[RESOLVED] More AdvancedLightMapping woe/bizarrety
by Steve Acaster · in Torque 3D Professional · 10/12/2009 (5:07 pm) · 14 replies
Is "bizarrety" even a word? And how many Z and T should it possess? Should that E be an I? meh ...
Anyhoo, my poor old Nvidia 7900GS 256mb has had it's share of AdvancedLightMapping issues over the Beta, burnt out spec, alpha everywhere lightmapped, alpha everywhere not lightmapped, no lightmapped meshes rendering, no none lightmapped meshes rendering - and now no lights affecting anything at all - hence my use of the phrase bizarre which singularly I can spell.
The various random issues seem without any particular cause, though I have suddenly noticed that they tend to occur on restart (and each and every subsequence restart) if I have not saved the level before exiting. So, if I notice oddities, alter something, save, exit, restart, it appears to be fixed - until I exit and restart without saving.
This is probably barking up the wrong tree but I get 5 or so of this in the console on loadup, nothing else of interest or suspicion though.
[FileSystemChangeNotifier::internalNotifyDirChanged] : [/shaders/procedural]
crappy 7 series GPUs FTW!!!!
EDIT: Come 101, all seems well. Good work, chaps.
Anyhoo, my poor old Nvidia 7900GS 256mb has had it's share of AdvancedLightMapping issues over the Beta, burnt out spec, alpha everywhere lightmapped, alpha everywhere not lightmapped, no lightmapped meshes rendering, no none lightmapped meshes rendering - and now no lights affecting anything at all - hence my use of the phrase bizarre which singularly I can spell.
The various random issues seem without any particular cause, though I have suddenly noticed that they tend to occur on restart (and each and every subsequence restart) if I have not saved the level before exiting. So, if I notice oddities, alter something, save, exit, restart, it appears to be fixed - until I exit and restart without saving.
This is probably barking up the wrong tree but I get 5 or so of this in the console on loadup, nothing else of interest or suspicion though.
[FileSystemChangeNotifier::internalNotifyDirChanged] : [/shaders/procedural]
crappy 7 series GPUs FTW!!!!
EDIT: Come 101, all seems well. Good work, chaps.
About the author
One Bloke ... In His Bedroom ... Making Indie Games ...
#2
10/15/2009 (5:07 pm)
Turns out its' bizarrity after all ... silly me. 
#3
1. bizarrity - adj: conspicuously or grossly unconventional or unusual;
The bizarrity of New Year's Eve, 2005 was one that will be remembered forever.
rofl
10/15/2009 (5:19 pm)
1. bizarrity - adj: conspicuously or grossly unconventional or unusual;
The bizarrity of New Year's Eve, 2005 was one that will be remembered forever.
rofl
#4
10/15/2009 (7:11 pm)
2000 and when? 
#5
10/15/2009 (8:58 pm)
I cannot remember if that is a laptop card, but perhaps there is an issue with that card compared to a software/driver issue? If it is in a regular computer might make sure it isn't physically defective. Look at the caps to make sure they are all in 1 piece or something.
#6
Seen the post on the caps thought i would post this on them, so theres no confusion on what to look for.
10/15/2009 (9:59 pm)
The surface mount caps today are junk they do not last long at all. If you suspect a cap problem check the top for buldging and if you see that then thats the defective cap or caps. Most of these caps today lose there dielectric and dry up, they will look physically look fine but you will never know unless you have an ers meter to check for the cap being dry inside. If a cap is bad by being charged beyond its rated b+ value it will open up on the top like a can being blown open, thats what the relief marks(cross pattern)in the top are for. Any black stain around the base of the cap means that cap is losing it's dielectric thru the lead holes at the base of the cap. The dielectric will soak into the board and destroy it over time.Seen the post on the caps thought i would post this on them, so theres no confusion on what to look for.
#7
10/15/2009 (10:32 pm)
As I have absolutely no idea about hardware, the more info the better, Harold. 
#9
10/15/2009 (11:30 pm)
excellent picture for seeing what to look for on the older caps. shows the buldge and the leaking dielectric. Usually on the new caps in his picture you will just see them buldged, very rare to see the dielectric like that at the top. the surface mount caps will leak the dielectric at the base usually or lose the uf rating or just go dry.
