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 / Question about Dark Physics software and hardware

Author
Message
huertaaj
17
Years of Service
User Offline
Joined: 3rd Dec 2006
Location:
Posted: 6th Dec 2006 16:13
I need to create some "INTERACTIVE" biology simulations for teaching my students at the university. I want to model things like gas molecules bouncing around in a box, membrane particles wiggling around, gas flow through pores, etc. Is this possible with Dark Physics? I can do this in 3D Studio Max 8 pretty well but the resulting 3D animation is not interactive, only viewable in a passive way. I would like to use DarkBasic and DarkPhysics to create 3D interactive materials for my courses. Is this possible with this type of software? Is it necessary for me to purchase extra hardware to get Dark Physics to work on my computer? Will my students need to purchase software or hardware to view the simulations on their computers?
Any help will be appreciated,
Huertaaj (total newbie on DarkBasic)
Bozzy
17
Years of Service
User Offline
Joined: 10th Sep 2006
Location: Birmingham, UK
Posted: 6th Dec 2006 18:38 Edited at: 6th Dec 2006 18:38
I don't know about how to do it, but if it is possible, you might need a phsyX card or whatever its called to do some of the stuff, im not sure though

I'm not very helpful am i lol

Bozzy
Argoon
17
Years of Service
User Offline
Joined: 16th Oct 2006
Location:
Posted: 6th Dec 2006 19:09
The only thing that i know and can say for now is that DP in is current version can only calculate liquid simulations(Gas,water etc) using the PhysX PP hardware, so yes for liquids your students will need to have a PPU, the rest can be made to work in software. Or you can wait for the new DP version using the new Ageia SDK that can do liquids in software but donĀ“t aspect a so good realtime performance.

Cheers
Hoozer
17
Years of Service
User Offline
Joined: 8th Sep 2006
Location: Bremerhaven (Germany)
Posted: 6th Dec 2006 20:17 Edited at: 6th Dec 2006 20:18
I would say it depends on how you want to simulate the objects!
If the "gas molecules bouncing around in a box"-thing is done with normal dynamic-rigid-body sphere and static-rigid-body-box objects, then I would say it should work fine, as long as you are not using 200 and more at once! (This number is just a guess, cause in my mini-game shown in my sig, it runs just acceptable with 120 objects, when the box-size 15 is selected! BUT I do not use the DP-collision-detection, maybe that's why it is not accelerated by a hardware-card!)

The "gas flow through pores"-thing might be realized with a particle-system and somes static-rigid-body-boxes with some holes in it (the holes are used as the "pores")! In the particle-demo from DP there are used 200 paricles at once in softwaremode inside a building that consist out of static-rigid-body-boxes, so yours would be just a "modified" version of that!

As "Argoon" said, actually DP needs a Hardware-Card for fluids! And some other things might be accelerated if one is present (although any test in the net I saw, there were no performance-boosts at all)!

Good luck with your stuff!


Hoozer

AMD 64X2 4800+ (939); 2GB; GF 6800LE (@12PS, 6VS; 380 MHz, RAM: 434 MHz)
DP-Sw-Mode-Comp-Entry (updated to V. 1.2):
http://rapidshare.com/files/3326313/PhysiX-Brick-Breakout---V--1-2.zip
VRMan3D
19
Years of Service
User Offline
Joined: 3rd Apr 2005
Location: New England
Posted: 8th Dec 2006 19:56
I would agree with Hoozer that yes, it's entirely possible and actually I think it would be quite simple to do with dramatic and very interactive results. And no for what you describe you shouldn't need hardware acceleration at all. The best thing (IMHO) about DP is that the performance is excellent even without the PPU card compared with other physics SDKs like Newton.

I'd be happy to help you if you'd like. I have created lots and lots of Dark Physics programs (that someday I will get around to finishing and releasing hehheh) and have used just about every command in DP extensively enough to find the bugs, etc. The project you describe would be a great use of DP, and I'd love to help and I'd love to let my (homeschooled) children check it out and learn from it, so even if you don't want my help perhaps when you're done would you mind sharing it with the forum here?

Best regards,
-=VRMan=-

World Famous 3D Screensavers
-- http://www.vrman3d.com --
huertaaj
17
Years of Service
User Offline
Joined: 3rd Dec 2006
Location:
Posted: 10th Dec 2006 15:34
Sounds good VRMan, please contact me about your proposal. Thanks to all for the good tips. So am I correct in understanding that if I purchase DarkPhysics, I can use it for my simulations without having to purchase any extra hardware and it will actually help my program to work better?

