Quick question:

In the ATI Radeon article, you said that you wouldn't be using SC:CT for benchmarking anymore.

"Unfortunately, we found that Splinter Cell: Chaos Theory's copy-protection scheme, StarForce, was tampering with our CrossFire results, so we'll be following up with those numbers in our video quality article. Suffice to say, it is because of StarForce that we will no longer be using SCCT as a benchmark."

But you used it in the Albatron review. Did you mean you would skip it only for certain cards?

It could be because the testing for this review was done before the 2900 review. It could also be because the benchmark has problems with Crossfire, which wasn't tested in this review. I don't think it is skipped for certain cards... just dual-cards since it wouldn't be a good basis for comparison. I'm sure Max will see this and clear it up for you.

Brian's first assumption was right: I benched this card before we decided to drop SCCT, when I was benching the HD 2900 XTs. But given the hard launch and nature of the HD 2000 release, we ran that article first.

StarForce is a huge pain in the groin. Not only does it tamper with drivers (I have to assume, given how bizarre the test bench reacted after running it for the first time after installing the ATI drivers, (installing it doesn't seem to have an effect until it's actually run) but also some research by others) it doesn't work with Windows x64, MCE (I think) and Vista. So here's a good game that's unplayable by a sizable and growing percentage of users.

We're going to be adding the next Unreal Tournament game for SM 3.0, and generally moving towards Vista as an equally important platform for benchmarking. I'm not sure yet, but I think we're also going to add S.T.A.L.K.E.R. plus whatever DX10 games (that support benchmarking) come out.

Since I'm full-time for TTL, we've got the luxury of playing around with the hardware benchmarking format, although we're really still transitioning.

While I've got you here, what are some things you'd like to see in a video card review?

I mean, I have a list of things to consider, but is there something in particular you want more or less of?

Thank you both for the information. I'm not what you would call a computer hardware enthusiast, but I find the stuff interesting.

I think it would be nice to see more direct comparison type articles. Bang/buck at price range X sorts of things that might help me decide which video cards to buy, or perhaps a little more information on why VidCard Y's Z feature makes is interesting/cool/worth the money.

You guys tend to mention these sorts of things in your articles, but I, for one, would like a little more detail.

As for the video card benchmarking stuff, I like the brief blurbs on the games, but when reading this article and the Radeon article in proximity I got a little confused because the blurbs were the same in both.

Anyway, thanks for asking. I'm looking forward to seeing some the new benchmarking format as it evolves.

That's because the games are tested the same so we copy/paste parts of the review that apply to other reviews (such as the majority of the "Test Setup" page until our test rigs change, and the "Testing - X Game" pages as you mentioned). There would be no point in re-wording those sections just for the sake of it.

Remind me again, why do we have the blurbs? It spaces the page out right, or something?

Maybe if we did a "test setup" run-down on each game, like:

DirectX 9
Shader Model 2.0
Version 1.0h some revision
FRAPS + Opening sequence
Low Detail: Texture size medium, font size extra extra large
High Detail: Some anti-aliasing up in tha hizzy
Et cetera, foo

we'd have a more informative header.

Well, the game IMAGES are in there to take up space next to the 300x250 ad-slot and to look pretty.

The blurbs are there to tell the reader how each game was tested and any special notes about that game.