Click for a bigger version
In terms of performance, Max Payne 2 actually runs faster than the original game based on the same machine spec, while looking better. What's the secret?

We did some major changes to the game engine, like Havok Physics that’s now used for collision checking. It’s a more efficient system than what we had for Max Payne 1.

Additionally our light map combiner was rewritten, and that reduced the primitive counts going to the graphics card by about a half. We also optimized our character update and skinning code - there are LODs in the skinning and the high detail version is SSE optimized.

If you could improve Max Payne 2, what would that be? Basically, your "it didn't make it into the game but I sure wish we could've implemented it" wish.

We’re just scratching the surface of physics gameplay. With more development time, I’m sure our guys would’ve figured out a lot of cool stuff to do with them.

There’s of course a long list of nitpicking wish list features that could be pasted here, but in the end all things said and done, I think the team’s pretty happy with how it turned out to be.

What part or aspect of Max Payne 2 are you most proud of? Why?

I’m most proud that we shipped the game. It’s a rare occasion for an independent developer as projects are getting longer and harder.

While the game's performance really isn't an issue with most of the newer video cards, what is your opinion of hardware vendors coming up with new drivers that specifically target Max Payne 2 for performance improvements with some sacrifices made to image quality that owners of the game have no control over? Would your opinion change if hardware review outlets consistently use Max Payne 2 (in their own ways, since the game does not feature any benchmarking mode) in their video card reviews?

This is a hot issue... I’m all for optimizations that add to the gameplay experience. When you take up the benchmarking aspect, the situation changes a lot.

Obviously the best for consumers would be that hardware would perform optimally with a "out of the box" game without needing to do application specific optimizations.

But if a vendor can make the game run significantly faster with optimizations, without ruining the visual experience, I'm all for it. As a gamer, I'll take +20 fps instead of anisotropic filtering for all texture layers any day.

But it would be great (and fair) to be able to disable ALL application specific optimizations for benchmarking use. Then the reviewers would have tools to analyze optimized and unoptimized performance and the possible quality tradeoffs.

Finally, what's next for Remedy? Max 1 and 2 are success stories. What have you learned from making these two games that you think will help you develop a better game in the future?

I could tell you but then… We’ve had a lot of fun and learned a lot from making Payne, and will take that into good use developing our next title. More on that later in the future.

 


We'd like to thank Markus for giving us the opportunity to chat with him. This interviewer, who personally just recently finished playing the game, was left mightily impressed by many things that make up a very good game -- a story that actually is engaging (I never once skipped any of the comic-style graphic novel interludes), a number of plot changes, great graphics, great performance and very polished production. And then there's also Ms Mona Sax, a sexy lady character that you get to play as in certain portions of the game.

If you haven't already bought this game, do yourself a favour if you're into mature action games and run to the store.

If you'd like to comment on this interview, please do so here.