LE 3 speed :

[YouGroove]

Well about speed, LE 2 is not known as a fast 3D engine, you must have a good hardware.

 

That's what i seen just playing some little games demos from the forums, even a little arcade game asks so much power for a poor scene and simple characters.

By turning off shadows and full screen effects it was more playable on my laptop.

In this lap top i just can run games like Dragon Age 2 by putting some settings down , but it is playable with as many or even more frames than the LE2 platform demo game or FPS demo with effects on.

 

So i have a question ;

Will LE 3 have a re write of the core engine ?

Will it be optimized ? Faster than LE 2 for some same game for example ?

Will it have good scene optimization options and features ? (shadows appearing with some distance clipping like some AAA games, technique that existes on old XBOX 1 games, progressive mip mapping etc .... )

Will it have some variable quality shaders (good quality, super quality ?

 

Shadows and full screen shaders are what can put down all FPS in games,

So will it be more optimisations in the future towards shadows and full screen effects ?

like

- very low res shadows , medium, high with different parameters of maps

- shadows dynamic updated only each 5 or 10 frames

 

 

For mobile i think it will have to be as good or close in speed to other 3D software having mobile support to be a good product, so im' not worried in that mobile area.

[Josh]

Okay, Leadwerks Engine 2 is possibly the fastest engine in the world. Other engines that are "optimized" have a fast framerate because they're not doing anything. I remember when Max3D was released, and all the Blitz people were telling me I didn't know what I was doing, and someone was going to show me how to make a deferred renderer that ran fast on **** hardware. Then they were shocked when it ran slower than Leadwerks.

 

That said, the PC market has fallen apart and hardware specs today are worse than they were three years ago. The default renderer in Leadwerks 3 uses forward rendering and doesn't use a uniform dynamic lighting, so it will have a fast framerate on ****ty machines.

 

In another six months the iPad might have a better GPU than the average garbage they're putting in PCs nowadays. All these consumers who keep spending their money on disposable craptops need to just buy a tablet they can drool on and wipe their fingers all over, they'll be better off and not be able to hurt themselves.

 

</nocoffee>

[Guest Red Ocktober]

"Leadwerks Engine 2 is possibly the fastest engine in the world. "

 

WWWWOOOOOAAAHHHH... wait a minute... let's see now...

 

Since my recent computer disasters i've been running everything on an antiquated Dell

Optiplex 270, with an equally antiquated geforce fx6200...

(laughter will not be tolerated!!! sympathy will be appreciated)

 

Doom3 runs nice and smooth... at pretty high settings,

 

Crysis demo runs smooth and fast... middle to high seetings...

 

Leadwerks editor/and/or compiled programs run like a slide show...

 

i must be confused... or somethin'

 

--Mike

[YouGroove]

Nop the game i tested on my laptop Dragon Age 2 , i pushed a bunch of parameters down, but it is at final playable with a complex scene (but with LOD) , characters (but with LOD), UI particles , some basic shaders all running.

And sorry but some Leadwerks game demos like the platform demo game was slow.

It was just a test i shared even if it didn't made shine LE engine , and i can just say i was just surprised as you after lauching it.

You should be realistic : I just put what i tested in my laptop, that's all.

What can i say : "Yes the platform game with effects on was fast"

Just to please you ? What will you gain ?

 

 

You should make some FPS tests on a non powerfull laptop ... but perhaps my laptop is too slow or i know it's graphic ship is not the best and i must push down lot of parameters down for games ?

So perhas the problem is lot more relate to it caus it's not really "Shader powerfull" ?

And this test is not to take account ? Well let's forget it so ... we will gain nothing, what matters is LE3 not

some LE 2 test.

 

And i wait LE3 for Android first, then PC in second position, os there is no worry

[Josh]

"Leadwerks Engine 2 is possibly the fastest engine in the world. "

 

WWWWOOOOOAAAHHHH... wait a minute... let's see now...

