Spanking new Particle Engine

#1 Before I get to it, let me describe briefly the structure of my particle engine:-

fxParticleReel
This defines the animation frames of any particle but does not contain any information on how it is used. As well as providing a mechanism to reuse particle animations, this construct also allows sorting by particle textures to reduce texture thrasing.

fxParticle
This defines many spatial (size, rotation, orientation) and environmental (gravity/friction/lights/fog) attributes of a particle, including a specific fxParticleReel to use. When the particle has been 'emitted', the attributes here define its' autonomous behaviour.

fxEmitter
This defines a method to emit particles. This is *not* linked to a specific fxParticle but describe a generalised method to emit particles and can generate *very* complex emissions from points, boxes, spheres, rings etc.

fxEmission
This combines a fxEmitter with a fxParticle to bring a specific emission into effect. From here, multipliers can scale various settings from within the particle/emitter objects to produce a unique emission. An emission controls overall aspects such as maximum particles allowed, sorting and culling.

fxEffect
This allows you to combine as many fxEmissions (or other fxEffects) into a single unit. Creating a realistic explosion can involve smoke, flames, sparks, shockwaves, some of which need time-delaying and interactions; all can be created from this one entity by combining various emissions/effects. The object can not only generate these combined objects but will emit them using a similar set of controls available with a fxEmitter. This allows you to generate *very* complex and dynamic effects with other effects/emissions moving in various directions.

>>> Continued on next post...

#2 As you may expect, these objects come with a wealth of attributes and so it's important to provide an intuitive method of generating them. This is something that has been paramount on my mind throughout my design and I think I've come up with a great way of meeting the problem. My GUI design allows attributes to be changed in realtime with results being immediately visible. Most attributes are editted over unit-time using a catmull-rom spline graph (as in Max Payne), not only for single scalars, ranges but also for things such as colours. The particle editor will compile these effects 'down' to scripts in the form of datablock definitions for the runtime system to use. These datablocks can then be editted manually if required.

Another very important concern is performance. Reading through the above descriptions, you can be forgiven for thinking, "great, but at what cost?". A valid question but like any system like this, scalability is critical in producing the correct effect for the target hardware. As well as supporting lesser-hardware and supporting many of the hardware accelerated features of modern cards, I've been researching various software methods in scaling systems such as these.

An approach that I've take is to allow critical parameters to have a normalised performance scaling factor. These control the scaling of each parameter in according with a system-wide scaling factor. For instance, setting a system-wide factor to 0.5 with any particular parameter having a local value of 0.9 results in a post-multiply of (1.0 - 0.45) 0.55 being made to that parameter. This can be applied to most attributes so that realistic scaling can take place e.g. max particles, LOD, size etc.

This (finally) leads me directly to the subject of this post. I am going to be writing the code to this after Christmas and I have a number of questions that I feel is important to ask at this juncture. The main reason for the post is that the development of such a system is no mean feat and therefore, unlike my previous code, I want to see if it has commercial potential for games developers within the TGE community. I have, of course, had a preliminary discussion with Tim/Jeff on this very subject but I would also like to hear your thoughts. The answers to these questions will drive the development of the GUI interface to a commercial product rather than an in-house dev-tool.

I would be most greatful if you would comment on some/all of the following questions:

- What features would you consider *essential*? (Please be specific as possible)
- What are your "would be nice ifs"?
- Would you prefer the product to be embeddable within the TGE or as a seperate app? (What are your reasons?)
- If it were to be commercial, would you pay no more than $5, $10, $15, $20? (Only if free? why?)
- Is the ability to build a library of effects and therefore reuse them a good idea to you?
- Would managing and compositing all of your games' effects and resources within the particle editor be useful?
- What do you think about purchasing TGE development tools like this?
- Any ideas on other dev-tools?
- What is the Matrix?


Thanks for your time reading this lengthy post and I look forward to your ideas....

- Melv.

#3 Holy guacamole! You've been busy. :-)

It sounds quite nice. Do you have any screenshots?

#4 "I want to see if it has commercial potential for games developers with the TGE community."

I had a thought in the shower this morning... Some of my thoughts from there are good... some are bad... I am too tired to filter this right now so I'll fire it off:

I would like to see some form of compensation for work I give back to the TGE Community... I have some real nuggets in store... But, I don't want to be held hostage providing support, fixing problems with HEAD, or really even answering basic questions. $20 doesn't even get your lawn mowed these days... though, some people will think they own you... Art Packs and Code Packs are distinctly different animals...

I have a few ideas:

1) Garage Games sanctioned contributions... People donate to a certain resource and there is a link to the people's names that donated with the monetary amount... if a couple hundred bucks was generated this way, it would be better than squat... Garage Games would see some of the donation... but not as much as PayPal :)

2) A licensing agreement that required the code be purchased for commercial use... that way it can be used by the community but if anyone actually sells a game... they have to buy a license...

3) Pay resources... no download access unless you pay...

Again, the main concern I have with this is being held hostage while making peanuts...

-J

#5 Hi Melv,

just a comment to this one

Quote:
- If it were to be commercial, would you pay no more than $5, $10, $15, $20? (Only if free? why?)