I already purchased the "Nuclear Glory" DLL for collisions. How does this relate to DarkPhysics? If I already have Nuclear Glory (which I have not installed yet), do I still need DarkPhysics?

Thanks,
Huertaaj
VRMan3D
19
Years of Service
User Offline
Joined: 3rd Apr 2005
Location: New England
Posted: 10th Dec 2006 17:50
Well it's up to you Huertaaj. Nuclear Glory and DarkPhysics do indeed do some of the same things, but DarkPhysics extends mere collision detection by simulating Physics interactions. With Nuclear Glory you could detect when your "gas molecules bouncing around in a box" hit the walls, but you would have to code the part of when they start spinning as a result of hitting at an angle. Basically DarkPhysics does a ton of the work for you so that you can just set up the objects, apply the initial forces and the physics engine will take care of making them bounce realistically off of the walls and each other. I think you would find DarkPhysics to be a huge order of magnitude easier to use than NGC and it would allow you to crank out your simulations much quicker than trying to do it yourself with Nuclear Glory.

Quote: "simulations without having to purchase any extra hardware and it will actually help my program to work better?"


Yep, no hardware required. If you'd like I could send you a sample program created with DarkPhysics so you could see how well it runs on your machine. I think seeing it in action would also help you understand the difference between it and NuclearGlory (or just watch the videos on the DarkPhysics website).

Of course another alternative is to use the free DBPro wrapper for the Newton Dynamics physics engine (sacrilige!) but I think the small price you pay for DarkPhysics is really worth it because:
1) It integrates with DBPro so well - one command to turn an object you loaded in with the DBPro Load Object command into a physical object (phy make rigid body dynamic box).
2) Newton performance can be really bad in comparison to the speedy DarkPhysics/Ageia PhysX system (without even using the hardware physics cards I mean which could potentially really increase your performance in certain situations, and Newton knows nothing about the hardware). (My Janger game went from an abysmal 7FPS to 500+FPS converting it from Newton to DarkPhysics for example)

Best regards,
-=VRMan=-

World Famous 3D Screensavers
-- http://www.vrman3d.com --
Hoozer
17
Years of Service
User Offline
Joined: 8th Sep 2006
Location: Bremerhaven (Germany)
Posted: 10th Dec 2006 22:29 Edited at: 10th Dec 2006 22:30
@ huertaaj:
Also DP is getting updated sometimes (I hope that will be soon)!
In the next update it might be possible to use fluids in softwaremode and the deformable objects become available! (Only when they convert/port/implement the AGEIA SDK 2.6.2 completely!)
That would offer you much more possibilities than any other physics-engine!

You can use your collision-DLL to start sounds or other effects when the collisions ocour! If you would use the DP-collisions for that, they allways check all objects and that's not too good for performance! (I never used them, but others told me they behave that way!) Maybe you should have also a look at the "DBP Collision DLL v2.0" from sparky, they are easy to use and very usefull and FREE!
http://forum.thegamecreators.com/?m=forum_view&t=74762&b=5

In the end it all depends on how do you think you want to make your simulation! (Use primitives or complex objects, particles or rigid-bodies, etc.!)


Good luck!

Hoozer

AMD 64X2 4800+ (939); 2GB; GF 6800LE (@12PS, 6VS; 380 MHz, RAM: 434 MHz)
DP-Sw-Mode-Comp-Entry (updated to V. 1.2):
http://rapidshare.com/files/3326313/PhysiX-Brick-Breakout---V--1-2.zip
Codger
21
Years of Service
User Offline
Joined: 23rd Nov 2002
Location:
Posted: 11th Dec 2006 07:34 Edited at: 11th Dec 2006 07:37


huertaaj
I was intrigued by you post and thought I would try my hand at "gas molecules bouncing around in a box" I would give the rest a go but ... I have no idea of the basic concepts

All the best

Codger





System
PIV 2.8 MZ 512 Mem
FX 5600 256 mem

Attachments

Login to view attachments
VRMan3D
19
Years of Service
User Offline
Joined: 3rd Apr 2005
Location: New England
Posted: 11th Dec 2006 10:35 Edited at: 11th Dec 2006 10:39
Hey Codger, well done, nice demo!