 

Since my recent computer disasters i've been running everything on an antiquated Dell

Optiplex 270, with an equally antiquated geforce fx6200...

(laughter will not be tolerated!!! sympathy will be appreciated)

 

Doom3 runs nice and smooth... at pretty high settings,

 

Crysis demo runs smooth and fast... middle to high seetings...

 

Leadwerks editor/and/or compiled programs run like a slide show...

 

i must be confused... or somethin'

 

--Mike

Leadwerks Engine 2 can handle any number of lights all casting dynamic shadows. Try doing this with any other engine and watch the framerate drop to 0:

 

It's awesome technology, and if it weren't for the consoles and Intel graphics chips holding everything back, the whole industry would be using this.

 

I don't know why consumers seem to have a fetish for low-performance hardware, but it's forcing me to adjust and build software for inferior machines. But we're going to get back to good graphics once Leadwerks 3 has been out for a while.

[Pancakes]

Josh has a point in that, on my GTX 275 Leadwerks was pretty slow. I was grateful to receive a 20fps once I added a few models and lights. But ever since I upgraded to GTX670 the dam thing refuses to slow down. I think Leadwerks is highly scaleable. Manipulating the editor feels much more like a "sandbox" because u just say I want a light here and u put it there. It takes power to simulate the physical world in that way. Everything in Unreal is faked and that's why it runs so fast and that's why AAA companies continue to use it.

 

If you haven't seen Leadwerks run on a GTX670 or above, I recommend u try it. You will want to slap yourself depending on how far of an upgrade you are making from what you had before. The GTX670 is literally 3x more powerful than my old GTX275, which in its own right was never a slouch to begin with. I get triple the performance in Leadwerks now. Where it would once sit at 20fps and refuses to drop below 60.

[Pixel Perfect]

Okay, Leadwerks Engine 2 is possibly the fastest engine in the world.

As much as I like LE2 this quote made me feel like I was in an episode of the Twilight Zone! Whilst I would never argue that LE2 is not a great engine do you have any real bench mark tests to back this statement up? LE2 is certainly not the only engine out there with deferred rendering.

 

That said, the PC market has fallen apart and hardware specs today are worse than they were three years ago.

The best hardware specs today for PCs far exceed anything in the past, this is again a completely meaningless statement that you keep regurgitating! The cheapest PC specs are lower than some of those available 3 years ago but this is simply in order to compete with all the cheaper mobile type devices that it finds itself in competition with and is simply driven by market forces. You get what you pay for, nothing has changed in that respect!

[Josh]

As much as I like LE2 this quote made me feel like I was in an episode of the Twilight Zone! Whilst I would never argue that LE2 is not a great engine do you have any real bench mark tests to back this statement up? LE2 is certainly not the only engine out there with deferred rendering.

I already posted my benchmark above. You can see the framerate and the number of lights.

 

 

The best hardware specs today for PCs far exceed anything in the past, this is again a completely meaningless statement that you keep regurgitating! The cheapest PC specs are lower than some of those available 3 years ago but this is simply in order to compete with all the cheaper mobile type devices that it finds itself in competition with and is simply driven by market forces. You get what you pay for, nothing has changed in that respect!

The PC market is dictated by the worst available hardware. People don't understand hardware specs but they do understand price. If you give them a low-priced crippled option, they will take it. They are trusting the vendor to give them that should be good enough to do anything they would want to do on the platform.

 

Most people don't need to graphics capabilities of an iPhone, but there's only one. If Apple offered a low-cost crippled iPhone, everyone would buy that one, and the 3D gaming scene on iOS would die overnight. Imagine if every single PC being sold had a discrete GPU on par with a GEForce 8600. PC gaming would be completely different.

 

Because the PC has no competition, so rather than improving hardware specs they've decided on a strategy of market segmentation. The idea is some people have $300 to spend, some people have $600, and some people have $1000, and you don't want to leave any money on the table. This is great for selling things like refrigerators, but when the hardware dictates what the software can do it's suicide. We're seeing the result of that now, as PCs are struggling to compete against mobile devices, of all things! We're right at an inflection point right now where PCs and consoles have lapsed so badly they are about to be eclipsed by mobile devices, even on graphics!

