(April 2016) Monthly WIP screenshot thread


My Ludum dare 35 Entry

It is game where you have to adjust the shape of your cubes to match the hole in the wall.

It was a lot of fun. I used a gradle file and intellij with latest alpha build of jmonkey.
It worked really well. Thanks for the awesome game engine.


I use baked AO and SSAO in Maker’s Tale. I rendered a texture atlas from blender of all the different ways the edges of my tiles could meet, then chose the correct texture at runtime for all the faces of all the tiles. You can sort of see it without the SSAO filter at the top of the Light And Shadow wiki

It may be worth noting I never use the standard jME SSAO filter in my projects, I dont like how it looks nor how it performs, I run a custom SSAO solution.

1 Like

If you want to use SSAO correctly it should be rendered as a pre-pass. Then you could add it as ambient contribution which wouldn’t effect diffuse or specular lighting. To combine SSAO with baked AO you can just do a min() operation to select whichever one is darker.

Also, for Minecraft-like games there’s a much more efficient and simpler algorithm:

Oh boy :slight_smile:
I’m kinda thrilled with the latest jme improvements.

The physical pipeline + tesselation shader + the ao system you work on bring jme really to new hights :slight_smile:
(if only we could get a decent deferred rendering, but I know how hard it is :slight_smile: )


I feel uneasy about this comment… sure by strict definition, AO only affects the ambient lighting, but that’s not the entire story, since the effects of AO can be cancelled out by direct lighting… My post pass AO filter takes an objects lighting into account when calculating the AO.

… I’m not to argue, I’m more than happy to be proven wrong so I can tweak my filter for a possibly better result =)

I hope that someday people will share their premium graphics code with the community.
But I can understand when people don’t share - since I’ve worked for several months on the upcoming framework I had doubts if I should really share that huge piece of work. ^^

I feel like that is a jab at me, which is completely fair and if so I take no offence, I release next to none of the code I write. It’s not that I don’t want to share, it’s more I don’t have the time, nor the will, to refine, tune, perfect, document, write test cases, explanations, uses cases and finally release stable chunks, from time to time I do, and every time its ALOT of work for me.

Perhaps a cheap cop out on my behalf, but that’s where I personally sit. I’m slowly working towards a releasable TTFX lib (TheToucher FX), which encapsulates all my tweaked and new shader systems, along with a runtime UI, if and when finally released hopefully there is some enthusiastic monkey that can tear it all apart and integrate any useful code into the core.


Not you specifically. I’ve seen lots of good stuff from several different people. :chimpanzee_smile:
My hope is, that some day games made with jME can compete with games made with big engines both visually and performance wise - and it’s not only nice assets that we need to achieve that goal.
Good games are always good “ambassadors” for their engines too…

“In-pass shadows” - do you mean the shadows are applied during processing the light? Not drawn over the scene with some post processor?

If i would share my deferred pipeline and shadows then probably noone would use it, because it is customized like hell. Imagine that I never touched the source code of the engine, instead of that I made many tricks to override the engine’s pipeline. For example - I don’t put lights into the scene and my material still have the stock’s default technique taken from Lightning. But due to the lack of lights nothing is rendered in the ‘main’ pass, it is used only for filling the shadows casters/receivers lists… All the magic I did is closed inside one scene processor, which contains two things: code for building g-buffer and a list of shadowmaps/light renderers.

Sounds interesting?
Trust me, it is better to write a new one from scratch than adapt this one to 3.1.

It is a bit different in your case because you’re using deferred rendering. That sort of implies that all lighting is performed as a post-pass. If jME3 were to support deferred rendering natively it would probably work as a SceneProcessor also which would default to rendering the scene to a GBuffer and then perform lighting as a post-pass.

Well, why not post it here too. Did that to help a user in another thread. Will be open source published in a couple of minutes…


What are we looking at here?

The app TestTerrainMesh.java - it’s the “triangle soup” version of a flat-shaded terrain mesh with pure random height points and some arbitrary triangles. There is also a “regular grid” version of that same terrain mesh demo. At the moment it’s just a mesh, not really a terrain like in jME terrain system. Here is the other thread: Custom mesh light glitch [Solved]

Hello there!
After years of trying random things and scrapping them immediately I have something that kind of works: random dungeon generator.
For now it just adds in parts with random orientation but I’m planning to fix it so there will be no open ends.
I don’t really know what to do with this generator though, so it’s just a fun side project for now. If anyone has some cool ideas what to do with it let me know :wink:

Gone insane with the dungeon size (it’s fully adjustable)


Post on github and a jmonkeylabyrinth plugin will eventually appear :wink:


Heh, IDK why someone would need a jmonkeylabyrinth plugin :confused:
But if you guys need this for whatever I’ll push it to GitHub once finished. Although was lazy and am using some lifehacks (manual stuff where automatic orientation recognition should be used).

We still in April :wink:

a 30 second video for the game play , sorry low resolution
1 - the camera is now Medium shot.
2 - lights revised to use Gamma correction.
3 - Graphics updates in location (most of the components has normal maps + diffuse map)
4 - other features like slight enhancement in NPC interaction will not be shown in video.
5 - I am still working as single developer in my spare time :), but i begin to expand the community around the game, and I give it to two friends to try it.

BTW: I moved my work to Unity3d just for curiosity to check out the possibilities, and I succeeded to migrate the graphics, and the camera, but it takes me 8 hours, then I moved back to the awesome JME, Unity is great no doubts but JME is awesome and great


Make a dungeon crawler game! :slight_smile:

Like this one? Skullstone - a dungeon crawler game

1 Like