Jump to content
BES

Still waiting on substance support...

Recommended Posts

As the title says ..substance support would be nice...I will keep randomly requesting it until its added smile.png

 

I mean its crazy easy to use...to add details to your low poly models(please ignore the mention of Unreal Engine 4 and just watch the video..THIS video is only showing substance painter and designer):

 

Its a method I am using for all my models... and its annoying to have to export things to textures to work in Leadwerks ..instead of just exporting the substance files with the model..

Share this post


Link to post

To answer your question, I don't anticipate this kind of thing happening soon. This isn't the direction I want to go in. Unreal and Unity support this, so if I add support I am moving our feature set towards the same as a product that is free. That is a recipe for destruction.

 

If I added exploding oildrums, physics ropes, even realistic fire or explosions it would create more utility for Leadwerks users than me promoting a third-party brand like this. Few people even make their own textures.

 

I really really want to avoid becoming the third free (by definition, this means "worthless") game engine with all the same features as the other two. That is a very bad situation to get into, unless you are interested in losing millions of dollars for years in hopes of eventually making a profit. While we're at it, why not rename Leadwerks a name that starts with the letter "U"? That would be automatic failure.

 

So that's where my thinking is right now and why I want to avoid spending time on features like this.

Share this post


Link to post
This isn't the direction I want to go in. Unreal and Unity support this, so if I add support I am moving our feature set towards the same as a product that is free. That is a recipe for destruction.

That's a weird statement and thought, Josh. blink.png

If you don't want to go 'their' route or direction, you shouldn't add terrain or vegetation or anything else that they also have in their engine. Adding support for something isn't following other engines as you may think, it's adding support for something that may be usefull for us, the game developers. People using LE care about what's in LE, and not what's in other engines.

Support for Substance Designer would be cool, and would even help speeding up the process of creating textures for the games we make. It's not that we can't make textures without Substance Designer, but Substance Designer is a pretty handy and usefull program for texture artists AND ESPECIALLY for those non-texture artists who have a rough time creating good textures.

 

I really really want to avoid becoming the third free (by definition, this means "worthless") game engine with all the same features as the other two.

Then i guess Blender is worthless too? Yet, i see there's tons of people using it. huh.png

Share this post


Link to post

Blender is only popular because it's free.

 

I want to focus on more high-level features and game content. I don't want to get dragged back down to low-level details. A lot of that stuff just isn't valuable and only appeals to a small segment of users. Adding integration with a $300 program isn't something a lot of our users will find useful.

Share this post


Link to post

I can understand the goal of LE. It must be quite hard finding a successful road among all the competition out there. The reason I'm now using LE is in fact its simplicity (specially with LUA). I have to admit that I miss some of the advanced graphical stuff offered by the big dragons, but its no showstopper in my view.

 

PBR is quite beautiful I have to say, but for my simple needs its not a big deal. I guess there are quite a number of "simple users as my self" out there whos primary goal is to make a small simple game with mostly premade components interacting by LUA scripting.

 

Of course I would want both PBR, much better water, animated clouds and all that, but again. Not a big deal for small simple games, but much welcome if offered of course.

 

One danger with the 'simple' approach Josh, could though be that LE ends up i just another FPS-creator, so watch out for that and don't throw away your graphic skills to much. LE must be better than the existing simple game creators although it doesn't have to compete with the big ones.

 

All in all I think LE is on the right track.

 

A BIT OFF-TOPIC (sorry about that)

----------------------------------------------

What I would like to see is more premade scripted game objects as different cameras, fires, explosions etc. that more or less could be dragged into the project.

 

Another thing would be that the flowgraph editor would be expanded and made better. That tool is great for making life more easy even for non-programmers. Right now its a bit to simple for some serious use. With some grouping of objects where a number of related objects could be grouped and folded to one box or expanded for editing it would be great.

Share this post


Link to post
Blender is only popular because it's free.

So, if Blender would be non-free (aka 'not worthless), then people would not use it? I thought LE supports Blender now? biggrin.png

 

I want to focus on more high-level features and game content.

Game textures (which is what Substance Designer does) isn't 'game content' or 'high-level'?

 

A lot of that stuff just isn't valuable and only appeals to a small segment of users. Adding integration with a $300 program isn't something a lot of our users will find useful.

As far as i know, the amount of people that use Substance Designer is growing and growing. Existing Substance Designers could even sell substance packs IF there was support for it in LE. wink.png

 

I can understand the goal of LE. It must be quite hard finding a successful road among all the competition out there. The reason I'm now using LE is in fact its simplicity (specially with LUA).