 

Look at this:

http://apc.io/products/

 

][/b]For openers, we started with an awareness that the purpose of a computer is to connect to the Internet. It is the Internet that now defines computing. When you begin here, magic happens.

 

Expensive, overpowered CPUs and bloated software are no longer relevant. With this awareness, we were able to drop power consumption to the point of making an energy-saving light-bulb jealous.

They're basically taking a shot at .NET managed code and laptops with expensive Intel CPUs that are just Facebook machines. ARM CPUs running simple native code apps are faster than x86 CPUs running bloated managed code. I don't know if this particular brand will take off, but this would have been absolutely unthinkable a few years ago.

 

Two huge things are playing out that began about eight years ago:

 

1. Microsoft never came up with a safe way for the user to install programs. As a result of this, everything that could migrate to the web did. The web is an awful platform for applications, but at least the user isn't scared to use it.

 

2. The PC market followed a strategy of segmentation, selling a lot of underpowered machines. This further damaged the market for desktop applications and games in general. Consumers didn't choose this because they wanted a ****ty computer, they chose it because it was a computer and cost less money than the other one.

 

On top of that, Microsoft announced to the world with Windows 8 that the last 20 years of interface design were all a mistake, and the iPad way really is better. What was a slow leak has now become a gushing hole. We're in the post-Wintel world right now, and it began October 26, 2012. Even Valve knows this.

 

We're essentially entering a complete reboot of computing. We're moving from a platform with no competition (x86) to a platform with LOTS of competition (ARM). This is why mobile moves so quickly...the market forces induce rapid improvement. Imagine if Android didn't have iOS to keep them on their toes. That's basically what the PC market has been like for ten years.

 

You can take advantage of this trend if you anticipate it. I don't claim to know what the outcome will look like, but there are a few ways you can prepare:

• Stay close to the metal. Native code provides speed, flexibility, and is the most future-proof. We can count on C++ and Lua continuing to work in this new world, whatever it looks like.
  
• Avoid getting locked into proprietary walled gardens. Don't get into systems that try to take over your life like .NET.
  
• Learn cross-platform open source technologies like GLSL, OpenGL, OpenAL, etc. This is why I make sure any third-party code we use is open source and free. I would not recommend anyone spend time learning DirectX, PhysX, or even Objective-C right now.
  

[Pixel Perfect]

Credible benchmark tests require stated hardware and operating environment and comparisons for a range of competing products. I don't see any of that here!

 

People don't understand hardware specs but they do understand price. If you give them a low-priced crippled option, they will take it.

You seem to have a very low opinion of people Josh if you sincerely believe this. Is this the current level of end user you are gearing LE3 to? I don't know of any PC gamers who buy such PCs even though they are available. The market was created for people who want a cheaper PC for primarily browsing the Internet, as I'm sure you are well aware, and they serve that purpose well.

[Josh]

Credible benchmark tests require stated hardware and operating environment and comparisons for a range of competing products. I don't see any of that here!

Why would I waste time on all that? I already showed you something I am not aware that any other engine can do. I'm not going around claiming with 100% accuracy there is nothing that can match it in any test, but calling LE2 slow for what it renders on screen is ridiculous.

 

You seem to have a very low opinion of people Josh if you sincerely believe this. Is this the current level of end user you are gearing LE3 to? The market was created for people who want a cheaper PC for primarily browsing the Internet, as I'm sure you are well aware, and they serve that purpose well.

Our software is designed for my user's users. They're the ones that buy games. I've been hearing second-hand complaints for four years that it won't run on a low-end laptop.

 

I don't know of any PC gamers who buy such PCs even though they are available.

In real life, I don't know a single person who is exclusively (or even primarily) a PC gamer. I am looked at as somewhat of an oddity for the fact I only play on PCs. Indies are shut out of the console market, but mobile gives us a way back into the mass market.

