If you can run this you\'ll probably be able to run TTR:T.
|
10-24-2005, 08:52 PM
(This post was last modified: 10-24-2005 08:53 PM by joeb.)
Post: #11
|
|||
|
|||
RE: If you can run this you\'ll probably be able to run TTR:T.
try highest settings vips, that\'ll cripple your card for sure
at default I\'m running it at about 20fps @ 640 |
|||
10-24-2005, 08:53 PM
Post: #12
|
|||
|
|||
RE: If you can run this you\'ll probably be able to run TTR:T.
Ok doki....
ATI Radeon X800Pro 256Mb AGP8x (Club3D brand). With 4xAA and 4xAF 1280x1024@85Hz Fullscreen = 28fps 1024x768@85Hz Fullscreen = 37fps 640x480@85Hz Windowed = 80fps With no preset AA or AF 1280x1024@85Hz Fullscreen = 30fps 1024x768@85Hz Fullscreen = 42fps 640x480@85Hz Windowed = 100fps Full system spec: Intel Pentium 4 Prescott 3Ghz 2Gb DDR400 MSI 875P NEO-FSIR2 Mobo Windows XP SP2 Hope thats of more help Vips |
|||
10-24-2005, 08:59 PM
Post: #13
|
|||
|
|||
RE: If you can run this you\'ll probably be able to run TTR:T.
hmm oddly enough the program seems to be using 100% cpu usage as well..
|
|||
10-27-2005, 04:55 AM
Post: #14
|
|||
|
|||
RE: If you can run this you\'ll probably be able to run TTR:T.
Well obviously it calculates the shape with the cpu, and the gpu only renders it with fancy shit...
|
|||
10-27-2005, 05:11 AM
Post: #15
|
|||
|
|||
RE: If you can run this you\'ll probably be able to run TTR:T.
17 FPS @ 640x480 with Radeon X300SE card. Why is it so bad?
|
|||
10-27-2005, 05:59 AM
Post: #16
|
|||
|
|||
RE: If you can run this you\'ll probably be able to run TTR:T.
Well, to put a bit of the way this system likely calculates HDR into perspective, let me explain some of what this system does:
In order to acheive the effect of HDR(in \"true\" nature: i\'ll go into that later on )you have multiple steps at which this thing takes. First, you need to take a copy of the currently rendered scene. Taking this screen, you downscale it multiple times untill you have a very small resolution of the original scene. This mipped version is then used to calculate the average intensity of the scene. Once we have our average intensity, we make 2 more versions of the scene, an under exposed, and over exposed, and combine them together, this is used to create the blooms. We take our combined screen, do some gaussian blur to the overbrighted parts, and then draw all this back onto the screen. This takes a crapload of passes to do \"properly\". Now, why did i say true with quotations around it? Because at the moment, it is *impossible* to do true HDR on a computer. Why? because HDR is where you have lighting exposure outside of the 100:1 range. However, pretty well all monitors do not go past that. This is where the blooming effect in HDR implementation\'s ingame come from. In real life, you arent going to get a great big blob of white bloom off a lit surface, no matter what the exposure is unless a) you\'re directly looking into a lightsource b)you are looking at a reflective material that is reflecting a very powerful light source, such as the sun. we use bloom ingame, even though it\'s unrealistic, to portray the fact that it\'s very bright, and outside our standard range rendering because we cant actually do that with a monitor. Anyways, the reason performance might be so bad is because this is \"true\" HDR, which generally involves doing a crapload of re-rendering, screen capturing, and post processing. However, the system we\'re going to be using doesnt require most of this, acheives the same overall effect, and uses only pixel shaders 1.1 -Reno out |
|||
11-07-2005, 06:54 AM
Post: #17
|
|||
|
|||
RE: If you can run this you\'ll probably be able to run TTR:T.
cool
|
|||
« Next Oldest | Next Newest »
|
User(s) browsing this thread: 4 Guest(s)