#10
After giving my GPU caps a quick fondling (you don't get to say that everyday!) they appear to be fine -- which is more than I can say for Nvidia's October driver ... which won't let me run in any windowed format except 4:3/1024x768.
I'll put it all my ALM issues down to a crappy old GPU.
10/16/2009 (12:56 am)
Interesting stuff.After giving my GPU caps a quick fondling (you don't get to say that everyday!) they appear to be fine -- which is more than I can say for Nvidia's October driver ... which won't let me run in any windowed format except 4:3/1024x768.
I'll put it all my ALM issues down to a crappy old GPU.
#11
From what im seeing a single core processor and these older cards are not going to cut it with this engine. Hopefully this will be fixed soon. There are still so many people running these older setups, i sure hope we get a bit of loving too. If we do not get the loving i see it hurting any makers game sales and will probably affect the engine sales down the road. What parent is going to buy there kid a top of the line machine to go play a video game(i know i would'nt).
I jumped into COD to see its cpu usage and it's only 17% on a fully loaded map and i do not lag.
10/16/2009 (2:09 am)
i'm having major lag issues with AL on my 7800gs card. The map i'm making is no where close to being finished(its very basic, maybe 30 trees all LOD are in, 2 buildings very small, maybe 50 bushes, 10 or 15 rocks) should'nt be lagging at all. My cpu usage is thru the roof(90 to 100%)all the time. It's so bad its unplayable, theres no way you can sell a game that acts that way. From what im seeing a single core processor and these older cards are not going to cut it with this engine. Hopefully this will be fixed soon. There are still so many people running these older setups, i sure hope we get a bit of loving too. If we do not get the loving i see it hurting any makers game sales and will probably affect the engine sales down the road. What parent is going to buy there kid a top of the line machine to go play a video game(i know i would'nt).
I jumped into COD to see its cpu usage and it's only 17% on a fully loaded map and i do not lag.
#12
This is just saying that its updated a file in that folder, and is reloading it.
Also, the performance issues are interesting that its on certain machines.
10/16/2009 (7:50 am)
[FileSystemChangeNotifier::internalNotifyDirChanged] : [/shaders/procedural]This is just saying that its updated a file in that folder, and is reloading it.
Also, the performance issues are interesting that its on certain machines.
#13
I had trouble running my TGE demo on a single core.
Harold, if you've experienced crashes, check that you don't have previous instances of T3D still running in the background - I've found that after a crash I sometimes have to go into taskmanager and close the app there. That can really eat memory.
10/16/2009 (10:38 am)
single core processorI had trouble running my TGE demo on a single core.
Harold, if you've experienced crashes, check that you don't have previous instances of T3D still running in the background - I've found that after a crash I sometimes have to go into taskmanager and close the app there. That can really eat memory.
#14
I have cpu usage of 89% to 97% on the dersetball map.
In burg the cpu usage is 89% to 100% it hits 100% as soon as i turn to see the machine animation it literally locks up then releases, major lag issues to the point that it's no use to even load that map.
specs: AMD Sempron 3400+ 2.0 ghz (single core)
2 GIG OF RAM
7800GS CARD
If GG wants to see this 1st hand, when i have a free day. (i'm 15 miles from your guys studio, so let me know).
edit: check the blank room running 92% this is crazy.
10/16/2009 (11:47 am)
I've seen that and experienced that. I'm not running out of memory. I only have one instance of it running. I have cpu usage of 89% to 97% on the dersetball map.
In burg the cpu usage is 89% to 100% it hits 100% as soon as i turn to see the machine animation it literally locks up then releases, major lag issues to the point that it's no use to even load that map.
specs: AMD Sempron 3400+ 2.0 ghz (single core)
2 GIG OF RAM
7800GS CARD
If GG wants to see this 1st hand, when i have a free day. (i'm 15 miles from your guys studio, so let me know).
edit: check the blank room running 92% this is crazy.
Torque Owner Bryan Morgan
Bunker 42 Games
I get a bunch of "[FileSystemChangeNotifier::internalNotifyDirChanged] : [/shaders/procedural]" stuff but no issues with the lighting on my Geforce 8900 GTX, besides what strange things can occur when toggling HDR without restarting.
BTW, You're probably wanting the word Bizarreness, not bizarrety