have no idea why adding one light has broken Vrad and I have no idea why removing the water 'fixes' the problem. I suspect that the water brush is not actually causing this but have no idea what could be.

    Edit 1.

    Hmmm. It's not the brush size, its where it is. A water brush on the surface works fine, all textures lit. One underground in a tunnel - zapp no lights. Now there are lights in this tunnel. Am going to remove these and see what happens.
     
    Last edited: Dec 22, 2011
  18. Riever

    Riever L2: Junior Member

    Messages:
    86
    Positive Ratings:
    0
    :thumbdown:

    Oh well removing the lights in the tunnel did not work. So I'm baffled. Water that is illuminated by the 'sun' has no problem. The level renders correctly and all the brushes are lit.

    The same water bush in a tunnel results in Vrad going mad, failing to light any of the brushes, including those lit by the sun. It also goes mad when calculating Bounce lights. This is an edited output of Vrad when it does not work, followed by one that does (minus water).

    There are no leaks and no invalid solids.
    Sorry in advance for such a long post.

    Dud output - fails to light brushes.

    Setting up ray-trace acceleration structure... Done (6.28 seconds)
    ]21537 faces
    23 degenerate faces
    2886845 square feet [415705792.00 square inches]
    206 Displacements
    745724 Square Feet [107384384.00 Square Inches]
    21514 patches before subdivision
    zero area child patch
    zero area child patch
    zero area child patch
    zero area child patch
    295508 patches after subdivision
    sun extent from map=0.008727
    257 direct lights
    BuildFacelights: 0...1...2...3...4...5...6...7...8...9...10 (79)
    BuildVisLeafs: 0...1...2...3...4...5...6...7...8...9...10 (255)
    transfers 31226041, max 2037
    transfer lists: 238.2 megs
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #1 added RGB(1, 1, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #2 added RGB(1, 1, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #3 added RGB(1, 1, 1)

    ... and so on till ...

    Bounce #97 added RGB(1, 1, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #98 added RGB(1, 1, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #99 added RGB(1, 1, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #100 added RGB(1, 1, 1)
    Build Patch/Sample Hash Table(s).....Done<0.1563 sec>
    FinalLightFace: 0...1...2...3...4...5...6...7...8...9...10 (32)
    FinalLightFace Done
    0 of 0 (0% of) surface lights went in leaf ambient cubes.
    ThreadComputeLeafAmbient: 0...1...2...3...4...5...6...7...8...9...10 (30)
    Writing leaf ambient...done
    Ready to Finish

    Object names Objects/Maxobjs Memory / Maxmem Fullness
    ------------ --------------- --------------- --------
    models 145/1024 6960/49152 (14.2%)
    brushes 4163/8192 49956/98304 (50.8%)
    brushsides 33438/65536 267504/524288 (51.0%)
    planes 24946/65536 498920/1310720 (38.1%)
    vertexes 40733/65536 488796/786432 (62.2%)
    nodes 11531/65536 368992/2097152 (17.6%)
    texinfos 5421/12288 390312/884736 (44.1%)
    texdata 169/2048 5408/65536 ( 8.3%)
    dispinfos 206/0 36256/0 ( 0.0%)
    disp_verts 16238/0 324760/0 ( 0.0%)
    disp_tris 25600/0 51200/0 ( 0.0%)
    disp_lmsamples 984620/0 984620/0 ( 0.0%)
    faces 21537/65536 1206072/3670016 (32.9%)
    hdr faces 0/65536 0/3670016 ( 0.0%)
    origfaces 13175/65536 737800/3670016 (20.1%)
    leaves 11677/65536 373664/2097152 (17.8%)
    leaffaces 27854/65536 55708/131072 (42.5%)
    leafbrushes 12783/65536 25566/131072 (19.5%)
    areas 11/256 88/2048 ( 4.3%)
    surfedges 161231/512000 644924/2048000 (31.5%)
    edges 97645/256000 390580/1024000 (38.1%)
    LDR worldlights 257/8192 22616/720896 ( 3.1%)
    HDR worldlights 0/8192 0/720896 ( 0.0%)
    leafwaterdata 1/32768 12/393216 ( 0.0%)
    waterstrips 2948/32768 29480/327680 ( 9.0%)
    waterverts 0/65536 0/786432 ( 0.0%)
    waterindices 50343/65536 100686/131072 (76.8%)
    cubemapsamples 0/1024 0/16384 ( 0.0%)
    overlays 27/512 9504/180224 ( 5.3%)
    LDR lightdata [variable] 18670744/0 ( 0.0%)
    HDR lightdata [variable] 0/0 ( 0.0%)
    visdata [variable] 5900061/16777216 (35.2%)
    entdata [variable] 238626/393216 (60.7%)
    LDR ambient table 11677/65536 46708/262144 (17.8%)
    HDR ambient table 11677/65536 46708/262144 (17.8%)
    LDR leaf ambient 32502/65536 910056/1835008 (49.6%)
    HDR leaf ambient 11677/65536 326956/1835008 (17.8%)
    occluders 0/0 0/0 ( 0.0%)
    occluder polygons 0/0 0/0 ( 0.0%)
    occluder vert ind 0/0 0/0 ( 0.0%)
    detail props [variable] 1/12 ( 8.3%)
    static props [variable] 1/63660 ( 0.0%)
    pakfile [variable] 6688/0 ( 0.0%)
    physics [variable] 1621067/4194304 (38.6%)
    physics terrain [variable] 73850/1048576 ( 7.0%)

    Level flags = 0

    Total triangle count: 64691

    This Vrad works. All I have done is deleted the water brush.