LE will not become harder to use if there was support for Substance Designer. They are both seperate programs, doing different things. LE is a game engine, Substance Designer a substance/texture creation program.

 

PBR is quite beautiful I have to say, but for my simple needs its not a big deal. I guess there are quite a number of "simple users as my self" out there

You don't need PBR to be able to work with Substance Designer and LE. Substance Designer is capable of exporting non_PBR textures or substances.

 

Anyways, i don't think that support for SD will come any time soon, but it would certainly be a great feature to have. smile.png

Share this post


Link to post

I remember similar arguments when PNG first started gaining notice. Liner colour space has become the standard for 3d graphics and in time this will have to be addressed if LE wish to keep abreast with emerging technology.

Share this post


Link to post

I remember similar arguments when PNG first started gaining notice. Liner colour space has become the standard for 3d graphics and in time this will have to be addressed if LE wish to keep abreast with emerging technology.

That's what I have a problem with. Why should we let a competitor dictate what emerging technology should be? Maybe they're not keeping up with emerging technology like our game launcher. I really dislike these types of threads.

Share this post


Link to post

I was mostly talking about “PBR” Linear colour space is a smarter way of working, it’s the evolution of the normal map we all grew to love. Substance designer is only one tool embracing the transition of many.

 

That said I think native substance support would be a great addition to your game launcher given the reduced file sizes.

Share this post


Link to post
One danger with the 'simple' approach Josh, could though be that LE ends up i just another FPS-creator, so watch out for that and don't throw away your graphic skills to much. LE must be better than the existing simple game creators although it doesn't have to compete with the big ones.

 

I agree with this one totally. As far as I can see, LE users are more or less computer graphics savvys that like to do things in a simple way. This things include low level as well as high level aspects. An LE is nice because you can do low level and high level model or texture manipulation. For me, this is the most important aspect of LE and a reason I'm using it.

 

Another nice thing is that a swinging door, for example, does not come as a black box, but as a model with a script attached - a script that you can read and modify to your likings. My point is that something like a swinging door is not built into LE. Rather, it's easily constructed thanks to LE's design.

 

OK, sorry for not talking about PBR and SD. An SD support would be a really nice thing and would not harm the aforementioned design principles. On the other hand, for people that want rather simple but powerful way to create procedural textures, there is an excellent free NeoTextureEdit. Unfortunatelly, it seems to be no longer developed. Nevertheless, I use it all the time for my textures.

Share this post


Link to post

Yeah, the door example is a good case of what we're all about, and I want to have more stuff like this. This is getting into another discussion, and I know the OP didn't mean anything bad.

Share this post


Link to post

PBR is a simple normal evolution of graphics, nothing dictated as we had textures before , then came normal maps that became standard after some time.

And cames now PBR that is already standard even on small engines like Paradox or Jmonkey engine.

GI and Sub Surface Scattering are also becoming more and more a standard ,something that exists from long time with Crysis engine,

 

At some point LE3 will have to incorporate PBR, because all tools will only do PBR materials today : Blender, Photoshop and DDO, 3D Coat , Substance, Mari etc ... The Diffuse/specular workflow is no more used, we use roughness and metalness attributes to display real materials.

 

There is still amazing stylized games not using PBR and doing great : Shadow Complex, Limbo, Journey etc ... It really depends if you target super realistic materials or if you give more importance to gameplay or target simpler graphics or a special style or fast to make graphics.

 

Journey_06.jpg

 

tumblr_nsjdcikkpT1s7o7zxo1_1280.jpg

 

1934-Borderlands_2_Stalker_Hunting.jpg?1352398974

 

What i would see as LE3 priority :

 

1- Better integrated Lua editor

 

2- Vegetation with wind parameters

 

3- Full GUI system

 

3 bis- Performance improvments (UE4 FPS demo runs faster then LE3 fps demo on my laptop or PC)

 

4- Plugin system : to have LE3 extend a lot and quickly, until then it will progress too slowly, and all future templates or editor extensions will rely on Josh only. Even Unity has opened some parts as open source rolleyes.gif

 

5- The flowgraph tool is just left as it is and it would need a good attention and some revamp like the ability to declare or change variables values in the flowgraph directly ,why not flowgraphs per object ?

 

6- Better physics or 3d engine : for example joint ball with the base joint moving is impossible and makes the engine crash blink.png , so we can't make simple hair physics or clothes physics for characters using joints sad.png.

