Tim Sweeney on shadows and HDR rendering on ATi cards

[XxStratoMasterXx]

@ Dave- Unreal Engine 3 was demoed publically before NV40's release?

 

[Dave Baumann]

@ Dave- Unreal Engine 3 was demoed publically before NV40's release?

IIRC they did the initial demos at GDC last year and they had only seen NV40 and R420 mere days before the show (I remember talking to Mark Reign about it at the NV40 editors day in April just after). The point bein is that most of the development of the public demo's had been done on R300/NV30 hardware prior to the public demonstrations. I do no recall them having rolled in the HDR into all the other demo scene they had done, but I believe HDR had its own demo scene. That was many months ago now, of course, and things will certainly have changed, they just haven't done any more recent tech demos.

 

[Bjorn]

And if that is the case the 20% figure doesn’t really tell us much in relation to the question since it can only come from a comparison of two things: either NV4x running the two modes, which doesn’t give an indication of performance for R300/420, or comparing NV40 with FP filtering/blending against R420 with the shader filtered method which is a performance comparison between these two boards and not indicative of the “total performance differenceâ€￾.

True. Hmm, needs some more explanation from mr Sweeney i guess.

 

[jb]

Perhaps I should ask for clarification?

Geez,

First it was JC post that left you almost more confused then when you started as his statements usally were left open to how you wanted to read it. Now Sweeny is doing the same??? Argggg.... Sorry week atempt at humor. But do let us now if he does reply!!

 

[KimB]

And if that is the case the 20% figure doesn’t really tell us much in relation to the question since it can only come from a comparison of two things: either NV4x running the two modes, which doesn’t give an indication of performance for R300/420, or comparing NV40 with FP filtering/blending against R420 with the shader filtered method which is a performance comparison between these two boards and not indicative of the “total performance differenceâ€￾.

No, but comparing the NV4x running the two modes can give us a ballpark figure of what sort of performance penalty the R3xx would incur. Anyway, more specific information just wouldn't help us at this stage, since any game based upon said engine is far from final, and thus the performance difference is sure to change significantly before final release.

So, what you should read is simply that the R3xx can do HDR in the engine, but it's slower. Whether this requires effects being turned down or off, or results in reduced image quality (i.e. for particle effects), we will have to see.

 

[Dave Baumann]

Thanks Chalnoth, that was a really useful post... no, wait, lets file that under the "Duh!" category.

 

[KimB]

See, Dave, you just need to learn to make your meaning more plain.

 

[XxStratoMasterXx]

Perhaps I should ask for clarification?

Geez,

First it was JC post that left you almost more confused then when you started as his statements usally were left open to how you wanted to read it. Now Sweeny is doing the same??? Argggg.... Sorry week atempt at humor. But do let us now if he does reply!!

EDIT: Nevermind, misread your post.

However, as Chalnoth has said, any game based on the Unreal Engine 3 is far from a completed state, and things will change. The intention of my question to Tim, was really a "how do you get HDR to work on R300 cards" rather than a "how much of a performance hit does using HDR on a R3xx card is noticed".