Chalnoth said:
demalion said:
Oh my, you seem to have selectively skipped some sentences, Chalnoth. Here, let me get the rest of that paragraph for you:
Chalnoth said:
In these situations, the FX's benefit will be in texture quality (high res + aniso will beat the texture quality of somewhat lower-res + aniso), whereas the 9700's benefit will be in edge quality at somewhat lower resolutions. Said another way, back to the old debate, what would you rather have? Higher resolution, or more FSAA?
Oh, not at all like ""the GF FX is the obvious choice for high resolution gaming", right? I must have misunderstood and the GF FX must have been the "more FSAA" choice?
Did you notice the part previously about more fillrate-limited games?
Did you notice the part about "partial quoting", neatly snipped here with no reply from you? Do your remember your argument about higher quality you seem to be trying to selectively snip?
Apparently not. Some more snipped text concerning "fillrate-limited games", written by me, ignored by you:
demalion said:
What you are doing (and you haven't changed, from molex power adapters, to this, to the next thing...you get "points" for consistency) is suiting what exists in your world to what you want to exist there, by, for example, arbitrarily excluding consideration of games that are able to offer both FSAA and higher resolution (a situation where, "coincidentally", the 9700, shines, and the GF FX does not, in comparison).
How about my response
specifically quoting your mention of "fillrate-limited games", again snipped, again making your question a waste of time:
demalion said:
I also stated that for older games, the FSAA quality and performance of the FX are of less importance, as it will look good and be fast no matter what. The real differences will be seen in more fillrate-limited games.
The thing is you lump Serious Sam and Quake III, and
any game that is not...hmm...let's say 10 fps faster average than UT2k3 at 1600x1200x32 with AA and aniso (some people consider 50 to 70 fps average playable you know) in with counter-strike under the heading of "older game" where "the FSAA quality and performance of the FX are of less importance", based solely on the fact that they can be playable at high resolution with AA. It is just a "coincidence" that as much as the FX will "look good and be fast no matter what" in these circumstance, the 9700 might happen to look
better, and, god forbid, be
faster...so these high resolution circumstances don't matter when comparing. A choice between "Higher resolution, or more FSAA", you say?
Why is it
only "fill-rate" limitation matters, and "bandwidth limitation" not matter? You'd think both matter for high resolution...or atleast I would. Let's see, is it because fillrate in actual games doesn't require bandwidth? Hmm...no, no, that doesn't make sense. It couldn't be because the GF FX has less bandwidth than the 9700 and more fillrate, is it? Nah, that would mean when you snipped:
demalion said:
"Disregarding positive things about the 9700" does not equal "highlighting positive things about the GF FX".
you just wasted my time and text again, and that couldn't be, could it?
Anyway, this is, very roughly, the comparison I expect (I don't think this has been done yet):
In UT2k3, for instance, the Radeon 9700 Pro at 1280x1024x4 (4x FSAA, that is) should perform similarly to the GeForce FX at 1600x1200x2 (The FX will likely be a little bit slower, but the performance should still be very comparable).
What about the 9700 Pro at 1600x1200x2? You know, the discussion you are selectively snipping wasn't about the GF FX performing faster at 1600x1200x2 with regards to the 9700, but your arbitrary (and now apparently inconvenient and therefore selectively removed by you here) determination that it had higher image quality while being faster, presuming the degree to which it was faster was conclusive, and ignoring all cases where it was not.
Yes, for example that aniso comparison you snipped, I mentioned in that last post, and that you snipped again.
To be fair, however, it seems the AA quality is going to be closer than most, if not all, screenshots have illustrated. In fact, to be fair, I mentioned just that in the other thread, and in fact allowed your assumption that 2x AA on both was equal in quality for comparison. Deja vu.
For games that are more fillrate-limited, as I stated, it's a mute point. Notice that I did single that out.
Yes I did notice that you blithely assume the 9700 versus GF FX at 1600x1200x2 is an open and shut case in terms of quality and performance.
The FX definitely isn't universally better for higher resolutions (since the 9700 can run just fine at very high resolutions w/ FSAA in most games).
Do you think I would have argued with something like this? Do you remember the comments like "absolutely better" texture filtering you used to support your argument?
Are you trying to rehash your argument here while pretending you didn't make such comments, under the banner of "I got picked on for highlighting the positives about the GF FX"?