Dolphin, the GameCube and Wii emulator - Forums

I'd like to start a new game benchmark. I'll provide the build with the settings preset. We just need to decide on the game. Since I'll be conducting the test, I'll decide which game we use.

The games I'd like to test with are:

1) Zelda Twilight Princess

2) Zelda Windwaker

3) Super Mario Sunshine

Most people should have these games, so cast your vote to the poll and I'll start the bench and provide the test build and the test area.

After one game has 10 votes, that will be the game that gets used. I'm sure my laptop will come in last with all of the fast Intel chips around, but it'll be fun anyway.

The game will be NTSC and the build will be a Windows x64 build.

CPU?
GPU?
Both?

I'm guessing both.

It's going to be an overall hardware test. I'll provide more information once we get the game nailed down.

I'll join :3

(09-18-2012, 08:24 AM)Starscream Wrote: [ -> ]The game will be NTSC

It excludes a lot of people

Yeah, but it would exclude me if it was PAL. I can't start a bench with a game region I don't have. :/

(09-18-2012, 09:05 AM)Starscream Wrote: [ -> ]Yeah, but it would exclude me if it was PAL. I can't start a bench with a game region I don't have. :/

Clearly, thats all that matters.

I have all three of those so it really makes no difference to me. I think that testing the system as a whole can be somewhat pointless though because you don't know the problem. I think load percentages could help if those could somehow be monitored and output as well. I don't do a lot of benchmarking so I wouldn't know. I'd be willing to do some testing with my MacBook and my desktop.

P.S. I was just messing with you on the first part.

---

Wait, no dangit I want to retract my vote for TP. WW is the better choice because TP may vary slightly between the GC and Wii versions, so it isn't a good pick.

You can measure the uncapped VPS instead of the gamespeed or convert the gamespeed. Either way you should be able to mix PAL and NTSC results.

My memory card save at the test point is going to be NTSC. If the guys using the PAL game are able to convert the save, I guess we can include both.

If you're going for extreme accuracy you can always separate the results into NTSC results and PAL results. That way everyone can participate and we can get as much data as possible.