 

Deferred rendering will make a comeback in the OpenGL4 renderer, but the reality of the amount of **** hardware people have (game players, not my users) is forcing me to take a giant detour until I can do it.

[Flexman]

I've certainly had my fair share of performance issues with LE2. But experience proved (and a big thanks to my pal DaveH for putting in many hours of experimentation) that you need to be strategic with how you use the engine. In that sense it's no different from many other engines.

 

By default most scenes, even near empty ones will chew the hell out of the fill rate because of all the buffers for the different worlds it's creating and updating in the framework. It's all too easy to use models that suck the performance out. Create minimal physics hull shapes. Tweak the LOD ranges and don't have shadows active on everything. Also improper use of shaders in materials, using texture channels that are not initialised, so on and so on.

 

Spraying grass and trees everywhere isn't going to win any performance awards. If you look at vegetation coverage in some games, take Apache Air Assault for example, bushes and grass can share the same scale as a house.

 

Games cheat like crazy for performance, performance testing during development is designed to focus on that, it's not something your average tinker developer will think about much unless they have to.

 

For what you get with Leadwerks 2, it's pretty OK for bangs per buck but you have to work at it to tweak performance. Some of the stuff we did just for the helicopter cockpit in our game required *several* re-organisations of the model hierarchy so that it worked in favour of how the engine culled objects. This is what games do.

[Pixel Perfect]

I'd certainly agree with that Flexman. A balanced overview and sound advice!

 

Why would I waste time on all that? I already showed you something I am not aware that any other engine can do. I'm not going around claiming with 100% accuracy there is nothing that can match it in any test, but calling LE2 slow for what it renders on screen is ridiculous.

Well maybe to back up the statement you previously made. Most decent engines can probably demonstrate something no other engine can do. I'm not knocking LE2 as I went out of my way to point out, just challenging the statement you made. If you don't want to be challenged don't make the statements! Further to that, I don't believe I ever claimed that Leadwerks was slow for what it renders on screen any more than I claimed it was probably the fastest engine in the world.

 

People make ridiculous demands upon hardware all the time Josh, this is not unique to Leadwerks. It's more a reflection of their stupidity than anything else.

 

I don't believe anyone mentioned that PC Gamers had to be exclusively PC only users! But by definition a PC Gamer requires the use of a PC does it not.

 

The simple fact is, game designers Indie or not have always tailored their games to match the hardware its intended to run on or they very quickly go out of business. So the whole argument about hardware is irrelevant, as a developer you simply chose which marketplace you are aiming for and design for that marketplace.

 

The mobile market has had **** video hardware for years despite it changing now, but that didn't stop people developing games for it. In fact most platforms are overrun with games.

[Josh]

I get fired up sometimes,but this was a really interesting discussion.

[Rick]

Even Valve knows this

 

The only thing Valve cares about in all of this is to stop PC OS marketplaces in their tracks because if we can sell our games through a Windows OS marketplace there is no need for Valve anymore.

 

Don't get into systems that try to take over your life like .NET.

 

The alternative WAS extremely painful, and some can still say somewhat painful still when compared to the ease of use and massive feature set in .NET.

[DigitalHax]

This is proving to be quite the informative thread! Certainly worth the read through!

[YouGroove]

Well i know C++ si the fastest, that' why i plan to use it.

For LUA and LE3 i'll see how it works how variables are shared etc ... if it comes similar to some engines allowing to make games only by script ?

But C# is not so bad, and allow to make games , it is used in some engines even open source and until now it works, even if you could be faster with C++.

 

Everything in Unreal is faked and that's why it runs so fast and that's why AAA companies continue to use it.

 

Well, this is lot more optimisation and shaders i don't think they are so faked, and if some engine could allow to produce great visual quality games that would run very fast, i would not matter if something would be faked or not.

