I've been testing Nexuiz for a long time on different hardware and software and I've saw that map "distonic" gives a lot less FPS than other maps on ATI cards while it doesn't look so complicated to compute by CPU or GPU. I've tried both open source and closed source drivers with my AGP Radeon 9600 and PCI-E X300 cards and while closed source driver was faster I get 5-10x less FPS on this map than on other maps on both drivers. This map is not playable on my graphic cards. I've tried the same Nexuiz versions on NVidia Quadro NV135M card with closed source NVidia driver and there's no such a big difference in FPS between maps. I thought maybe never kernel+drivers+xorg+nexuiz will help with it but it didn't. Now I use Kubuntu 9.10 with new open source drivers which support GL_ARB_vertex_buffer_object and GL_ARB_occlusion_query and I've enabled those extensions in Nexuiz, too. On low settings I get ~15-25 FPS on "distonic" while on other maps I can now play with >60 FPS most of the time. I wonder what's doing this slowness and if it's a bug in graphic card hardware/map/nexuiz?

Moved

What I could see in this map, that it uses an awful lot, highly detailed patches, though that shouldn't slow the map that much down.
Make sure you don't have realtime dynamic lights on. That's a real FPS killer on this map. Seems like whoever converted this map to Nexuiz, used all the original light entities of the map as dynamic lights.

I don't use realtime dynamic lightning because it's really FPS killer for most maps in Nexuiz on my graphic cards.