Some old games having such faked physics really shine , even today.

 

7- Improved prefabs ( putting multiple simple physics child objects in a prefab and it will go crazy when instanced)

 

7 bis - Physics collision layers and some graphic table set up to say what layer collide with what other (more easy to setup collision)

 

And later PBR and GI smile.png , why not.

 

 

 

LE3 is one of the most easy with Lua and one of the most easy framework, or when you are a coder or want full control on the engine , new features are coming in a good way recenlty , it would just need more DLC (level art and models, weapon and script , character and AI , car and AI , complete template games etc ...) for all beginners to start.

 

This is just my opinion on PBR and LE3 , and because i am full buzzy on Unity and i am more a 3d artist then coder perhaps it is not a good opinion at all on these points , so don't take it too seriously rolleyes.gif

Share this post


Link to post

More high level features is better than low level features, high level I mean more functions in the engine, like bone/ skeletton features, export and import features and so on, then we can write our own editors and tools to make life easier.

Share this post


Link to post

Gonna go through some of YouGroove's points.

 

1 - Josh said that they are putting research into how to make this better.

 

2 - Maybe this will come with the Vegetation system in 3.7, but I doubt it. I'd put it in the suggestion box if you really want it.

 

3 - In the LE library, there are headers for UI elements such as buttons, sliders, and such. Josh said that these will be added slowly.

 

3 bis- Performance improvments (UE4 FPS demo runs faster then LE3 fps demo on my laptop or PC)

 

3b- OH MY GOD THIS! I understand that OpenGL is a different animal than DirectX, but it's just silly that a mid-range PC like mine can run UE4 at 60fps while it looks like a Hollywood blockbuster film, while the LE showcase map or the Vectronic Demo dips to 45fps at times, and it looks like the Source Engine! (Not that's a bad thing) This really needs to be fixed/optimized better, no one should have a GTX 980 just to play any LE game at 60fps.

 

4- I think this is because Josh want's a streamlined editor so it's easier to nail down issues. +1 for dreaming though.

 

5 - There is no doubt that the flowgraph editor needs to be more of a feature than a "Oh, here's a thing, feel free to use it if you want" element. I sometimes forget it's there and mostly use the entity field in my scripts.

 

6- Yeah, when I show someone else the engine, it's mostly the physics that throws them off. Of course, other people and myself are use to Havok Physics, but that has a $25,000 royalty fee.

 

7- Yeah, the prefab system can get a bit screwy if you're doing something more than a simple model with a script. Issues I have most with it is when you instance CSG.

 

7b- I don't really understand this.

 

And about your whole PBR argument, there should be an easier/better way to make shaders, Your visuals are all about the shaders and how you use them! Shader design should be made for artists in-mind. Of course, this is easier said than done, but it's something to think about. We can't have Shadmar make all of our cool shaders for us forever.

 

If you feel like your project would do a lot better on a different engine, then why aren't you doing just that? I'm developing my game on LE from Source because with Unity, and UE4, I got so overwhelmed as a solo developer. They were powerful, but too powerful for how simple I wanted Vectronic to be. Plus, them publishing royalties, while Leadwerks has none. You simply buy the software, and you're good to publish - you own your game.

Share this post


Link to post

Yeah... I didn't mean to start anything ...I just think substance support would be a nice edition to Leadwerks Engine..

 

Not having the support impedes workflow though...since, as I said ...you have to break down the substance to individual textures after painting the model(substance painter) and/or export them to textures when making the substance(substance painter and designer).. the model doesn't always turn out how you wanted it when you painted it when its in Leadwerks Engine..

 

When all you should have to do is transfer the substance file with the model then assign it to the model ..

 

As everyone knows by now I work with multiple game engines...including Leadwerks ...but they all support substances(except for Leadwerks)... its why I decided to use substance painter/designer for all my models...painting directly on the models is much easier.. being able to change the texture in code procedurally is also a big help..for simulated real-time damage ..among other things..

 

I just think the engine would benefit from adding this and PBR shaders in general..

 

I like Leadwerks Engine and would like it to evolve...but evolve with the developers currently using it to make games by supporting the many tools the developers use outside of Leadwerks.. that includes plug-in support also..its been much requested by the developers using Leadwerks Engine...yet the engine isn't evolving to support the developers requesting it..

 

EDIT: When I talked with the substance devs..they said they were willing to work with Josh to get the support added..