For free would be nice and anyone would be grateful, but hey pays this your bills ???
We also sell our content packs and doesn't give them away for free. It doesn't matters if you model some Art or code some functions, it takes time. Time you spend in your freetime, time you can't spend with your wife. So why not get paid for this ???

- Sven

#6 Absolutely!

I mean, roughly estimate the amount of hours you work on it.. quite a lot eh? Would definatly deserve payment..

Peoples argument would be, "But the whole engine is $100"

e.g $20 = 20% of the original engine cost.

But i know how much work goes into things like this, and other everyone should realise it.

I myself would easily pay $50 for something like this.

#7 Impressive as usual Melv.

Here's a few things that would be nice to see. Some mentioned and some can be done in the engine but it might be difficult. Some of these might also be better in a physics dev tool (which would also be nice unless you combine them anyway):

Shockwaves, beams (straight, curved, connecting 2+ points), timed multi-stage combinations, object "shielding" effect, object movement trails, constant gravity (repel/attract), Distance based effects, Water-waves/tides/whirlpools, Electrical/lightning effect not just environmental use, attaching any effect to objects, liquids (pouring water for example), multi-shape fields, directional volume effects like fire and water that don't kill frame rate, ability to move the effect (tornado for example), triggering attached effects on objects on/off. I think a lot of these might be possible with what you described but I'm not positive.

I'm sure there's a lot more but those are a few off the top of my head.

It seems like it would be best to have it merged with TGE to keep it updated but I can see having it as a seperate tool too. If you make it I'm sure it'll be worth $20. I wouldn't have a problem paying that much for an addition like this. Most packs are $10-20 so it's fair.

Having a library of effects would be useful.

#8 I would gladly roll out a bit of cash for something like this with no hesitation at all.

I would love to see it as an independant app away from the engine itself as that would give you greater flexibility. As long as it is easily ported into torque so it will work in game the seperate app is the way to go imho.

The aspect of being able to build a library of effects and being able to re-use them is awesome and to be able to build the effects outside the TGE is even better since you would have the flexibility of trying before you add it. Throw in the UI that you mentioned and I feel it would be a huge success.

Matrix.....hmm... Thats a place all of us computer junkies dream of but always wake up right before we get in...

#9 This sounds great! I'd pay for this - especially if a library of effects was included. $20 easily seems like a great price.

#10 Guys,

I intend to keep the actual particle processing abstracted away from Torque itself so if other opportunities present themselves then I could port it quickly. To save development time, I intend to use the TGE to create the editor; the only question being whether this comes in the form of a mod or not. Protecting the product may push me down the standalone app route but we'll have to see.

The proving ground for this will be Strategem and so you'll all get to see some of its capabilities sometime early next year.

@Sabrecyd: All the effects you mention here would be possible with my system.

I fully intend to keep everyone up to date with development as it progresses and it *must* progress as it is required for Strategem.

I don't apologise for considering the commercial potentials of such as system as it is a non-trivial endeavour for someone with a day job but I do want to mention that this doesn't mean that any development I do will be exclusively commercial from now on. Free FX for all!

- Melv.

#11 Holy carp (not a typo) Melv - that is awesome, awesome stuff.

I wanted to throw my hat into the ring and basically say HELL YES I'd pay $20 for this.

In fact, I'd pay more. You see, when I consider that the *average* content pack goes for $20 and then compare this with what you're proposing, I have consder the fact that there is a fundamental difference between them.

When I buy a content pack (haven't yet, but getting ready to do so very soon) I realize that I get what's there and nothing more. To add to this library of objects requires either personal effort on my part in creating additional assets, or additional expenditure in purchasing more assets.

What YOU are proposing is a toolset through which I can create and modify my own effects, thus creating a "library" if you will, of particle effects that I can call upon at will. Furthermore, because this is a tool, and not a content pack, it grows with me dynamically - everything I add to it is one more trick in the toolbox.

That said, I'd pay more for this than a content pack (meaning no offense to Bravetree or any other content creators, I already think you guys lowball yourselves, but am very thankful that you're willing to do so to make things affordable to the rest of us ;-).

One thing I'd like to add, is it would be nice to have a tool, perhaps linked to the engine like the show tool, that would allow visual particle editing and perhaps even a nice gui to facilitate it. Don't know if you've "been there done that" and wouldn't be surprised if you had.

Great stuff - looking forward to it.

#12 Yes, it does take allot of time to create content but its a mistake for programmers to think their time is more valuable than an artist's time. Fact of the matter is coders and artists need each other to complete a game. You guys might create the functions but it'd still look like crap if you didn't have artists to make it look pretty.

Having said that let me also say I'd be willing be pay between 20 dollars, same as the content packs from Bravetree. If you charge more than that then you create a money barrier where only a small number of developers can afford the resource, hence you've put a chink in Torque's armor. The Torque community is supposed to be about low budget developers supporting each other in creating games, not feeding off each other as a source of revenue. Making games in the manner we do and then trying to sell them on the open market is a crap shoot to begin with. You don't need to make things harder on developers by demanding large sum of money for the best resources.

