Shadow calculation and lightmaps explained
Moderator: English Moderator
Re: Shadow calculation and lightmaps explained
Thanks, I will check once more to be sure it's really 3x faster ON THE SAME MAP ^^
Re: Shadow calculation and lightmaps explained
We have tested the computing of the same map with 2 similar configurations. One based on ATI Radeon the othe based on Nvidia GeForce :
The map is : http://sm.mania-exchange.com/maps/1276/ ... -au-centre
(copyright NONO ^^).
My configuration :
Asus P8P67 - Core i7 2600k@3.4GHz - 8GB DDR3 1600 - ATI Radeon 7950 => 16 minutes for Ultra quality
Franck's configuration :
Asus P8Z77 - Core i7 3770k@3.5GHz - 8GB DDR3 1600 - GeForce GTX 680 => 11 minutes for Ultra quality
The real difference between the 2 CPUs is around 5%. The computing is 45% faster with Nvidia !
If you want to benchmark with this map too. Please post here your results with the same informations (Motherboard, CPU, Mem, Graphic card).
The map is : http://sm.mania-exchange.com/maps/1276/ ... -au-centre
(copyright NONO ^^).
My configuration :
Asus P8P67 - Core i7 2600k@3.4GHz - 8GB DDR3 1600 - ATI Radeon 7950 => 16 minutes for Ultra quality
Franck's configuration :
Asus P8Z77 - Core i7 3770k@3.5GHz - 8GB DDR3 1600 - GeForce GTX 680 => 11 minutes for Ultra quality
The real difference between the 2 CPUs is around 5%. The computing is 45% faster with Nvidia !
If you want to benchmark with this map too. Please post here your results with the same informations (Motherboard, CPU, Mem, Graphic card).
Re: Shadow calculation and lightmaps explained
So its not 3x faster 

Re: Shadow calculation and lightmaps explained
Indeed and sorry for that ... I based this on the average time my friend Franck tell he had for computing all the map of its server. Not exactly the same maps, just an estimation, ...
It's why I thought it was better to test with a same and huge map
The best test would be to replace the graphic card in the same computer and compare. But I have only Radeon at home.

It's why I thought it was better to test with a same and huge map

The best test would be to replace the graphic card in the same computer and compare. But I have only Radeon at home.
Re: Shadow calculation and lightmaps explained
Hello, I'm the friend of Niarf
,just some precisions, some map take 5 minutes to calculate and others take 20,it's very inconstant
. I think 10 minutes is a good average.


Re: Shadow calculation and lightmaps explained
Translated from Problèmes texures, Fix nicely illuminates the issue with wrong lightmaps on the servers
:

[b][color=#FF8000]Fix[/color][/b] wrote:Yes, the maps are provided by the servers and transferred to the players. If the servers did not calculate the lightmaps again, the players will get "buggy" maps. And especially when blocs have been modified, the place where their lightmap information is read and stored might change.
To illustrate this, imagine the lightmap is a worldmap.
In the beta 1.3 the block "arch" draws light in France, but after some modifications to debug and optimize, since the beta 2, it wil read its lightmap now in Germany... so if the lightmap was calculated again properly, the correct information will really be located in Germany, but with a older lightmap from the previous version, Germany has the image data for lights of the poles or bumpers as an example... therefore the arch will show shadows and lights which are not meant for it.
In the beginning we had systematically a new calculation at the smallest detected modification... even if we modified the block only a little bit without any effect on the lighmaps.
Now the detection of differences is only done when you open the map in the editor to avoid destroying that the online experience with constant lightmap recompute before starting a match.
Since we are in beta, the modifications might be very important from one version to another, as many blocks often go back and forth between hammer and anvil.
Re: Shadow calculation and lightmaps explained
Thank you very much TMarc 
I modified one or 2 words, because I felt some of my sentences 'intentions' weren't easy to translate.

I modified one or 2 words, because I felt some of my sentences 'intentions' weren't easy to translate.
Re: Shadow calculation and lightmaps explained
the question is now, which words Fix has exchanged, and where 

Re: Shadow calculation and lightmaps explained
hammer and anvil , sureTMarc wrote:the question is now, which words Fix has exchanged, and where

A lot to gain ?
Who is online
Users browsing this forum: No registered users and 1 guest