obviously these demos were done a while ago and do not take into account recent speed-saving features such as triangle caching and object culling. but for an indicator of raw speed, they're pretty good.
Rob On Mon, Feb 16, 2009 at 11:43 AM, Rob Bateman <[email protected]> wrote: > test of triangle limit can be found here: > > http://away.kiev.ua/away3d/techdemos/ > > look for "Triangle limit demos" about half way down the page > > cheers! > > Rob > > > On Sun, Feb 8, 2009 at 3:09 PM, Makc <[email protected]> wrote: > >> >> well, with this test people can check themselves how well their brand >> new ferrari pc or good old ford pc will ride your road, whilst now all >> you have to tell them, is that some guy in some car was known to ride >> at 3k poly per frame, which is pretty much nothing. any test with >> sufficient number of options that everybody can play with will do >> better than that. >> >> On Sun, Feb 8, 2009 at 1:16 AM, Fabrice <[email protected]> wrote: >> > >> > What do you mean by nothing? >> > I'm just pointing out that this kind of test can be misleading. >> > >> > Its like trying to find out howmany cars can ride and how fast on a >> given >> > highway. >> > A lot if they are all Ferraris riding same high speed, a lot less if you >> > drive in the Netherlands a rainy monday morning at 8:00 am. >> > In the end you know very little of the highway itself... >> > >> > Fabrice >> > >> > On Feb 7, 2009, at 11:14 PM, Makc wrote: >> > >> >> >> >> little is still better than nothing >> >> >> >> On Sat, Feb 7, 2009 at 5:23 PM, Fabrice <[email protected]> wrote: >> >>> >> >>> map sizes, tris size, mapping, material types, angles, stage.quality, >> >>> browsers, player versions... etc all are factors of influence. >> >>> I've seen scenes running slow under 500 tris... same for renderings >> with >> >>> just 250 tris drawn of a 60 000 tris model... >> >>> Its a nice test, but tells very little in the end. >> >>> >> >>> Fabrice >> >>> >> >>> >> >>> On Feb 7, 2009, at 4:02 PM, Sean McCracken wrote: >> >>> >> >>>> >> >>>> I'll be doing a test just like that. I'll post my findings Monday. >> >>>> >> >>>> Sean >> >>>> >> >>>> Sent from my iPhone >> >>>> >> >>>> On Feb 7, 2009, at 5:55 AM, Makc <[email protected]> wrote: >> >>>> >> >>>>> >> >>>>> Hum, does it mean 3001 triangle would be too much? Someone should do >> >>>>> online test app where user could specify fps + features used, and >> the >> >>>>> app would add triangles until fps goes too low. >> >>>>> >> >>>>> On Fri, Feb 6, 2009 at 3:53 PM, Li <[email protected]> wrote: >> >>>>>> >> >>>>>> I guess a quick answer would be around 3000 triangles with motion >> >>>>>> (someone >> >>>>>> correct me if Im wrong). However, Away3D uses something called >> >>>>>> triangle >> >>>>>> caching. The idea behind this is that if an object hasn't moved, >> and >> >>>>>> the >> >>>>>> camera hasn't moved, then the object would look the same, so why >> >>>>>> redraw >> >>>>>> it? >> >>>>>> Doing stuff like this allows for a very larger amount of triangles >> as >> >>>>>> in >> >>>>>> the >> >>>>>> Intel game demo in the away3d site which uses more than 100,000 >> >>>>>> triangles! >> >>>>>> >> >>> >> >>> >> > >> > >> > > > > -- > Rob Bateman > Flash Development & Consultancy > > [email protected] > www.infiniteturtles.co.uk > www.away3d.com > -- Rob Bateman Flash Development & Consultancy [email protected] www.infiniteturtles.co.uk www.away3d.com
