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.

Work in Progress / Universe Generator Part 2

Author
Message
Fluffy Rabbit
User Banned
Posted: 17th Jul 2012 05:08
Hey, I see a half-sphere! That means you're just half a step away from releasing a demo!
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 17th Jul 2012 13:16 Edited at: 17th Jul 2012 13:25
Quote: "Hey, I see a half-sphere! That means you're just half a step away from releasing a demo! "


It's true in a way. The grid is probably the hardest part. It's stupid but the Z will not move to the front. I should have it fixed today, and then start on the rules. Thanks for comments.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 17th Jul 2012 15:55
Wow, I fixed it, and it looks perfect. That's supposed to be what Spacetime looks like without the bending. I think it's a bit creepy. But anyway, you decide what you think, only comment on the game.



Dimis
12
Years of Service
User Offline
Joined: 12th Jun 2011
Location: Athens, Hellas
Posted: 17th Jul 2012 16:03
Looks good, I don't think that it is creepy. Actually it reminds me of that magnet - toy thing that I saw on youtube, I think it is called NewCube or something like that.


TheComet
16
Years of Service
User Offline
Joined: 18th Oct 2007
Location: I`m under ur bridge eating ur goatz.
Posted: 17th Jul 2012 16:10
I know I've been harsh to you, and to be honest I still don't agree with what you're doing.

However, I don't want you to think that this project is useless, in fact, I really dearly hope that you will prove to me with this project, and others, that we're wrong. You have been considered crazy, but perhaps you're just becoming sane in an insane world~

Best of luck to you, and so far it's looking good, Pincho.

Just out of interest, how does that look performance wise, and how many objects are you planning on rendering?

TheComet

Zotoaster
19
Years of Service
User Offline
Joined: 20th Dec 2004
Location: Scotland
Posted: 17th Jul 2012 16:15
I agree, it's better to actually see something than to hear some theories without any proof.

As for your demo, are you using actual spheres? That'll make it very slow. It might be better to have a circle textured on a plane that's always facing the camera - it'll look the same but be a lot faster.

"everyone forgets a semi-colon sometimes." - Phaelax
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 17th Jul 2012 16:20 Edited at: 17th Jul 2012 16:30
At the moment I'm using sphere 6,6, which is a low poly sphere. But anyway, the grid is hidden, so the performance loss will be to do with using it as an array. And the array is at least 13 times bigger than the size of the chosen spacetime scale. But at least I'm not putting in Time, which would make a single particle use up all of the computer. I could do a demo of time though, just to show what it looks like.

That whole object on the screen scrolls around OK though if anyone is interested in the performance of DBPro. There are just 20000 sphere 6,6. I think I can handle about 60000. Once hidden, it can go higher I think, on a normal computer, and using an array.

Yodaman Jer
User Banned
Posted: 17th Jul 2012 17:34
That looks really cool! However you're achieving that, it looks like it could be used to create some types of solar system in a game someday.

Quote: "It might be better to have a circle textured on a plane that's always facing the camera - it'll look the same but be a lot faster"


I agree with this, as your project expands in order to get faster results it would be much more efficient to do it this way. But if you insist on using spheres you should look into Dark Occlusion, as it will only render spheres directly in front of the camera and not the ones behind anything, so that would help as well.

-Yodaman Jer

Not contributing much of anything useful to the forums since September of 2007.
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 17th Jul 2012 17:44 Edited at: 17th Jul 2012 17:52
I should make plains face the camera for the colours, and the sphere will be hidden. It also has the right physics to create demo's as well, so a bar magnet demo can be just plugged into it, and that would need some 3D I think. I've never really used plains much, so I don't know how well they will work.

Matty H
15
Years of Service
User Offline
Joined: 7th Oct 2008
Location: England
Posted: 18th Jul 2012 01:37
Rendering performance should not really be an issue with this type of project. I would think you would do thousands of loops for any one sync?

You are going to be seriously under powered no matter what computer you run on since you are attempting to simulate the universe

It's good to have something visual to represent your simulation but it's not absolutely necessary I would think.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 18th Jul 2012 02:13 Edited at: 18th Jul 2012 02:21
As an artist I just want to watch it, and see what it does. NASA save out Data on a Super-Computer, and watch the Data a few weeks later. I don't even have a Hard Drive big enough for that. but I've left it open ended, so you type in a number for how big you want it to be. Unlimited I suppose apart from DBPro limits. I'm only on scale 10.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 18th Jul 2012 13:29 Edited at: 18th Jul 2012 13:57
Just doing a few tests. I had a particle limit before, and now I decided on a border limit so that you always have a sphere. With a particle limit you could end up with a sphere with the end missing off it. This particle count seems too high, I'm not sure if my counter is screwed up I couldn't find a fault in it. This is the FPS maxed out at with a border of 100, which is scale 10...



So if I use plains, I suppose I could try to merge them together with alpha edges to get smooth transitions.