 

I played Guild Wars 1 , and the LOD were so faked if we can say that, some model buildings in the far distance was onlyt a bitmap and not so well oriented if you took your time to look at it, as you come near it switched from the bitmap to some low poly modelsLOD polygon models.

I prefer a game with lightmapping instead of real lights, if the result is the same and if i can gain lot of frames.

What matters is the result and not what is coded behind caus the player even casual will never see your code or what technique you used.

[Flexman]

I think the point is,it's not the engine 'faking' so much as the talent and technical skills using assets in a way to create lush visuals to fool the eye. The black art of visual game development. Some of these guys keep their secrets close to their chest. Sometime the engine adds a button to do it, "Generate lightmaps" in Unity comes to mind. One-click scene improvement.

 

And now everyone and their dog are doing stuff like baked ambient occlusion to create extra depth.

[YouGroove]

Ambiant occlusion is standard nowadays, i dont think there is so much engines without SSAO, specially with computer power increase (you have the choice to turn it on or off usually).

 

I remember some article talking about a DreamCast console game where it was hard coded the visibility of sections of the level

[Daimour]

So we can conclude that LE2 is very fast engine but nobody (except couple of people and may be Josh) can use it properly to rich decent FPS and nice picture. It's very very tricky, I know.

 

That's why we can't see tons of games created with LE2. So will this situation repeat with LE3? May be to have a good engine is not enough?

 

Josh, did you consider ways to teach people use your engine(s) properly? And teach them to make decent games with engine? I know, you are a programmer and not a teacher. But think about it.

 

You teach people -> They make many decent games with LE -> More people see it -> More people buy LE

[Rick]

It's only "tricky" if you are trying to make some game on par with AAA studio. Most indie games could easily be made and it wouldn't be "tricky" with LE. The fact of the matter is most devs don't finish their projects. For a game engine it's a numbers game. You just have to get a very large number of devs so that the small % of them that will actually ever finish a game will do it with your engine. The lower dev count you have the lower % that will finish a game (and that can be 0% with a low enough number of devs). I don't think LE even comes close to user count to something like Unity or UDK.

[Pixel Perfect]

I... The fact of the matter is most devs don't finish their projects. For a game engine it's a numbers game. You just have to get a very large number of devs so that the small % of them that will actually ever finish a game will do it with your engine. ...

I would completely agree with this. Even a fairly simple game engine is beyond the capabilities of most would be game designers if they are having to code it themselves and have no prior experience of writing a game engine. Whilst LE2 has some issues (what engine doesn't) I believe it is perfectly capable of forming the basis of a viable game engine and has been for a while. It is a numbers game as Rick says!

[fallingbrickwork]

That's why we can't see tons of games created with LE2. So will this situation repeat with LE3?

 

I'm sorry to say but i think the situation will be exactly the same for LE3 for the reasons stated in replies above.

 

I remember a thread that started a while ago titled, "What are you going to make with LE3?" (or a similar title). I think someone should start a thread titled, "What have you finished with LE2 in the x number of years it's been out?" but I don't think there would be many posts. The nature of the renderer pulled people to LE who had the romantic notion of making a AAA game or similar; and for obvious reasons - the realtime shadows etc. For small teams and hobby users everyone knows this is just what it is, a romantic bedroom idea.

 

Now, Josh is trying to buck this trend by opening up the mobile market, which should increase the user base; or will it?! I don't think it will (at all). infact, even a percentage of his current 'loyal' userbase won't upgrade because the renderer isn't up to that of LE2. I think the difficulty for LE moving forward will be to find it's place. It is a great engine, no doubt about it, but.... whereas once it was the only affordable engine with great visuals, it no longer is. It is now trying to enter an already saturated mobile market, where it will struggle to compete.

 

On the question of speed etc... i agree that the engine will look nice on a GTX680, but if you took those system specs to a publisher, they would probably laugh at you and say, "Pass!".

[Rick]

I think Josh needs to pay some of us to make games for his engine to get the word out

[YouGroove]

Yes LE 2 is not very known, because it lacks real 3D games specially.

