Hello all!
I wont to ask you one question:
On WebPlayer FullScreen (on Intel G41 Express) I was got 200 FPS (counted by FRAPS) in IE, FF, Chrome and scene is totally empty - clean project with only default camera.
On standalone which builded from the same project, I have got 450 Fps
I know this video card is not the best, but why so big difference between ActiveX and StandAlone? Are this normal?
FRAPS will kill your FPS, were you using that on your Standalone, too?
Put in the FPS script on an object and test it like that…
Fraps will not affect the framerate if not recording.
The difference is always that large, cause inside the player the plugin is bound to informations and other aspects of the browser where it lives inside, while the standalone is totally unbound and exists on itself.
Browsers have a sane throttle on plugin stuff too
Thank all, for the answers
FRAPS will kill your FPS, were you using that on your Standalone, too?
I was use it on WebPlayer and Standalone. As Hintze mentored, “FRAPs only affects your fps when you are actually recording video, when not recording it is pretty accurate and not taxing the system very much at all.” (С), today Fraps is fastest app to count FPS, I think they difference from real is not much than 2-4%, it is applicable for me.
Put in the FPS script on an object and test it like that.
All I will get with this is the rate which Unity sends data to the video card. This is good information, but not all information what I need in my research. I was need pure performance of my hardware at last point for compare several video cards.
The difference is always that large, cause inside the player the plugin is bound to informations and other aspects of the browser where it lives
inside, while the standalone is totally unbound and exists on itself
You right, dreamora, some piece of performance was payed when object was embeded to browser, but look in to the my research results, they clearly shows, that FPS is locked on ~60 FPS when Webplayer in browser windows and to 200FPS when Webplayer works on FullScreen.
Test plan
Test results
Nvidia 8400 GS
62 fps Window WebBuild
200 fps Fullscreen WebBuild
707 fps Window StandAlone
1124 fps Fullscreen StandAlone
triangles.exe 1288 fps Window (!!!)
triangles.exe 1920 fps Fullscreen (!!!)
Intel GMA 4500 (G41)
62 fps Window WebBuild
200 fps Fullscreen WebBuild
435 fps Window StandAlone
682 fps Fullscreen StandAlone
triangles.exe 846-952 fps Window (!!!)
triangles.exe 1024 fps Fullscreen (!!!)
So as you can see, Window WebBuild and Fullscreen WebBuild are locked near to 60 and 200 FPS, because this FPS is not depended from VideCard and browser name. This is strange for me, I wonder if some one knows how to turn off this lock.
And one more thing, simple Direct3D application triangles.exe shows 50% performance on similar empty scene, is there some trick to achieve similar performance on empty scene? In Fastest mode config I was turned off all what possible to speed up project 
P.S. hippocoder, I was not really understand what you were mean
Listen, plugins like flash and unity have a resources cap. So that adverts don’t hog 100% of the users CPU. This doesn’t mean unity is running slower, it means that unity is taking up less overall resources and won’t go too far over the max speed limit. Why are you worrying about this?
Interest opinion, I was know this before, but newer associated with Unity Webplayer thank you for this idea.
Why are you worrying about this?
I am only interesting why is so different 60 and 200 fps is normal for my simple project) Is this affects FixedUpdate (fixedTimestep) or not?
At least, I wonder if you know answer for other question, which culprit makes empty unity project about 2/3 slower than simple empty Direct3D application ( http://www.drunkenhyena.com/cgi-bin/dload_tracker.pl?file=/download/dx9/triangles.zip )?
On this screenshot setup of my test application exists: Clip2Net — screen capture tool for Windows, Android, iPad, Mac, Linux
Outside of having more bandwidth, I don’t understand what difference anything above 60fps makes?
It could just be because I come from a background of outputting to TVs from consoles (and back in the 8-bit Atari/C64 days, no one really cared about framerate), but post-60fps seems imperceptible.