Share this post


Link to post

GI would be sick and a PBR shading model. The 3 things that I would like to see in Leadwerks really badly would be, Global Illumination doesn't have to be real time, I would be happy to use lightmaps. PBR, this is something I am dying to see in Leadwerks! Cloth and liquid simulation.

 

I would throw money at my screen for these 3 things, be it a Kickstarter, personal funding to Josh or whoever. I really want to see those 3 things so badly. I am using UE4 for a project and yeah the fancy stuff like, blueprints, persona, cascade its all nice to have but its the simplicity of Leadwerks which makes it number one!

 

Coding wise I feel Leadwerks is as simple as it gets thats coming from a total noob at programming, (ask Rick he will tell you lol). I feel that graphics should be a focus now, I read through this forum last night to see if there was any activity on this, may be a post from someone with a glint of news. I know the focus has been on level editing tools, a lot of Steam integrating which is awesome. Please focus on graphical stuff, get Leadwerks up to date with current tools. A lot of people are picking up these new PBR enabled tools on Steam now, content/assets are gonna pretty much be PBR ready too. It is only a matter of time until old ways are gone and it all goes standard (which is coming fast).

 

BES: Back in March I asked about this https://forum.allegorithmic.com/index.php?topic=1134.0

 

I think Substance support would be a benefit for Leadwerks. There are a lot of people using the Allego stuff adding support for it in Leadwerks would open up to a whole new bunch of users meaning more projects with the engine pushing it even further. Personally I can only see it as a positive. May be I'm biased though :P

Share this post


Link to post

My biggest problem is the massive fps drop that Leadwerks has. One more day runs at a good 4 - 15 fps on my computer, my game was RUTHLESSLY optimized so that the lights wouldn't kill the fps. Any time I want to add a light, I have to think do I really want that there. There can never be more than 4 lights on, and the Directional light will cripple my game play, YET I can run Dying Light at 60fps, all Source games except Cs:Go in excess of 100 FPS. There are some optimizations to be had. If we could work on the engines rendering above all else, I would be happy.

 

-A Frustrated AMD R9 270x 4gb Crossfire user.

Share this post


Link to post

My biggest problem is the massive fps drop that Leadwerks has. One more day runs at a good 4 - 15 fps on my computer, my game was RUTHLESSLY optimized so that the lights wouldn't kill the fps. Any time I want to add a light, I have to think do I really want that there. There can never be more than 4 lights on, and the Directional light will cripple my game play, YET I can run Dying Light at 60fps, all Source games except Cs:Go in excess of 100 FPS. There are some optimizations to be had. If we could work on the engines rendering above all else, I would be happy.

 

-A Frustrated AMD R9 270x 4gb Crossfire user.

 

Check to see if your pointlights have any shadows enabled. Disabling them should give you some frames back.

 

Yeah, even I am not satisfied with how the Vectronic Demo came out in performance. Of course, looking back, there was somethings I've should have done better, but overall I kind of struggled keeping my framerate at 60fps. I've went back in my demo map, and made sure all my lights followed this guideline, and still my 3 room demo hits the high 40's, and I noticed the framerate was higher in a packaged build with 3.5. (I still got that build if you want it Josh!)

 

I'm not saying they should roll back, and have you bake the lighting again like in 3.0; but I feel like if developers could bake the lighting, and use deferred lighting if needed, games would overall run better, and the shadows wouldn't fade out at a distance, or do stuff like this. But I assume that you can't/would be a chore to have the best of both worlds.

 

From how Josh put's it, Leadwerks and games developed on Leadwerks are targeted at people who are interested in indie experiment games, but to me, it feels like you need a PC that a "Yo Bro Shooter" would have to have the best experience with the engine. I think a higher priority than PBR or substance support is better optimization. I'm gonna run a few tests to see what causes the slowdown mostly. (Whether be shadows, amount of objects, amount of lights, polys, etc).

 

There is no doubt that Leadwerks needs more games for it. If there was a at least few top notch quality games that looked and ran awesome, Leadwerks would gain popularity and more people would be using it (or know more about it to spell it correctly!) If developers are having issues actually creating their game with performance/workflow issues, ether no games will be developed for it, or the engine will be known for low quality "games". If you know how to code shaders, your games today can look really nice, but that's overall the hardest, and most complicated part of the entire engine.

 

I really hate to open this can of worms, but I think it's important for the product as a whole for people to speak up and openly talk about the difficulties of using Leadwerks beyond "Oh, I can't do X, Does not have Y, bad engine". If you can't get your shelled out map (a map in nothing but dev textures and little models) to play well at a good framerate, and you went through your code through a fine comb, you can forget about how cool you wanna make it look; it'll still run like trash! Leadwerks has so much future potential, I feel like a full update on improvements and optimization would be richer than an update with a new feature or two.

