Bump Bump Bump
Bump bump bump
Allowing to define vertex density per landscape component is also very very needed. Some areas can be pretty low poly, some other areas might need more than 1 vert per meter and so on.
This thread is gold! I ran into most of the issues listed myself and could not have summarized it better. Here are a couple of additional ones I encountered and found no mention of here. Note: This list was compiled during a project that ended half a year ago, so some items might not be relevant anymore.
1.) As an artist/designer I cannot set a mesh for a foliage type in the foliage painter panel when that mesh is already used for another foliage type. This would be useful for painting the same mesh with entirely different settings (e.g. size and inclination settings).
2.) As an artist/designer I cannot define in the landscape layer panel which layers I want to be normalized against each other. I can only choose whether a layer should be normalized against all other normalized layers, or not to normalize it, but I cannot have it normalized against a specific subset of other such layers. Suggested solution: Allow for definition of “Normalisation Layer Groups” in the layer manager. A Normalisation Layer Group would define/hold those layers that should be normalized against each other. The system should allow for the definition of an arbitrary number of such groups with an arbitrary number of layers per group. This would help to make complex procedural grass type painting easier, which atm has to resort to normalizing layers in the shader graph, which quickly becomes hard to manage with increasing grass type variety.
3.) As a technical artist I want to be able to control the distance blending and maximum intensity of the Normal texture that is automatically computed and rendered for/on the landscape actor.
4.) As an artist/designer I want to be able to use the Foliage painting tool even if no landscape landscape actor is present in the level. The tool is already able to paint on anything, including static geometry, not just landscapes. A landscape actors presence should not be required in order to use it.
5.) As an artist/designer, I want to be able to assign already painted foliage instances to another foliage type. I.e. I want to be able to select some instances of a type, Right-click over another foliage type in the foliage painter panel and choose “Add Selected Instances”. Expected result: Selected instances of the current foliage type are assigned to the destination foliage type and replaced with the appearance of the destination foliage type. This is useful for replacing objects that have already been carefully placed and need to be replaced with another one without having to repaint or reposition them all over again.
6.) As an artist/designer I want to be able to manually control the rotation and scaling of individual foliage type instances in the models local space instead of only world space (which can be very counter-intuitive)
7.) As an artist I would like to have more control over random rotation of grass type instances (i.e. per axis).
8.) As an artist/designer, I want to apply computations in the transform panel to multiple selected objects, not just a single one. E.g. multiplying the scale of objects along the x axis by 10 by entering *10 in the corresponding parameter field should be allowed even for multiple selected objects (currently only works for a single selected object).
9.) As an artist/designer I would like the editor to remember layer paint tool strength separately from the sculpting tool strength so that it does not need to be readjusted every time when switching between painting and sculpting.
10.) As a technical artist I want to query a landscape layers world space normal from a shader based on a world space position. This would help in blending static model geometry normals with landscape actor normals in cases where the landscape actors normals are decidedly not pointing upwards (areas with pronounced slope).
10.) The last one is a bug I believe to have already reported, but I don’t know if it was ever addressed: When using the “Replace References” function from the Content Browser -> Asset Actions menu on a foliage type, scenes using that foliage type will contain orphaned foliage instances of that replaced type. That is: The foliage type deleted from the scene, but painted instances of that type remain in scene and belong to no foliage type in the foliage painter panel. This becomes obvious when removing all foliage types from the panel, which leaves only the orphaned instances left standing in the scene although no foliage type is left present in the panel.
Just bumping this for some developer attention.
Bump Bump Bump
I’d guess it is indeed good time to up the thread a bit, particularity about integration of this PR:
GDC Bump Bump
GDC 
Bump Bump Bump
I’ve heard multiple people say spline decals would be easy, so if it is why haven’t we gotten them yet? As somebody focusing on the art side of things I would love to add it myself, but i don’t code.
Either it’s not easy to make, or we just need the right person to add it to the engine. A lot of of people and teams would benefit from it.
Bump Bump Bump ^
Yes, bump bump.
[taken from another spline-decal post]
I want to bump this Thread as well. At the moment we’re looking for certain solutions like spline-decals. This is one of the little features that would speed up the daily routines in UE4, especially for environment based tasks.
