Announcement

Collapse
No announcement yet.

Idea: Promote simpler map designs for pub play

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • replied
    Originally posted by tidu View Post
    Nah, if the base game has the assets then they are not packed with the map. I do not know if the WIP assets in the editor are in the build though. My guess is that they are not since there's no pak to utilize them yet. In that case they would be included in your pak. It's the same as it's always been, it's just that we're starting off with fewer assets and 2k+ textures are the norm now.
    Hmm, I think I'd like a citation before agreeing with this.

    Each builder will have different compression settings and lightmap resolutions on their BSP surfaces and meshes, although a lot of this work is done at build, there is still those settings per build for each of the elements within the map.

    If you were to use the Unreal Engine as an independent game developer, you would want to have different compression for your textures per level/map/section/area for optimisation, and then you crunch for the platform (PS4/PC/Mobile/Colecovision).

    With Unreal Tournament, we typically build at 'epic' settings and let the individual downgrade accordingly, even so - when testing we we are aiming for 120 FPS and this is achieved normally by turning the dial up and down for lightmap resolutions and checking the near-endless settings and configurations via the viewport and texture density etc etc etc. In the good old' dayz, fog was added to hinder draw distance, now we got LOD and it's stuff like this which is really the key to optimization.

    No two maps or builds are the same, and neither are the settings used to cook them. So, I'm gonna need a lil' bit more before I join your side of the table...

    Leave a comment:


  • replied
    Originally posted by Scinbed View Post
    Sorry to go off topic but.. is this true? Seems cray to me. Gonna have to do some research.
    Nah, if the base game has the assets then they are not packed with the map. I do not know if the WIP assets in the editor are in the build though. My guess is that they are not since there's no pak to utilize them yet. In that case they would be included in your pak. It's the same as it's always been, it's just that we're starting off with fewer assets and 2k+ textures are the norm now.

    Leave a comment:


  • replied
    Originally posted by Scinbed View Post
    Sorry to go off topic but.. is this true? Seems cray to me. Gonna have to do some research.
    It is. When you "bake" a map into a .pak file, it encloses every asset that the map uses which isn't used in the base game (and some that are) and includes it as part of the pak.

    Leave a comment:


  • replied
    Originally posted by WHIPperSNAPper View Post
    Nowadays all of those files are bundled into the map file, which makes things neater I guess, but at the same time if 10 different maps use the same textures, they all have to get downloaded 10 times instead of just once.
    Sorry to go off topic but.. is this true? Seems cray to me. Gonna have to do some research.

    Leave a comment:


  • replied
    Originally posted by Mitch Mitchell View Post
    IIRC this is why map builders in dayz gone by used to team up and put together 'map packs'; running them on their own servers - I think there is room for collaboration on this scale.
    Back in the UT99 and UT 2004 days, you just downloaded the map file itself which only contained the map file. The textures and music for the map came as separate texture and music files, and if players already had those files, they wouldn't need to download them again. Nowadays all of those files are bundled into the map file, which makes things neater I guess, but at the same time if 10 different maps use the same textures, they all have to get downloaded 10 times instead of just once.

    Leave a comment:


  • replied
    Originally posted by HenrikRyosa View Post
    I do agree with much of your thoughts there, especially that Epic are likely going to push for as impressive visuals as possible. There is no reason that can't be done with simpler and smaller overall levels though, pushing biggest details out to the background or what we'd formerly call a skybox. Initially I was sort of under the impression that's what they would do, but I guess it's not always so easy, especially when you want to make a map built on castle ruins, grass and the like.
    Yah, I think they could/should also release lighting/environment maps as a baseline - so, all the post processing & lighting setup for the particular mesh releases, coupled with something like this:

    https://youtu.be/g5mojzojJfU

    ...would greatly assist the map making process.

    If you are going to go and share your meshes and you own the engine, you should go the whole way and really commit to helping the community. I got all the time in the world for Epic, but they are surrounded by swathes of technical people and artists. I think sometimes they (perhaps) forget we, are not.

    So yah, I agree with you.