Keep in mind what happened with WildTangent. They operated under the low budget/free banner during its initial stages. But as soon as they got all the free alpha testing they needed from developers they switched to their rediculous pricing license. What happened then? There was a mass exodus of developers who left in disgust. I think the same thing will happen here if everyone developing tools for Torque get's in their heads they can make tons of money by charging prices that are beyond the means of most of us.

#13 @Kirby: That's good to hear.

@N R Bharatae: Large sums of money? $5, $10, 20$! I don't think my time is more valuable than anyones. Getting some money back is not to put food on my table but to justify the huge effort required to make something like this, not only to myself but to my wife as well!! Over the next 6 months I guarantee that I'll be releasing plenty of free code that's taken nearly a year to develop ... free ... nothing ... nada. We all have to realise that this site is not about making games for free, it's about making games on a lower budget. $100 for the engine, $10 dollars for a particle engine, $20-40 for some content packs and you're sorted. That is a low budget.

I don't see the WildTangent thing happening here. I was one of those disgusted developers who left WildTangent; I was very active on the forums and produced a bucket-load of custom objects for the community absolutely free. I don't intend to change that here but there will be some projects that are non-trivial and need some other form of justification other than simply being a cool thing to do. Also, don't forget that the kind of monies that are being discussed here are not in the $100's of dollars but in the $10s. The price of a good meal (at least in the UK).

Anyway, I don't want this thread to turn into a monetry debate. I would rather hear what *features* people would want.

- Melv.

#14 How much people are willing to pay is probably more related to how much money they have. I would pay $50 for what you are offering.

Being able to make a library of effects would be of great value to me.

The GUI could be standalone and that would be fine with me. As long as I can work on an effect in the GUI edtior and then get the EXACT SAME effect in my TGE based game, which it seems clear, would be the case.

I also dont hold to the idea that coder's time is worth more than Artists. ( I split the "pie" of my game with *two* other artists even-steven. Meaning we each get a third...meaning I valued their time as exactly as valuble as mine). That being said I belive a tool like this has the potential to be MORE valuable than a content pack and therefore worth more money (to me).

You asked about managing all effects from such a system. Also not a bad idea.

You talked about performance and scaling and such. Is there any plans for supporting shader based effects for "high detail" modes?

#15 I remember you from the WildTangent boards, Melv. I also remember all the stuff ytou gave to that engine. I'll tell you I have the greatest respect for your ability. But I don't agree with you. If I followed your logic I would charge project leaders everytime I supplied a model or 2D graphic.

By large sums of money I meant the fifty to one hundred dollar range. As I said I'd be happy to pay 20 for your particle engine though I know its worth tons more. I'm simply concerned that coders will start to see dev tools for Torque as a means for making money and loosing sight of the whole reason Torque exists.

#16 $20 sounds about right if it comes with 5-10 sample/base/template effects to build off of.

My biggest concern is for the real-time editing. The faster I can get the core look of the effect down (sliders, wheels, color picker, etc.) the faster I can get to the time consuming job of tweaking it to look just right (playing with specific values). It needs to be easy to get in-game and ideally be editable in-game.

#17 If it looks good, and I think I'd have a use for it. I'd pay for it, I'd also say go for the $15-$25 price range.

#18 I think I mention this on Joshua Tread about Torque PowerTools.

I, me as myself speaking for me :). I would be willing to pay for any software, tweak, editor, or tool that will further help me develop the game of my choosing.

My running total so far is thus:

Torque: $100.00
Milkshape: $25.00
Cart. Shop: $65.00 (BobtheBuilder thks!! for the CSMtoDif)
PShop Pro: $99.00
VC++ 6.0 $99.00 (going have to upgrade this one heh)

Plus there is a certain City Builders content pack I been looking at :)

As you can tell most of the stuff is under $100.00 US dollars and well within range of the everyday person Like me :) who works a normal job. :) Not Blue or White collar just a regular job :))

Go ahead and attached a price tag to your services guys, I know you want to give back to the community. But time is the only commodity I just dont have a lot of so tools make it possible for me to produce stream-flow content creation.

Eventually any good investment pays off. Will it be a indie MMORPG game, will it be a great educational product. I dont know but I will get my money back :)))

So yes!! I would be willing to pay for something that can help me make a great game.

Edit: Effects I would like to see

Force Fields Ala Borg, Star Trek, I would like my summoner class in my throw up energy barriers that depend on the attack can deflect or absorb attacks. eg. Summoners cannot wear armor of any type. Essencers are the same. Swirling particle etc., that whirls around the characters.

Elemental Creatures compose of nothing but Particles would be cool for my Game: :))

Powering up effects: Summoners and Essencers must gain mana in battle. Fighters and Rangers drawing fighting spirit energy for thier attacks. (Think Dragon Ball Z..heh)

#19 Well said John.

#20 All the features I wished for above could come true?! Hell yes!

Can I place a pre-order? ;)

If this is made as a stand alone dev app, how will that work? I just mean, how would the particle/physics effect actually be added to the engine after it's created in the app? Maybe a dumb question, but I'm just curious.