ClydeFrog
Members-
Posts
13 -
Joined
-
Last visited
ClydeFrog's Achievements
-
Dzon liked a post in a topic: MCS Project [UE4]
-
Smoo liked a post in a topic: MCS Project [UE4]
-
Jeff liked a post in a topic: MCS Project [UE4]
-
z3filus liked a post in a topic: MCS Project [UE4]
-
the_raven liked a post in a topic: MCS Project [UE4]
-
There will be force powers - in the future. Right now, we are still working on different game modes, maps, character customization, and adding yellow/medium style to the melee system. The team is very small, but we are chugging along. We should release a new version pretty soon. More info on the project's discord: https://discord.gg/ZR5F4x
-
ShakeThatSalt liked a post in a topic: Open source Jedi Knight in a modern engine
-
afi liked a post in a topic: Open source Jedi Knight in a modern engine
-
Boothand liked a post in a topic: Open source Jedi Knight in a modern engine
-
Tempust85 liked a post in a topic: Open source Jedi Knight in a modern engine
-
Open source Jedi Knight in a modern engine
ClydeFrog replied to afi's topic in Art, Media & Technology
Hello Everybody, this is a great topic I've been working on the exact thing being discussed in this thread - and I have a bare bones network capable (i.e. dedicated server) prototype. I plan on uploading a video in the next few days so you guys can see the progress so far. Basically, this is what I have so far: 1) One game mode - duel. Two players fighting (obviously), and the rest are spectators. All clients run through a dedicated server. 2) Three basic attacks (right, left, overhead), along with their respective blocks. For testing purposes, I have a bot as client #2 - who can never be hit because he blocks everything perfectly (more on this later). 3) A light saber blueprint which handles all of the light saber stuff (i.e. it's blade color, turning on/off, etc..) Also, the hit detection system is done by the light saber itself. And this actually ended up working quite well. I used the animations off the Adobe Mixamo site (mocap which can be downloaded by anyone for free). The skeleton used for the animations was re-targeted to the UE4 dummy (very easy to do with any skeleton really). A note about the animations. They definitely need some more work, but they are good enough for now. I'm using root motion for the combat - so the player attacking lunges forward a bit and takes the camera with him. So during the attack, you lose player control until the saber hits something (which stops the animation) or the animation swing ends. I tried the combat without root motion and my personal opinion is that you lose all tactical aspects of the combat - and face to face combat pretty much turns into players running around trying to get a lucky hit. Just my opinion, ultimately the community will decide. In either case, the way it's set up right now it is very easy to switch between both. As for the blocking, what would be your opinion on this? Should it be automatic, player driven, or a hybrid based on something like the angle the player is facing? I'm still working on this every spare moment I have. I think that even the most basic of foundations is enough to get the ball rolling on a project like this. Also, I would really appreciate any suggestion you guys might have on the combat - as I'm still tweaking the shit out of it. On another note, I have yet to write a single line of code (and I do know C++). Yes, it might be required at some point - but so far it has not been needed for anything. I've done no optimization work at all and so far it runs fine (and I have an old graphics card). So after I clean up the code, and more importantly, document it - I will setup a Git project. Hopefully, a lot of people will be willing and able to contribute. Agreed 100% Any animations would be great, if you could re-target them to the UE4 skeleton. Exactly, get the bare bones combat working and then take it from there. -
ClydeFrog liked a post in a topic: Open source Jedi Knight in a modern engine
-
ClydeFrog liked a post in a topic: Open source Jedi Knight in a modern engine
-
ClydeFrog liked a post in a topic: Open source Jedi Knight in a modern engine
-
ClydeFrog liked a post in a topic: Open source Jedi Knight in a modern engine
-
Aidor liked a post in a topic: UE4 feasibility test (porting Kejim WiP)
-
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
Looks like the roughness is way too low, and the roughness maps are too high contrast, leading to a "wet" look on almost everything, which is made even worse with UE4's screen space reflections. -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
After digging into and implementing UE4's animation system(s) I am very optimistic... Does anyone have any combat animations? I was able to download some from Adobe's Mixamo site (everything there is free for the time being), and I managed to use quite a few for the combat system I'm working on, but you can never have enough. And I suck really bad at animating. Thanks NightShadow. Any help is always welcome. As soon as I have something basic (bare bones) in place, then I will know and be much more specific on exactly what and were help will be needed. Agree 100% - focus right now is only on multiplayer. I'm not sure if Disney would give a shit if nobody is making money off the IP. By the way, the combat system I've built so far was done without writing a single line of code. UE4's visual scripting system can be used instead. I thought Epic Games was just talking bullshit when they would say you could build an entire game using only visual scripting - but they were right. Funny this is, I'm a software developer and I know C++, but I've yet to even use it on my project. -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
