(08-28-2013, 02:43 AM)u_liekmudkipz Wrote: [ -> ]吾'll post the benchmark of my 古ld and cranky desktop later... It's armed with a Core 2 Duo processor along with a crappy GPU.
I'm definitely interested in this. I just noticed that there's a lack of Core 2 Duo benchmark results and I wanted some. Luckily I may be able to provide just that in the near future...
Also, if the GPU is really crappy, you could set the internal resolution to 0.375x.
Here's the results of my elderly desktop...
Software
OS: Windows 7 x64 Ultimate
Dolphin Version: x64
Game Version: NTSC-U
Hardware
CPU: Intel Core 2 Duo E7500@2.93GHz (OC'ed to 3.53GHz)
GPU: OC'ed Nvidia GeForce 9400GT 1GB GDDR2 (Core: 680MHz / Memory: 454MHz / Shader: 1750MHz)
RAM: 2GB DDR2
Display: 1366x768 monitor
Results (FPS/VPS/SPEED):
Save 1: 52/104/173
Save 2: 44/87/145
Save 3: 66/132/220
Screenshots:
Save 1:
![[Image: ebu3bKE.png]](http://i.imgur.com/ebu3bKE.png)
Save 2:
![[Image: KqX0yYR.jpg]](http://i.imgur.com/KqX0yYR.jpg)
Save 3:
![[Image: Uxezrjt.png]](http://i.imgur.com/Uxezrjt.png)
Disabling the Antivirus and switching to basic theme gave me additional 3-5 FPS.
@Nintendo Maniac 64:
Looks like that I'm the 2nd C2D user here on the benchmarks LOL! Back in 2010 when I started to lurk the forums, I saw a lot of Core 2 Duo/Quad users on the old forums. And today, a lot of people here have switched to Sandy/Ivy Bridge chips.
I've noticed that in the first and the third tests fps can vary a lot. It depends on Beedle's shop. If it's on screen, it drops about 10 fps or even more. In my opinion the second test is more reliable, and that should be used in the results chart. In my results I've waited till that ship disappeared from screen.
Yeah, it's good that you did that. I've already stated that numerous times throughout the thread, but sadly not everyone complied to it.
Due to that there are definitely some inaccuracies in the result list. Give or take +- 10FPS from most results.
10!?!
More +/- 3 if you're being generous.
(08-29-2013, 12:02 AM)NaturalViolence Wrote: [ -> ]10!?!
Yes. First test:
Before -> 94-98fps
After -> 105fps
Of course, this can be proportional, so if someone has low fps anyway, the drop would be less noticeable.
I just ran the bench mark for save 1 on a stock 4770K, and I saw frame rates from 75 to 92. Of course 75 and 92 are the extremes, but depending on the position of the boat on its loop around the island, I would get a stable 89ish or a stable 77ish. Meaning the frame rate stayed the same for maybe 5 seconds or more and wasn't the result of a spike.
With the upcoming 4.0 release, it might be nice to make a benchmark that consists of a .dtm file that'll get everybody into the game at the same frame x. Then you could have Dolphin stop itself after y number of frames and log its FPS to a file. When it's done, strip out the data before frame x occurred, and average them all together. Probably just do this from the command line to make it easier.
I don't really know what obstacles would be in the way of that, but off the top of my head:
1) I don't know if running a dtm on multiple computers with different specs and IO speeds would result in the same emulation.
2) I don't think there's a way to launch a recording from the command line.
3) There isn't a command line option that'll stop the emulation after a certain number of frames, or any GUI option for that matter.
Maybe I'm just making things too complicated. Anyway, it's just a little idea I had.
I think it's a good idea. I'll I'm saying is don't expect it to completely stabilize fps.
Also the reason you experience such high fps variance is because your cpu is so god damn fast. The variance is about +/- 10% in most cases. Which for most of us means around +/- 6 fps (I miscalculated originally) at 200% and +/- 9 fps at 300%. But you're getting closer to 300% and 400% then 200% and 300%. So for you it's more like +/- 8 fps and +/- 12 fps. But it's still 10% just like everybody else.