Aside all of that “set something free, 12 mill assets, etc”, i would love to see Epic creating some sort of intern task-force that is building “plugins” for their own engine, maybe based on quarterly voted feature requests and then selling these on the marketplace. I don’t see a problem paying, lets say ~$30 for an official supported spline-decal system.
Someone hire this guy, he is a genius 
I fully support this idea ^
Sadly, I doubt we’ll get such a thing in this engine.
The typical “We’ll work on it if it aligns with our internal needs” has made UE4 stall for a while now. And even when they work on something that aligns with their internal needs, it usually stops receiving support as soon as their internal needs change i.e world composition is all abandoned, landscape in general. Recently they also stopped improving volumetric fog itself. Their engineers aren’t to be blamed or anything like that. But to be honest Epic’s goal doesn’t seem to be developing a “game” engine any longer i.e car demo in previous GDC was all about something other than games. Recent GDC’s hightlight was RTX which has nothing to do with games in current decade, etc.
Look at where Unity has gotten, they’re improving the hell out of it. I think in a year or two, if Epic don’t catch up a lot of people are going to start switching engines again if Unity starts to offer more than UE4 does. It makes no sense when you see some engine in 2005 had spline decals and UE4 in 2018 doesn’t. Along with many other old-basic features.
Additionally, if it’s really a matter of not willing to spend money-time on having someone work on what community needs, you Epic, could release your 12 million dollar assets on Marketplace as paid content, and use that money to contract someone for working on what people need for “PC/Console Gaming”. I doubt any serious studio is going to make a game with those Paragon characters since they don’t want their reviews to get nuked over night.
I have no doubt that some Epic Engineers are still with us here and doing outstanding jobs in supporting the community. I believe they’re spending time on things that are not on their daily task list because they love their work and love to see what we’re doing with it. An community like us has needs that are viable to look into because these things are important to stay competitive. Some of these needs can be an hilarious task to code as well, but they also became an industry standard somewhere else.
For all of this, Epic opened up their their Trello voting system. Some features got high votes, but stay untouched. I don’t know anything about the intern decision pipelines but i believe that external companies are an important part to a) keep Epic alive, b) making decisions, based on important deadlines like GDC. This is ok! Business as usual as long as you can pay your bills.
This sadly may not only corrupt the voting system, because the engineers got their tasks from “somewhere” by “someone” who “maybe” does not care about hobbyists if “Disney” is paying a skitload of money to see Captain Phasma in glorious Realtime Raytracing. Epic also started to split the Unreal Engine into two branches. The “free” UE4 and the subscription based Unreal Studio.
I worked for several Architecture Studios and i see only one reason why Unreal Studio exists. $49 per month (per Seat?) for Datasmith support and some royalty free substance materials. Yeah, big architecture companies like the term “enterprise edition”. Literally the same features like the free one but you can tell your customers that their projects are made with the “enterprise edition”. - Marketing is so easy, wearing black business suits, isn’t it? -
From an artists perspective (the one guy wearing nerd-tee’s at marketing meetings) it may or may not speed-up anything. You can still do your Architecture projects with the free one till “somebody” will decide that new features like certain GI systems will be an exclusive part of the enterprise edition. And then, we may see some shitstorm like EA got with their Battlefront II “decisions” with the hashtag #epicfail.
That’s why i came up with the idea making this an task force thingy. Unreal Studio is a way to get an monthly income to pay everybody working at Epic. It’s cool, why not. So, why not going a step further, making an $25-50 sub as well. Not for architecture and visualization, but for real game designers like us who don’t need Datasmith but other essential features that are already an standard in other engines like Unity or CryEngine. You can still implant these new features into the free UE4 once they’re stable. I’m ok with it as a paying customer.
I’m writing this because i started my journey with Epic in 1998: I got the game UNREAL for free at their Powerslide Game booth on CeBit and within its outstanding Level-Editor. I loved using it because it was way ahead of other editors like Hammer at this time, followed by all of their later Level-Editors.
Now 20 years later, i’m still with Epic because of their community and their outstanding engine. Aside this, Unity became stronger and will end up being the “Maya” for serious game developers while UE4 may become the lazy 3DSMAX for Architecture and Visualization. You can see that UE4 is already good enough to meet all requirements in this sector.