Seppuku Arts
Moderator
19
Years of Service
User Offline
Joined: 18th Aug 2004
Location: Cambridgeshire, England
Posted: 18th Jul 2012 14:14
Would trying Dark Occlusion help boost your frame rate? There's a lot of particles we can't see and DBP is most likely drawing them.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 18th Jul 2012 14:30 Edited at: 18th Jul 2012 14:56
So I tried plains and with transparency the FPS dropped to 3, and without transparency stays at 4. Strange, but I am drawing all of them behind the front ones. I'm not even rotating them yet. It looks very smooth though compared with none alpha.



Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 18th Jul 2012 14:35 Edited at: 18th Jul 2012 15:26
Quote: "Would trying Dark Occlusion help boost your frame rate? There's a lot of particles we can't see and DBP is most likely drawing them."


Dark Occlusion could be a solution.

EDIT: Oh the Demo of Dark Occlusion doesn't work on my computer. The buildings just vanish. I'll try upgrading my DB Pro, it's only on version 7.5

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 18th Jul 2012 20:43
Ok, so inside each array fits the zero particle which uses the rules...



Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 19th Jul 2012 14:16 Edited at: 19th Jul 2012 14:40
So I noticed that I had slightly lost the shape that I was aiming for. I had squares where there were supposed to be Hexagons. So I have just tweaked it a bit...



The aim is to get nature's shapes to self build. The blueprint that I use is a snowflake. If the model self builds a snowflake then it is the right shape. I made a snowflake test some years ago, and the shape can be built from the model that I have here.

Now all of that will be an array. It will be an array where each Dimension can rotate. That would be very hard to program. So by using models as an array you can spin each dimension, and still find out which dimension is next to which dimension. For example if a Z dimension orbits around a central position 180% it just moves 2 units back. I can still find out which unit it is next to to pass energy along the chain. This is a fast way to use images with no collision detection. Imagine trying to check for collision with 180000 objects. Quite slow. But I have put every particle in an array that can spin. And it will be faster than most other methods of collision detection. I am trying to use particles as particles and as little maths as possible.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 19th Jul 2012 21:46 Edited at: 19th Jul 2012 21:51
I have put a link in here, but I shall not comment on it. It has something to do with my model...

http://www.sciencedaily.com/releases/2012/07/120719132949.htm

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 20th Jul 2012 11:44 Edited at: 20th Jul 2012 11:49
Inside all of that goes the energy. And I will probably simulate time afterall, because it is needed in lo-res for certain Cause / Effects. Not sure what the gap is down the middle... We are now at the sort of interpreted, low, visual range. So this is the sort of thing that you will see most of the time on the screen...



Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 20th Jul 2012 12:51 Edited at: 20th Jul 2012 13:25
The gaps were because the energy was low. Once I fill particles with energy they become bloated with information.

The first image is just a 2d energy error, but it looked like Saturn so I kept the image. And the second image is full energy levels, like a leaf for example. And I wanted to check my particle counter, because it seemed to be jumping up too high, it should be 84502 + Energy, I have found the error.



Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 20th Jul 2012 16:46 Edited at: 20th Jul 2012 16:57
And now I have the rotations following the camera. Maybe they look better facing the middle? They are only being used as graphics so the rotations don't affect the physics.



Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 20th Jul 2012 21:13
Just another shape that popped up...



Fluffy Rabbit
User Banned
Posted: 22nd Jul 2012 00:09
I would suggest having the plains facing the camera, since that would mean they're visually closer to spheres. How long until you release a demo?
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 22nd Jul 2012 02:12 Edited at: 22nd Jul 2012 02:14
Quote: " I would suggest having the plains facing the camera, since that would mean they're visually closer to spheres. How long until you release a demo? "


I just have to make sure that I can track all of the particles, and I'm almost out of memory. Then I would like to hide the particles behind other particles. The last obstacle is deciding what happens when energy hits the edges of the array. Entropy can let it escape, I can't afford to let it escape, I need to maintain it so that I get faster results than normal. I could either use a gauss to rebound the energy back in a random way (which may actually exist), or allow it to escape but replace the particles with virtual particles, which fill gaps in the universe. There are a few choices.

A few days I suppose. I can't wait to see the full effect.

Latch
17
Years of Service
User Offline
Joined: 23rd Jul 2006
Location:
Posted: 24th Jul 2012 07:48
@Pincho

Kevin Picone created an object dynamic cache exmpale some time ago for DBC. I imagine it could be adapted for DBPro. Basically, it caches objects and only displays the ones you need or are visible. It may be useful to you. Here's a link:

object cache

Enjoy your day.
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 24th Jul 2012 15:24
Quote: "
@Pincho

Kevin Picone created an object dynamic cache exmpale some time ago for DBC. I imagine it could be adapted for DBPro. Basically, it caches objects and only displays the ones you need or are visible. It may be useful to you. Here's a link:"


It could be useful, but I get a broken link...
Quote: "Internal server error. Please contact system administrator."