I took the liberty of modifying your nice demo a little to use a material with no friction and 100% restitution (bounces don't drain the velocity), and set damping to zero. This eliminates the need for the 'MaintainVelocity' and makes it just use the pure physics. (And I fixed the leaks in the box corners =)

I kind of did this for Alfredo's (huertaaj) benefit mostly since he's brand new to DarkPhysics.

Here's the modified by VRMan version:


And Alfredo, my latest email should be in your inbox, I look forward to hearing back from you tomorrow.

Best regards,
-=VRMan=-

World Famous 3D Screensavers
-- http://www.vrman3d.com --
huertaaj
17
Years of Service
User Offline
Joined: 3rd Dec 2006
Location:
Posted: 11th Dec 2006 15:21
Thanks guys!! I really appreciate your interest and support. Since I don't have DP yet, I can't try out your demos, but I'll purchase and install it today for sure. I can't wait to see how this looks!
Huertaaj
huertaaj
17
Years of Service
User Offline
Joined: 3rd Dec 2006
Location:
Posted: 11th Dec 2006 15:29
Hey, this is strange! I looked into purchasing DP and it says that I mujst have the full version of DBPro installed and it must be updated to version 6.2. I checked my version of DBPro which I just purchased last week and found that I have version 1.054. Can anyone explain what I need to do here?
Thanks,
Huertaaj
Hoozer
17
Years of Service
User Offline
Joined: 8th Sep 2006
Location: Bremerhaven (Germany)
Posted: 11th Dec 2006 18:11
@huertaaj

Here you can see the upgrades:
http://darkbasicpro.thegamecreators.com/?f=upgrades

Your version is "5.4", so it's the normal pack without any patches/upgrades!

When you have done all upgrades (the last upgrade/patch "6.3" should be enough), it should look like this: "1.063"

I hope it helps and that I'm not wrong!


Hoozer

AMD 64X2 4800+ (939); 2GB; GF 6800LE (@12PS, 6VS; 380 MHz, RAM: 434 MHz)
DP-Sw-Mode-Comp-Entry (updated to V. 1.2):
http://rapidshare.com/files/3326313/PhysiX-Brick-Breakout---V--1-2.zip
Codger
21
Years of Service
User Offline
Joined: 23rd Nov 2002
Location:
Posted: 12th Dec 2006 02:55
Some nice touches VRMan3D

In your code you stated
Quote: " Not sure why this line won't compile for me:
` phy set continous cd 1.0
` I thought it was supposed to be in the 1.01 update? Oh well hehheh.
"

The reason this does not work is that contiuous cd only works with cubes not spheres ( all TGC bullets need to be cubes seems strange but true?).
If you increase the thickness of glass you can reduce the tunneling effect

I never thought about modifying the material to remove loss yours is a much better solution

You modified the box building function thats cool but you changed it to a cube building function


huertaaj

I included the executable so you can try until your DP is ready see download button on my origional post


All the best

Codger

System
PIV 2.8 MZ 512 Mem
FX 5600 256 mem
VRMan3D
19
Years of Service
User Offline
Joined: 3rd Apr 2005
Location: New England
Posted: 12th Dec 2006 18:13
@Codger

Quote: " phy set continous cd 1.0
` I thought it was supposed to be in the 1.01 update? Oh well hehheh.
"
The reason this does not work is that contiuous cd only works with cubes not spheres ( all TGC bullets need to be cubes seems strange but true?)."


Actually yeah I remember reading about that, this is why the other command phy set rigid body ccd ID, 1 gives you the skeleton error message but what I mentioned above, phy set continous cd 1.0, actually won't compile for me. This is supposed to be a system wide CCD setup that affects everything and I've never been able to get that command to compile ("Could not understand command at line bleah"). Although I am running a specially modified version of the Dark Physics dlls sent to me by Mike, I am sure I was never able to get this command to compile. Does it compile for you? In other words I mean you can't even run the program with that line in it, not just that it generates a run-time error.

Quote: "I never thought about modifying the material to remove loss yours is a much better solution"

Glad you like it, I wasn't trying to be holier than thou or whatever, I figured you'd appreciate that change.

Quote: "You modified the box building function thats cool but you changed it to a cube building function"

Yes indeed you are correct. It doesn't do the same thing as yours at all since it's a cube but I figured I'd leave it in anyway since it fixes the leaks at the corners.