Yes, that is exactly the goal of this project. And I do agree with Teancum, a sparse environment will always look like a sparse environment...especially when we are all used to the detailed environments of modern games. Geometry "clutter" goes a long way in making an environment more realistic (pipes, cables, debris, etc.) - and there is absolutely none in JO. As Lancelot mentioned, adding extra geometry and polish would come later. And yes, it is possible to add missions (maps) together via streaming based on player position (like what is done in open world games) or even the old fashioned way, by actually merging both mission maps and positioning them correctly. Correct, there is no color grading, no post processing effects, no proper reflection. Global illumination is the default for the baked light maps and there is no global occlusion at all, only the default "screen space" occlusion and the occlusion maps used by the textures. Yes, a shooter + saber combat would be the ideal, but one has to start somewhere :-) I'm trying to put together a highly "tweak-able" animation system in UE4 for saber combat. The point being so I can slap a single duel level on a server and ask all the fine people of this forum for some feedback. -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
Personally, I think the best thing to do would be something similar to the way Epic Games is doing the new Unreal Tournament. It's entirely community driven, where anybody can contribute by adding maps, code, models, animations etc. Of course, that doesn't mean anybody can add anything they want and expect it to be in the final game (in the case of Unreal Tournament, Epic ultimately decides what community content makes the grade) but as long as there are some clear guidelines and goals it certainly can be done. My personal preference would be a saber melee game - multiplayer (and/or bots) combat. I think it's vital to start with something as manageable as possible. Start with only one map and only one game mode (for example one duel map). Put the project on Git hub so everybody has access to contribute, using only the main branch on some dedicated servers for public consumption and testing. Keep iterating and updating the combat system until the community is happy with it - then start adding the polish. -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
Just the saber combat alone would be worth doing, considering that it's been 15 years. -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
Hello again everybody, I finally had some time this past week to start working on my project again, this time on some interior sections of Kejim. I decided to work backwards and start building from the end of the map. @@SomaZ - I already RMA pack all of my textures. I also use the alpha of the base color for height maps, and the alpha of the RMA for emissive maps (so technically I do RMAE). When I'm finished, I plan on releasing everything I've done to anyone who wants it (map, textures, shaders, etc..) Here are some more screen shots: -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
@@Psyk0Sith Biped rig is fine - the bones can be re-targeted inside of UE4. Do you have a link to the model? @@Teancum You must be thinking of UE3. UE4 does not have a license - the engine, tools, and source code are all free. Epic Games takes 5 percent of your game sales after it reaches a certain revenue. It it's not a game (i.e. visualization, simulation software etc) then Epic takes nothing. -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
Absolutely! Is he already rigged? -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
Asgarath83, your options are pretty limited if you're forced to work with a 32 bit OS. UE4 is 64 bit only, as SomaZ mentioned, you could try Unity or maybe you could try the previous version of Unreal Engine (which is called UDK - which is also free). The only "professional" grade engine with all required tools out of the box (level, particle, landscape, material, animation, cinematic editors etc.) that I know of which supports C++ is UE4. In fact, all of the engine source code (C++) is included with UE4. UDK has all of these features as well, but no source code is provided, and all game logic is done within its scripting system. -
UE4 feasibility test (porting Kejim WiP)
ClydeFrog replied to ClydeFrog's topic in WIPs, Teasers & Releases
Thanks everyone, I appreciate it. @@ensiform, the tool I use to export the models is called Noesis. It basically converts quake maps and md3 files to *.obj or *.FBX files - including their textures and material ID's. Sure, that's all you would need for the most basic material shader in UE4 but just using that would be a total waste for an engine that uses physically based rendering. It would not look good or "updated" at all. So, to answer your question, I try not to make any 3D asset from scratch - I just use the source models and maps. I cannot avoid, however, re-texturing all of the assets. -
Hello to all. I'm a game enthusiast/hobbyist who works primarily with Unreal Engine 4. Jedi Outcast ranks at the top of my list as one the greatest games ever conceived. I recently played through it (again) and I thought it would be interesting, as a side project, to test the feasibility of remaking the environments in UE4 without making everything from scratch. More specifically, unless it is absolutely necessary, use only the assets included in the original. So I started with the first SP map "kejim_post". In the screenshots below - the only asset which was replaced was the shuttle (which was far too low poly), and I also added some object detail to the cliffs (smaller rocks, etc..). The rest was done by creating a normal map for each object (or each texture if it tiles). This might sound like a lot of work but it actually wasn't (that was the whole point of the exercise). Almost all textures are panels with an assortment of basic shapes jutting in or out (knobs, lights, switches, etc.). Creating the normals in photoshop using Quixel was quite easy and fast. Seems like it might actually be feasible after all. Here is what I got so far: