[Papervision3D] Flash 10 Optimization idea: replace "precise" with Flash 10 3

[Ralph Hauwert]

HW accelerated by default : upscaling to fullscreen.
On Mac OS X a couple of things are OGL assisted during drawing, but  
that mearly seems to make up in speed what the platform was lacking a  
long long time (and still does).

The new WMODE'S like DIRECT and GPU enable other HW accelerated parts,  
but as noted by Adobe and Tinic Uro, this doesn't necessarily mean a  
speed increase.

Another interesting is threading.
Since Moviestar, drawing is threaded, the drawing area is split over  
the cores / cpus.
Pixelbender runs threaded, again, over vertical splits.

Anything AS doesn't run threaded.

Ralph.


On May 19, 2009, at 4:04 PM, Alan Pinstein wrote:

> Oh really? Is there a good list of what is hardware-accelerated? I  
> never could find one.
>
> Alan
>
> On May 19, 2009, at 9:40 AM, Tom Richardson wrote:
>
>> Flash 10's draw triangle is not hardware accellerated.
>>
>> On Mon, May 18, 2009 at 6:24 PM, Alan Pinstein <apinstein at mac.com>  
>> wrote:
>> Well the way I was using it was to just use it in drawTriangle() in  
>> BitmapMaterial. So I *think* that PV3d has already done the z- 
>> sorting at this point, meaning it wouldn't be an issue.
>>
>> The vectors are what I'm grappling with ATM myself.
>>
>> But since it's the actual rendering that takes ~75% CPU load, I am  
>> hoping the hardware acceleration of just this tiny bit could  
>> produce significant gains. Definitely worth trying.
>>
>> Alan
>>
>>
>> On May 18, 2009, at 6:40 PM, Jake Lewis wrote:
>>
>>> The problem with drawTriangles is that there is no z sorting of  
>>> the triangles.  For a simple scene made up of one simple (ie non- 
>>> concave) object this is fine ( ie it will work for your virtual  
>>> tour Alan) but a generic API such as PV3D will need to have the  
>>> triangles sorted.  The overhead of remarshelling all the vectors  
>>> could well negate the advantage of this approach.
>>>
>>> Jake
>>>
>>>
>>>
>>> On Mon, May 18, 2009 at 5:43 AM, Clox <ogggan at gmail.com> wrote:
>>>
>>> Alright, thanks =)
>>> I've looked into it a bit now, and I agree with what you should,  
>>> shouldn't
>>> be too hard to implement.
>>> But still, some 3D knowledge is definitely good to have.
>>>
>>> Hope you manage to get it working =)
>>>
>>> And like I said before, I'll help if there's anything I can do.
>>> --
>>> View this message in context: http://www.nabble.com/Flash-10-Optimization-idea%3A-replace-%22precise%22-with-Flash-10-3-tp23514763p23594273.html
>>> Sent from the Papervision3D mailing list archive at Nabble.com.
>>>
>>>
>>> _______________________________________________
>>> Papervision3D mailing list
>>> Papervision3D at osflash.org
>>> http://osflash.org/mailman/listinfo/papervision3d_osflash.org
>>>
>>> _______________________________________________
>>> Papervision3D mailing list
>>> Papervision3D at osflash.org
>>> http://osflash.org/mailman/listinfo/papervision3d_osflash.org
>>
>>
>> _______________________________________________
>> Papervision3D mailing list
>> Papervision3D at osflash.org
>> http://osflash.org/mailman/listinfo/papervision3d_osflash.org
>>
>>
>> _______________________________________________
>> Papervision3D mailing list
>> Papervision3D at osflash.org
>> http://osflash.org/mailman/listinfo/papervision3d_osflash.org
>
> _______________________________________________
> Papervision3D mailing list
> Papervision3D at osflash.org
> http://osflash.org/mailman/listinfo/papervision3d_osflash.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://osflash.org/pipermail/papervision3d_osflash.org/attachments/20090520/cb01ef6e/attachment.html>