Share this post


Link to post

I agree, performance on my system is abysmal. Default template, two csg rooms and 4 lights makes it crawl at 20-30fps in run mode. Until this is fixed no feature will make it better.

Share this post


Link to post

It should be noted that some performance issues such as this one are driver side, and the Leadwerks Dev Team can't really fix alone.

 

I've played around with the AI and Events showcase map, and it ran fine as expected after I turned all lights that were buffered into Static+Dynamic. There is a framerate drop going around the first corner (dips down to 27fps) and while a crawler is moving, (I made them chase me all over the map, and I got my frames back when they died) but the shadow mode is correctly set to Dynamic. So that must be ether navigation side, or something in the script/animations; haven't not really looked.

 

So right now, just follow this guideline on shadow modes, don't use buffered lights, and while shelling out your map, just set the ambient light to a bright color (almost fullbright) with no lights so you can see what actually impacts performance as you develop your map and scripts. Having deferred lights are great, but you still need to be careful. I've still yet to figure out why the Vectronic Demo runs so poorly still, but that might be because of my garbage ways of doing things, I dunno. Again this was with one map, so I gonna do a bit more testing with lights models and such.

 

When I go back to Vectronic, I'll make sure to note what things make the game stutter.

 

Sorry that this topic derailed to the max.

Share this post


Link to post

Yeah...performance is an issue...but it was only an issue for me when in debug mode..

 

After publish there was a steady 120fps ... even on a low end laptop..

 

I thought it was fine...but if you guys think that's an issue..then I guess it would be a priority over anything else..

 

My current leadwerks game has been on hold for a while(working on other things atm)...so I haven't tested anything to see if performance has dropped since previous versions..

 

I think I made my point(or points) anyway..

Share this post


Link to post

My biggest problem is the massive fps drop that Leadwerks has. One more day runs at a good 4 - 15 fps on my computer, my game was RUTHLESSLY optimized so that the lights wouldn't kill the fps. Any time I want to add a light, I have to think do I really want that there. There can never be more than 4 lights on, and the Directional light will cripple my game play, YET I can run Dying Light at 60fps, all Source games except Cs:Go in excess of 100 FPS. There are some optimizations to be had. If we could work on the engines rendering above all else, I would be happy.

 

-A Frustrated AMD R9 270x 4gb Crossfire user.

 

I'm surprised (and disappointed), I get about the same fps and I'm on an APU not an AMD R9 270x 4gb which should be way more powerful. Believe me I've tried to optimise the heck out of it. Definitely would like to see some performance in the engine.

Share this post


Link to post

You could just add a light that is static shadows only, or no shadows. Dynamic shadows, especially point lights, have a significant cost.

 

I'm surprised (and disappointed), I get about the same fps and I'm on an APU not an AMD R9 270x 4gb which should be way more powerful. Believe me I've tried to optimise the heck out of it. Definitely would like to see some performance in the engine.

That's because your game is CPU-bound, not GPU-bound. This is likely due to the number of entities in the scene. The culling distance setting will make a big difference because it will prevent all small objects in the distance from even being iterated through. Your scene has a similar performance profile to "The Zone", which is what I originally implemented the distance culling algorithm for.

 

Set as many objects as possible to "near" or "medium" culling range. Even zombies should be culled by distance, STALKER doesn't show characters past a certain view range. This will massively reduce all those small objects in the distance that are being rendered. (Hopefully you are not turning on occlusion culling for each object, the default settings are best.)

 

This system is what allowed "The Zone" to run in real-time.

Share this post


Link to post
Set as many objects as possible to "near" or "medium" culling range.

 

I've already done this on every object in the map. Nearly all objects/furniture in buildings are mostly set to near, some medium. Big outdoor things like buildings and pylons are the only things set to max and other medium sized outdoor things like fences, cars are set to far. I have zombies set to far because medium gives too much popup on important outdoor things like the enemies.

 

The only thing I can think of to address this then is to bring in a closer/denser fog in the game and then reduce the max/far things down to far/medium so there is no popup. It will change the look of the game but who knows maybe it will be better. I'll do a test copy of the map like this and see if this helps the performance.

 

(Hopefully you are not turning on occlusion culling for each object, the default settings are best.)

 

Nope not doing that. Thanks for the advice.

Share this post


Link to post

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...