If you have the manpower, UE4 is valuable for developing outstanding games because bigger game companies can hire people to code the things you need as an artist. Without these people, i can stay here and waiting for sunshine, or making the decision to switch the engine to use the existing tools i need. I don’t want to make this decision yet, Maybe someone at Epic read these threads and start to think about why people, who are not hobbyists at all, crying to pay for the engine they use to get the features they really need for their next game project.
I you love something, don’t set it free and just let it go.
These aren’t separate branches. Unreal Studio includes additional stuff: Datasmith, Substance plug-in with some sample materials. Nothing was cut or altered in “Unreal Studio”. It’s mostly about licensing.
Well, that would simply decrease their profits while increasing paperwork and “fiscal work”.
I heard that even Unity (no engine has more users, I guess) makes most of the money from online services like ad system. Subscriptions are important, but the only source of their income.
I’d like to draw a red line between feature requests that noticeably affect the business and other feature requests that don’t.
For example you can be cool with not having dynamic GI. (Despite dynamic GI being there in almost every other engine) you can be cool with not having it because they might not be in a situation to do it and it’s understandable because it’s a lot of work.
Then there are other tiny little features such as decal displacement, or spline decals for making -roads- and I don’t think there’s any justifiable reason in the universe for not having it. Hmmmm I believe working on these tiny little -usable- features would have even made people a lot happier than seeing an RTX feature which would probably cost like $6000 worth of GPUs to run on a single machine. A feature that even if you manage to run, your customers can’t which renders it useless for “gaming” because no one wants to spend 6000 dollars for running some fancy feature of a 60 dollars game.
[USER=“69”]Just Krishna[/USER] Thanks for clarifying. I have to say i’m not the business guy at all, just seeing it from a very low point like reading news and certain reactions. I’m not against Unreal Studio, but it looks confusing for people who already use the substance plugin and source (i got an allegorithmic sub). Aside Datasmith, i don’t see a difference so you’re right, might just be for license reasons.
@Maximum-Dev
Regarding GI, we had some longer talks in our company two years ago because we were close to switch to Unity because another competitor used Unity’s GI solution and just looked better out of the box, even with less detailed textures and assets. We’ decided to stay with UE4, didn’t baked our maps and faked GI to get the “dynamic” we needed. Of course you can fake systems all day long, but in the end you’re loosing dev-time to create these fakes and probably run into other problems like performance and stability issues.
RTX might become viable in the next 5 years so it’s ok for developing and showing its capabilities. You’re right, the main concern here is about the tiny feature requests that are on the road map (not only on Trello but Forums as well) for a long time untouched that could open up so many fresh possibilities to keep up with other game engines. The big question: How is it possible for a handful of people to tell Epic that these tiny things are as much important “today” as some RTX feature in the unpredictable future.
Yeah I agree, GDC announcements were basically targeting film and media projects (RTX), I think they want to diversify which is good but not the best for game development focus… (spline decals and displacement decals yes please)
I doubt RTX will be viable for games as quickly as everyone wants to think, remember VXGI has been around for a long time and is still not widely regarded as product ready. On that note VXGI 2.0 was announced with improvements that were very exciting, but with the focus on RTX and epic doing unreal studio etc it seemed like it got less excitement than it should.
even without GI and other lighting features it would be nice to have the existing lighting for outdoors cleaned up a bit, it seems needlessly complicated to create a realistic lighting scenario!
Indeed, it is unrealistic to expect a fully deployed GI solution, running in under 1ms done in UE4, fully in-house next month,
But (quote from first post) not having these covered:
1. Pen pressure support for landscape painting. Ticket.
2. Improve landscape sculpt tools.
3. Brush falloff curve editor.
3. Expose terrain heightfield / normal map to material editor. Sample those in terrain’s pixel shader. Weight maps too.
4. Generate the dominant physical material map from the landscape material to better support procedural landscape texturing.
Is a bit disappointing.
And not having this one in stock builds in 2018:
1. Texture arrays.
Seems like a sin to me.
On the bright side, tessellation CSM performance issue actually received some love in 4.19
It was the only thing, that was addressed from the list in OP, or did I miss something ?