Huertaaj, I'll try adding to this message the compiled executable as well (Like codger did on his). You should be able to download and run it without owning DarkPhysics or DBPro since it's a standalone executable. He and I did this so you can try it out on your system before purchasing the tools to build the code. (the red 'Download' button in the lower right corner of the message)

Best regards,
-=VRMan=-

World Famous 3D Screensavers
-- http://www.vrman3d.com --

Attachments

Login to view attachments
Codger
21
Years of Service
User Offline
Joined: 23rd Nov 2002
Location:
Posted: 13th Dec 2006 04:45
VRMan3D
Quote: "phy set continous cd 1.0,"


I had used CCd earlier with Cubes and it did work but when I tested it by changing the Sphere to cubes on the molecules demo I recieved the same error ay you did. I guess we will have to wait for the next update

huertaaj
How about describing some of the properties of the biology simulations that you described perhaps a refernece or two and we can start a programming challenge

System
PIV 2.8 MZ 512 Mem
FX 5600 256 mem
Codger
21
Years of Service
User Offline
Joined: 23rd Nov 2002
Location:
Posted: 15th Dec 2006 03:47
huertaaj

I saw this on TV (Chanel 7 USA) it is an amazing animation that shows what is going on within a cell. On the telly it had a commentary that explained some of what was going on. My favorite part is the structure that show a "creature" pulling a large sack of debris along a "road" that was constructed just for that purpose out of thin air.

I dont expect to knock up something like this in DBP over the weekend



http://www.xvivo.net/press/harvard_university.htm

System
PIV 2.8 MZ 512 Mem
FX 5600 256 mem
huertaaj
17
Years of Service
User Offline
Joined: 3rd Dec 2006
Location:
Posted: 16th Dec 2006 22:08
Hi Codger:
That's a great series of animations. It was obviously modeled in a professional 3D application like Maya or Max. I also use Max 8 to make those kinds of 3D animations for teaching my own students many concepts about cell biology, but my animations tend to not look so polished. Mine are a bit more simplistic, mainly due to time constraints. My objective is to be able to make "interactive" animations to be use in the classroom and for the students to view on their computers. I have tried over many years to use VRML viewers and other tools here and there so that my students can view my animations in an interactive way, but none of those viewers/tools hang around very long and before I know it, they don't work or are not supported any more. I'm wondering how hard it would be to take a standard animation made in Max 8 (keeping it somewhat simple), and create a free-standing viewer myself. That's one of my goals with DBP.
Huertaaj
huertaaj
17
Years of Service
User Offline
Joined: 3rd Dec 2006
Location:
Posted: 16th Dec 2006 22:17
Hi all:
I'm wondering if I should start a new thread with a cell biology animation "challenge" instead of putting it here. I think I'll do that.
Also, I tried the .exe files provided by both Codger and by VRMAN3d. The one provided by Codger was good but had a large number of spheres that were bouncing straight up and down or horizontally, not true to the characteristics of "random thermal motion". The .exe file modified by VRMAN3D was much more realistic because it solved that problem. I'm wondering how hard it would be to make it so that the camera rotates around the center of the box and can zoom in or out, as if the camera were attached to the surface of a sphere viewing the box from any angle. The way the camera movement works now makes it difficult to see the box from all sides. Also, the camera should move faster for more interactivity.
Huertaaj
Codger
21
Years of Service
User Offline
Joined: 23rd Nov 2002
Location:
Posted: 18th Dec 2006 01:48
huertaaj
Quote: "I'm wondering if I should start a new thread with a cell biology animation "challenge" instead of putting it here. I think I'll do that. "


Sounds great I will try to participate


Quote: "I'm wondering how hard it would be to make it so that the camera rotates around the center of the box and can zoom in or out, as if the camera were attached to the surface of a sphere viewing the box from any angle. The way the camera movement works now makes it difficult to see the box from all sides. Also, the camera should move faster for more interactivity."


It is very easy to have the camera point to the center "Point Camera 0,0,0,0" will do the trick

As far as the camera moving faster this is easy to adjust by modifying the control camera parameters but I my machine it is almost too quick now.

VRMan3D
I noticed that your modification stop the spheres from rotating as they should when bouncing of other surfaces. The effect is easier to notice by applying a texture and removing the color object routine. it can be adjusted by setting the following parameter

phy set material dynamic friction NoLossMat, 0.9

however when this happens if two spheres collide they can stop

huertaaj what does happen when two say oxygen molecules collide ?


Codger

System
PIV 2.8 MZ 512 Mem
FX 5600 256 mem

Login to post a reply

Server time is: 2024-04-19 22:36:56
Your offset time is: 2024-04-19 22:36:56