 

What to say , it's easy to fail like Torque 3D or DX studio engines for example

Why because of commercial bad decisions, no correct bug fixes, not enought advert ? not enought mobile games (T3D engine) etc ...

And lot of people flammed Torque 3D because of bugs and blocking bugs , i remember a team becoming crazy making a game with it, then switched to another engine and they never didn't thaught on coming back to T3D.

 

Critical bug fixes is an issue to avoid in a 3D game engine.

Nowadays what you can see is 3D engine having and editor with easy workflow like drag and drop features, panels for physics, camera etc ...

LE 3 is also going in that direction to open up to 3D artist or simple beginner programmers.

The language reamins to be easy, and it seems lot of stuff will be now available by default (make the world and entities placement, load all that with one command etc ... )

 

Yes i believe in LE 3 caus fo it's new fluid and quick workflow, it's scripting from what i have seen seems to be in some way like what is done elsewhere (you could make some game entirely in script it seems).

It will have new visual stuff for some similar programming, you will be able it seems (i don't have tested anything) to create some sort of tools or plugins to incorporate on the editor.

So yes with Mobile also you are open to all Android tablet and Phones, so big potential of customers

 

What i hope to see , and we will see how it comes, will be some plugin templates (like doen in another well known 3D engine).

That's what will attract 3D artists, 3D beginners, new comers to game making :

Ability to put and adjust in panels the player gameplay basics (physics, controls, camera), just that is crutial to already put some default 3D model character that is the player in the level to test gameplay and level design.

 

So yes, if LE 3D will be enought open and easy to add features, these panel templates will be i think a critical features to attract more and more 3D artists and people

 

Even World studio with lightamps, included directy in the editor will allow you to wor like in UT3 , you will be able to create rooms or level design stuff fast. even people will be bale to craete levels using ony Wordl studio (specially for retro or casual games on PC or mobile !).

 

I don't knwo if it will be some trial version for new comers to LE 3 ?

But this is the other point that will make people decision, if they can produce and put gameplay (code or template panels) easy and fast, with some beginner 3D game tutorials (third person game , FPS game, Racing game).

 

Last point i already pointed out will be the advertisment :

So adverts on some CG sites and others, but priority will be the GAMES maed with LE 3.

 

 

Another 3D indie popular engine was not popular before V2 , it was Mac only, but already the great workflow editor attracted windows people , even some have baught some Mac just because they found easy script and wonderfull editor.

But was made it popular have been : MOBILE.

I've seen lot of programmers, simple 3D artists making little games, even good graphic ones, it was more and more games .. i've seen a basic 3D artist knwowing anything about programming making a game on mobile with some help in script.

Nowadays there are visual programming languages for non programmers at all.

 

So yes , LE 3 mobile games will bring advertisment and success.

If a commercial PC game would be made with it , be sure it will make lot more advertisment for PC game making with LE 3.

It's the gallery of games that promotes a game engine, so it will be our games also that will rise LE 3 popularity also.

 

When i go to some 3D engine teh two things i look at are :

- Forum topics not dead (and enougt people)

- Games selled with it

That shows the engine is running good.

(For example Shiva 3D that is a good engine have lost more than two years with people wanting the version 2 to improve lot of stuff ... they have loosed customer that have gonna to another 3D solution)

 

( I had some idea : an optionnal LE 3 logo diaplay at startup on mobile games , for people who would decide to promote LE 3 generously )

 

So yes i'm optimist, because mobile games are not big PC games (even if they are Big mobile games also), they can be made minimalistic in textures, models, but with great gamelay( even casual).

So lot of us will be able to put games on Android, it is much easy to publish for Android than on PC big platforms( steam, desura ).

 

Who knows perhaps 3D artists will take the step and go for big 3D games on PC with big terrain worlds

[Canardia]

The problem with Shiva is that it's too GUI oriented. Other than that, it's the only engine on the market which can do realtime shadows on mobile.