[Omniscaper]

I know your 1/20/04 build is still a work in progress but  it is already excellent!!! My Trek ships have always shown texture anomolies because of their many subobjects. I've been slaving away, trying to make them single meshes in order to get rid of the texture nasties. I'm aesthsteically inclined =). You SCP folks just saved me from this particular obsessive cumpulsive disorder. Now I can concentrate on more releases. THANK YOU THANK YOU THAN YOU. I'll gladly donate once I'm out of art school, one more semester to go and then I'm a rich gallery showing artist. In the mean time I need the cash for painting my masterpieces.

[Bobboau]

eah... wha?
(post screens:D)

[Omniscaper]

Its evident on the my very first picture of the Galaxy Class. Click on that thingy under my name. (I just noticed that "thingy" a couple of days ago) Its on the high poly version, check out the nacelles.

[KARMA]

I think he means this: http://3dap.com/hlp/staff/setekh/omniscaper/galaxy_high04.jpg
it looks somehow like z-buffer error to me, since he's talking of differnt subobjects, but it's hard to say without seeing the model. Like if the nascelle's plating is a subobject surrounding another subobject, nascelle's blue area(don't knoe the right word), for all the lenght, too close to each other. But as we know the z-buffer error was already fixed, so I don't know

[Drew]

theres a new build? or do i need to compile it?

[Bobboau]

oh, that, it's a clipper error, I've been meaning to fix that, looks like someone got to it first

[Lightspeed]

whee thanks -- I havent had time to try yet but I had those problems too

Is the HT&L displacement bug fixed?

(particles floating off your shots when rotating the camera, the targetting box around a ship floating off it when rotating, etc)

-edit: no, it's not fixed, I just saw it

[Lightspeed]

Alright, I've been testing the build again

Note: I've also re-tested with the 12_05_03 build since I have some extra content (weapons + planets, hi poly herc, etc) since I last tested

COMPARISON:

12_05 build - light blue = better performance, dark blue = worse performance

1_20 build - yellow = better performance, red = worse performance

Polygons & Textures

Detail Level 1:

85 FPS | 85 FPS

Please note that Detail Level One is 9 (nine!) Hercules fighters on screen (and YES, I use the hi-poly version )

Detail Level 2:

42 FPS | 44.3 FPS

Detail Level 3:

42 FPS | 42.5 FPS

Detail Level 4: (Insane amount of detail)

28.3 FPS | 28.3 FPS


ANI performance

Low Use:

min: 85 FPS | 85 FPS
max: 85 FPS | 85 FPS
avg: 85 FPS | 85 FPS

Medium Use:

min: 85 FPS | 85 FPS
max: 85 FPS | 85 FPS
avg: 85 FPS | 85 FPS

High Use:

min: 42 / 27 FPS | 42 / 35 FPS (It's swapping between 42 and 35 all the time)
max: 85 FPS | 85 FPS
avg: 42-27 FPS | 42-35 FPS


Weapon Effects
(using the standard particle rendering)

Low:

min: 85 FPS | 85 FPS
max: 85 FPS | 85 FPS
avg: 85 FPS | 85 FPS

Medium:

min: 78 FPS | 85 FPS
max: 85 FPS | 85 FPS
avg: 85 FPS | 85 FPS

High:

min: 50 FPS | 61 FPS
max: 70 FPS | 85 FPS
avg: 61 FPS | 75 FPS


Looks pretty good to me

[J3Vr6]

Um, what's the difference between the blue and the yellow? they both say 12-05

[Lightspeed]

oops. I typoed. *fixes*

[SadisticSid]

Phew, thought it might be shiny new 256MB 9800 Pro on the blink already. Now to persuade Phreak to compile an Inferno build...

[phreak]

i have one built.  i just cant get FRED to compile

[Bobboau]

just wondering, who fixed that?

[J3Vr6]

How about doing a test with more ships in mission, to be comparable to an original campaign mission.

Have a couple of capships, some shivan swarming, etc.  Also would be good if you put your own system specs down so we can see what machine is running those tests.

[RandomTiger]

Originally posted by PhReAk
i have one built.  i just cant get FRED to compile

Thats probably something to do with me, what does it say?

[phreak]

Code: [Select]


libcimt.lib(filebuf.obj) : error LNK2001: unresolved external symbol ___pioinfo
c:\Games\Freespace2\fred2_open_inf.exe : fatal error LNK1120: 1 unresolved externals


ok i got it working.  i had to remove the iostream usage from TCP_Socket.cpp and set libcmt.lib libcimt.lib and libci.lib to be ignored by the compiler.

[Lightspeed]

Originally posted by J3Vr6
How about doing a test with more ships in mission, to be comparable to an original campaign mission.

Have a couple of capships, some shivan swarming, etc.  Also would be good if you put your own system specs down so we can see what machine is running those tests.

uhm.. detail level 4 is:

9 Hercs
1 Arcadia
5 Amazons
6 Herc Mk IIs
6 Myrmidons
6 Maras
6 Perseus
5 Fenris cruisers
1 Sathanas
1 Hecate

All in your screen at the same time.

Not enough?

Look in the "request for testing" if you want my test files.

[Omniscaper]

Oh, my bad. I recently upgraded my ATI Catalyst drivers to 4.1 and that did the trick. I recently been tracking a system crash problem and reverted to the older 3.9 drivers, and the texture anomolies reappeared on all subobjects. I guess its a lousy driver, unlike the newest release. Sorry to stir up a commotion.

[J3Vr6]

Wow, were those the hi poly fenris models?