Latch
17
Years of Service
User Offline
Joined: 23rd Jul 2006
Location:
Posted: 24th Jul 2012 22:05
Oh, sorry about that. Here's a download

Enjoy your day.

Attachments

Login to view attachments
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 24th Jul 2012 23:40
Quote: "Oh, sorry about that. Here's a download"


Oh thanks, I'll take a look.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 26th Jul 2012 17:17
My bathroom was being decorated for a few days, so I have just got back to this. I've managed to delete a lot of objects to get some memory back. I shall put them back as I bend spacetime which is faster than moving them, so that has worked out well. Now I have more memory for my array, and can get that working.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 26th Jul 2012 20:40 Edited at: 26th Jul 2012 21:14
Well I got two big Arrays in, and it didn't crash from lack of memory. But how to find out how much memory is left? System memory reads 2048, but I have 3072, so something is wrong there...




I suppose, if I could figure out how much memory is left I could make the program self configure for each computer.

Task Manager so far....149,984k It runs out at about 250,000k (DBPro crashes to Editor), so I'm not sure what's happening there.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 1st Aug 2012 21:46 Edited at: 1st Aug 2012 22:06
Just an update. I am currently working on a test to make sure that I do not lose grid positions as the particles rotate. In John Conway's Game Of Life the grid was a 2D static grid of X/Y coordinates. However, I am simulating Space Time as a grain structure which is a hypothetical scientific theory. My grain rotates, which would be like a golf ball full of dimples rotating, and you need to put a ball bearing in each dimple. That means that I have to track the dimples as they rotate so that I don't lose a neighbouring grid position. So imagine the Conway game where the grid is moving around like a puzzle, and the X/Y coordinates are all getting moved. It makes it harder to move the blocks around in code. In 3D it is harder still...



The test with the numbers above is coded so that as the bunch of particles rotate together, the numbers must remain in the same places, and not change at all. I can then use the numbers to point to the limb that they are next to. The limb will change, but the numbers will not. This is a situation created by our decimal system, and I am having to convert it to a sort of counting ring system. A counting ring seems to match the Universe better than decimal. I have to maintain this counting system for unlimited particles, so I am making sure that it works with this visual test.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 3rd Aug 2012 17:07 Edited at: 3rd Aug 2012 17:26
Well, I've got the numbers to be stationary now. The interesting thing about trying to copy the Universe is that you get some insight into the possible physics of unanswered questions. There have always been questions, and I feel that I can give new answers to questions that usually have no real answers. Like...

"Is the centre of rotation through particles stationary?"

I get an answer. I shall not comment on it. It might be deemed as trolling.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 16th Aug 2012 10:55
Just an update There should be a new video soon, and I have a musician making a fantastic soundtrack.

Fluffy Rabbit
User Banned
Posted: 16th Aug 2012 12:01
@Pincho-

Is it really necessary to promote the potential key to everything with music?
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 16th Aug 2012 13:15
Quote: "Is it really necessary to promote the potential key to everything with music? "


I think that a silent video is a bit boring.

Fluffy Rabbit
User Banned
Posted: 16th Aug 2012 13:54
Quote: "I think that a silent video is a bit boring."


Surely, your program can simulate sound waves.
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 16th Aug 2012 14:07 Edited at: 16th Aug 2012 14:08
Quote: "Surely, your program can simulate sound waves. "


Maybe, but the computer can't pick them up. Anyway I can't produce enough particles for that.

Fluffy Rabbit
User Banned
Posted: 16th Aug 2012 14:10
@Pincho- Then maybe you could make it go "BOOM" whenever there is a reaction over a certain magnitude.
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 16th Aug 2012 14:32 Edited at: 16th Aug 2012 14:34
I don't think that test 2 will be at the scale of a boom. Most likely some swirly patterns. Maybe when I get to about test 10, I will get some idea of the physics.

Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 17th Aug 2012 02:33
This Video demonstrates a 3D rotating Array that maintains a local coordinate system. Although the object on the screen is rotating in all 3 dimensions, it's separate parts are still maintaining an ordered structure. The numbers of each array are monitoring the rotation of each part. This allows for fast object tracking, and collision.

http://www.youtube.com/watch?v=gd8RAQLTb7Q&feature=youtu.be

Yodaman Jer
User Banned
Posted: 17th Aug 2012 06:20
Music brightens up any kind of video. Looking forward to seeing it Pincho!

-Yodaman Jer

Not contributing much of anything useful to the forums since September of 2007.
Pincho Paxton
21
Years of Service
User Offline
Joined: 8th Dec 2002
Location:
Posted: 17th Aug 2012 12:49
Thanks! I've just heard a sample of the music, and it's more than awesome, it's double awesome!

Login to post a reply

Server time is: 2024-03-28 19:45:58
Your offset time is: 2024-03-28 19:45:58