Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

Dark Physics & Dark A.I. & Dark Dynamix / best fps performances = objects vs particles

Years of Service
User Offline
Joined: 26th Oct 2002
Location: Rama spaceship
Posted: 13th May 2013 11:06
Hi all again.
I am really inside a total war to obtain best fps performances.
With my best code optimization and reduction and
using the new video card I got so and so double fps but for my project is not yet sufficient although it is already very realistic.

Using few words my project point to have a airplane plunged into thousand of "air particles" to simulate the air flow in the airplane shape.
Currently my "air particles" are little DBPro boxes with their DarkDynamix box collision shape.

I am really curious of what could happen if I start to use darkDynamix particles instead darkDynamix objects.

So some question (sorry if you think these are really basic question but I totally out the particles matter):

1) Do you think using particles instead objects I could get more fps?

2) can the darkDynamix particles move the object that affect?
In other words, currently the "DBP objects air particles" can push and move the airplane, exactly as the air in a real airplane.
Can I obtain the same using darkDynamix particles?

3) can I turn off the particles gravity?

4) can I set the position of a particle as how I do using objects?
Years of Service
User Offline
Joined: 26th Oct 2002
Location: Rama spaceship
Posted: 13th May 2013 12:48
Sorry for MattyH, I forget to add the [dark dynamix] tab in the title..
Matty H
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 15th May 2013 00:23
I don't think particles can affect other actors, only works the other way around, I think.

Your project is very interesting but very cpu intensive. Not sure how you plan on simulating air? There are millions of air particles in just a small space, what kind of resolution are you aiming for? eg 1000 particles per square meter.

And do you have a way of only simulating particles near the plane, eg relevant to the simulation?

It's definitely not a normal problem and may take a custom solution.

You can position PhysX particles manually and you can turn off gravity, and they would be quicker to simulate.

Years of Service
User Offline
Joined: 26th Oct 2002
Location: Rama spaceship
Posted: 15th May 2013 11:38 Edited at: 15th May 2013 11:40
Hi Matty H, yeah I think too my cpu is suffering a lot..
And If the particles can't affect the "movement/rotation/inclination" of the airplane so I can't use particles system.

What I would and what I can about my project are 2 really far things but I try ALL to make these 2 things getting more near possible!

I tested a lot of way to make an airpane surfing in the air and I finish to choose to use 4 different "air particles" zones.

1) tail zone that influence the vertical drift, stabilizers and rudder (500 air particles)
2) Left Aileron zone (50 air particles)
3) Right Airleron zone (50 air particles)
4) Body & wings zone (600 air particles)

1200 air particles since now (If I want to mantain a minimum fps speed...). Really far from the real thing but already extremely realistic!!

I would to use just 1 zone for all the airplane but I am too far about performance to can do that.
If I using just 1 zone so I should to use at least 4 times more air particles to mantain the actual "resolution".

I always knew that I more good in ideas than in programming skill then I know that a project such this require a lot of skill especially using the appropriate programming language (C++).

DBPro + DarkDynamix (plus a my friend, captain of boeing 737 as technical consultant)permit a bad programmer as me to DO all that! but I must to face the damn performance issues.

I am near to change my CPU too (after chenged the video card), from 2 core 3.40ghz to 6 core 3.5ghz hoping I will get the right extra power just to can add some other air particles mantaining at least 25 frames x seconds.
Years of Service
User Offline
Joined: 16th Jun 2009
Location: Altis
Posted: 23rd Jun 2013 10:31
I don't think you could ever get an accurate simulation using physx boxes/particles, even with millions, physx is designed for "approximate" physics for real-time games. It would be better to use aerodynamics physics equations for something like this (that's how they would do it in flight simulator).

Login to post a reply

Server time is: 2023-03-30 19:13:02
Your offset time is: 2023-03-30 19:13:02