What's the verdict on Halo's built-in Timedemo?

[ragejg]

Is it a usable performance indicator? Is it buggy? Does it not enable/disable feature options correctly? Is it worth benchmarking with?

... I have separate gameplay benched already with fraps, but I also have plenty of timedemo results ... I've heard rumblings in many forums debating the timedemo's validity... causing me to question it myself...

:confused:

[saturnotaku]

It does a good job in doing what it was designed to do. However, I don't like the way it interprets results. Fraps' measurements are more to my liking.

[ragejg]

It appears to read fps even on the transisitons and the stutters that go along with them... on an average, teh halo timedemo results I've tabulated are ~2pfs lower than when I measure each sequence with fraps... right now the 5900's doin 38 @ 10x7, all highest IQ, no aa/af, specular on @ stock, 41 OC'ed... FRAPS gives me 40/43... without specular, the timedemo gives me 40 and 43, and fraps gives me 42 and 45...

... my cpu must be limiting the perf. increase of disabling specular...

... and why in the world does Particles get turned off in the timedemo? I have it enabled in halo's menu options... :confused:

[digitalwanderer]

It sucks & it's silly!

Why benchmark the cut-scenes? They ain't very representative of the games actual gameplay. :rolleyes:

[bkswaney]

Is it built in the demo?
I have not seen it anywhere? :confused:

[ragejg]

Originally posted by digitalwanderer
It sucks & it's silly!

Why benchmark the cut-scenes? They ain't very representative of the games actual gameplay. :rolleyes:

Um, dig... I guess from what I've read, the cut scenes are in fact not part of the equation... TThere's an explanation for it somewhere ... in some faq file, I think.... and in that explanation it speaks of separately benchmarking the scenes... duh...

That duh was to me for making the incorrect observation... :p