Gfx Benchmarking: Flash/Flex vs. WPF vs. WPF/E vs DHTML

« Adobe Seminar ("3 Briefs") Feb 07 - Source code

Update: Flash 9 Experiment - Fireworks Alpha Mask + Webcam »

Feb 28

Gfx Benchmarking: Flash/Flex vs. WPF vs. WPF/E vs DHTML

Flash, General, Silverlight/WPFe, WPF, Flex Add comments

I'm afraid to say I got different [relative] results to those published on the site (the Flex cache as bitmap version was by far the highest FPS on my machine), but I'd have to dig into the source to see how these have been put together...

Link to the examples

Link to the results and discussion.

Some personal notes. I tend to think a Flash only (non-Flex framework) example would be a better option to a Flex app. Also the Flex example uses a Timer, he probably should try making use of the ENTER_FRAME event to see what results that gets. Also worth trying the fullscreen-then-back trick the Papervision3D guys are using to get a perf boost. Any other tests out there? Something like this would be interesting

Overall though the performance is great on all the examples so things are heating up

7 comments

§ Alexey Gavrilov said on : 28/02/07 @ 08:09
DHTML version does uses transparent PNGs -- otherwise you would see balls corners obscuring each other :-)

Looking forward for your results and thoughts!

Alexey
§ Richard Leggett® said on : 28/02/07 @ 08:11
Hi Alexey, I was seeing things, I thought the WPF/E and Flex versions had 50% opaque fills in the center, my bad :)

I've put my results on the original blog entry.
§ oz said on : 28/02/07 @ 08:42
Richard, could you check the links on the examples page. I'm getting framerates in the mid 70s for the Flash/Flex version and framerates in the low 60s for the casheAsBitmap version.
§ Richard Leggett® said on : 28/02/07 @ 08:47
Hey Dan,

That's extremely strange. My cacheAsBitmap results are much higher than the regular ones.
§ oz said on : 01/03/07 @ 04:32
Okay, tested again just to be doubly sure and got the same thing. I'm using FF 2.02, FlashPlayer 9,0,28,0. I think we'll just have to chalk this up to quantum interference or dark matter interactions. :-)
§ Richard Leggett® said on : 01/03/07 @ 06:10
I'm sure we spoke about putting a stop to your dark matter experiments!
§ oz said on : 02/03/07 @ 09:01
Did I write "Dark Matter"? No, no, no, that was a typeo. What I meant to write was "Bark Matter". I'm totally off the Dark Matter experiments. Cold turkey.