    Setting up ray-trace acceleration structure... Done (6.29 seconds)
    21430 faces
    23 degenerate faces
    2870252 square feet [413316352.00 square inches]
    206 Displacements
    745724 Square Feet [107384384.00 Square Inches]
    21407 patches before subdivision
    zero area child patch
    zero area child patch
    zero area child patch
    zero area child patch
    293589 patches after subdivision
    sun extent from map=0.008727
    257 direct lights
    BuildFacelights: 0...1...2...3...4...5...6...7...8...9...10 (80)
    BuildVisLeafs: 0...1...2...3...4...5...6...7...8...9...10 (255)
    transfers 30857548, max 2038
    transfer lists: 235.4 megs
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #1 added RGB(1382200, 980035, 679566)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #2 added RGB(281515, 167511, 98811)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #3 added RGB(75416, 37647, 18693)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #4 added RGB(22841, 9529, 3911)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #5 added RGB(7736, 2690, 893)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #6 added RGB(2833, 822, 217)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #7 added RGB(1112, 268, 55)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #8 added RGB(460, 92, 15)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #9 added RGB(199, 33, 4)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #10 added RGB(89, 12, 1)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #11 added RGB(41, 5, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #12 added RGB(19, 2, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #13 added RGB(9, 1, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #14 added RGB(4, 0, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #15 added RGB(2, 0, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (0)
    Bounce #16 added RGB(1, 0, 0)
    GatherLight: 0...1...2...3...4...5...6...7...8...9...10 (1)
    Bounce #17 added RGB(0, 0, 0)
    Build Patch/Sample Hash Table(s).....Done<0.1540 sec>
    FinalLightFace: 0...1...2...3...4...5...6...7...8...9...10 (32)
    FinalLightFace Done
    0 of 0 (0% of) surface lights went in leaf ambient cubes.
    ThreadComputeLeafAmbient: 0...1...2...3...4...5...6...7...8...9...10 (32)
    Writing leaf ambient...done
    Ready to Finish

    Object names Objects/Maxobjs Memory / Maxmem Fullness
    ------------ --------------- --------------- --------
    models 145/1024 6960/49152 (14.2%)
    brushes 4162/8192 49944/98304 (50.8%)
    brushsides 33432/65536 267456/524288 (51.0%)
    planes 24940/65536 498800/1310720 (38.1%)
    vertexes 40615/65536 487380/786432 (62.0%)
    nodes 11474/65536 367168/2097152 (17.5%)
    texinfos 5418/12288 390096/884736 (44.1%)
    texdata 166/2048 5312/65536 ( 8.1%)
    dispinfos 206/0 36256/0 ( 0.0%)
    disp_verts 16238/0 324760/0 ( 0.0%)
    disp_tris 25600/0 51200/0 ( 0.0%)
    disp_lmsamples 984620/0 984620/0 ( 0.0%)
    faces 21430/65536 1200080/3670016 (32.7%)
    hdr faces 0/65536 0/3670016 ( 0.0%)
    origfaces 13174/65536 737744/3670016 (20.1%)
    leaves 11620/65536 371840/2097152 (17.7%)
    leaffaces 27739/65536 55478/131072 (42.3%)
    leafbrushes 12761/65536 25522/131072 (19.5%)
    areas 11/256 88/2048 ( 4.3%)
    surfedges 160801/512000 643204/2048000 (31.4%)
    edges 97393/256000 389572/1024000 (38.0%)
    LDR worldlights 257/8192 22616/720896 ( 3.1%)
    HDR worldlights 0/8192 0/720896 ( 0.0%)
    leafwaterdata 0/32768 0/393216 ( 0.0%)
    waterstrips 2960/32768 29600/327680 ( 9.0%)
    waterverts 0/65536 0/786432 ( 0.0%)
    waterindices 50499/65536 100998/131072 (77.1%)
    cubemapsamples 0/1024 0/16384 ( 0.0%)
    overlays 27/512 9504/180224 ( 5.3%)
    LDR lightdata [variable] 18461224/0 ( 0.0%)
    HDR lightdata [variable] 0/0 ( 0.0%)
    visdata [variable] 5810022/16777216 (34.6%)
    entdata [variable] 238626/393216 (60.7%)
    LDR ambient table 11620/65536 46480/262144 (17.7%)
    HDR ambient table 11620/65536 46480/262144 (17.7%)
    LDR leaf ambient 51212/65536 1433936/1835008 (78.1%)
    HDR leaf ambient 11620/65536 325360/1835008 (17.7%)
    occluders 0/0 0/0 ( 0.0%)
    occluder polygons 0/0 0/0 ( 0.0%)
    occluder vert ind 0/0 0/0 ( 0.0%)
    detail props [variable] 1/12 ( 8.3%)
    static props [variable] 1/63658 ( 0.0%)
    pakfile [variable] 5821/0 ( 0.0%)
    physics [variable] 1620430/4194304 (38.6%)
    physics terrain [variable] 73850/1048576 ( 7.0%)

    Level flags = 0

    Total triangle count: 64479
     
    Last edited: Dec 22, 2011
  19. grazr

    aa grazr Old Man Mutant Ninja Turtle

    Messages:
    5,436
    Positive Ratings:
    3,573
    If you're using "expensive" water you can only use it at 1 hieght. Use "cheap" water textures for other heights. Such as where you may have used them as puddles etc.

    I wont bother explaining too much as it's a fiddly topic, but in a lamens sense it's memory related, it allows the engine to focus high quality reflections on a single plane.
     
    Last edited: Dec 23, 2011
  20. Pocket

    aa Pocket func_croc

    Messages:
    4,489
    Positive Ratings:
    2,219
    Aren't all the default waters in TF2 cheap, anyway? I've never seen reflections in official maps that weren't obviously cubemap-based.