    Leave a comment:


  • replied
    Originally posted by WHIPperSNAPper View Post
    IMHO one of the biggest concerns with complexity is file size. Not everyone has fast Internet; some people are stuck with 1.5 MB and 7.0 MB/sec DSL. It's kind of a downer to have to sit and wait while a 200 MB file downloads. It just disrupts the flow of the gaming experience.
    PSi is 57 MB and Genku (the one pictured above) is 89 MB.

    Supporting your point, perhaps this is why they are popular - and ASDF is not that large, but also comes with the engine so...kinda don't have a choice.

    Not sure where this is best aimed. Engine/cooking optimisation? Texture compression?

    IIRC this is why map builders in dayz gone by used to team up and put together 'map packs'; running them on their own servers - I think there is room for collaboration on this scale. Perhaps you would be able to get something like this for UT4? Rally support for a team of map designers to put together maps of the design and simplicity this thread is talking about and put a map together per team member and release a map pack?

    Even nVidia released a map pack for UT3...

    Leave a comment:


  • replied
    IMHO one of the biggest concerns with complexity is file size. Not everyone has fast Internet; some people are stuck with 1.5 MB and 7.0 MB/sec DSL. It's kind of a downer to have to sit and wait while a 200 MB file downloads. It just disrupts the flow of the gaming experience.

    Leave a comment:


  • replied
    i think DM-Ironic fits this perfect!

    Leave a comment:


  • replied
    I do agree with much of your thoughts there, especially that Epic are likely going to push for as impressive visuals as possible. There is no reason that can't be done with simpler and smaller overall levels though, pushing biggest details out to the background or what we'd formerly call a skybox. Initially I was sort of under the impression that's what they would do, but I guess it's not always so easy, especially when you want to make a map built on castle ruins, grass and the like.

    Leave a comment:


  • replied
    /signed

    Yup.

    I'm on board for having maps like this - it's one of the reasons I ported Genku to UT4:



    Simple 3-colour map with 1 overpass, 2 main floors, each with 2 sub levels internally.

    This map is apparently very popular, and if you haven't played it yourself - the link is below.

    The original designer (Andrew) sent me a PM saying he is happy it has now been ported and this is partially what I hoped for; I also hope he begins to map again for UT4 as he ticks every box for what this thread is about.

    So happy was I with his design, I also ported PSi, another design for what used to be called 'low-poly' and meant for competitive play - and if you look at the success of ASDF, there is definitely an audience for 'uncluttered' designs.

    I forget who it is I am quoting here, but Unreal Tournament is in a very strange place right now regarding design. There are those which say 'next-gen has to have ├║ber graphics with loads of post processing fx and blinky-bling-bling fancy-pants shineyness' and yet what happens when a new map gets put on a server? Everyone drops the graphics down to the barest minimum to see if they can gain an advantage. And yet, you make a 3 part video diary on the making of a map talking about how many people were involved and the different departments required - which is great for promotional material and general PR, but does leave a slightly bitter-sweet taste in the mouth of people who do not have access to the resources you do.

    So, on one hand Epic is pushing their engine to new heights of awesomeness, and on the other we got the most basic of basic maps being one of the most popular. We need to ask why this is.

    I seriously believe there is no middle ground here, and if anything we need the following to be better, to make this game better (in no order):

    • Better BSP tools/map building tools (this could and probably should have its own forum)
    • A GPU-powered lighting/baking system, or simply sending CPU instructions to the GPU to crunch at build
    • Easier goal-oriented systems configurable by the mapper and/or server administrators
    • Colour-systems for surfaces
    • Larger characters/scaling (I have absolutely no idea what the term is)


    Y'see Epic, I get it - I really do: you want a game to show off your engine and get people downloading the engine for free and building levels for you. I also get this game is in Alpha. All this adds value to your product and your engine because the positive words will spread. BUT...you gotta meet us halfway here, because if you would have told me the editor for the new Unreal Tournament would be using an ancient BSP tool coupled to the creaking CPU light-mass system 5 years ago - I would not have believed you.

    Guys, you get to a certain point and the editor just slows down to unworkable levels, I know it's not something you really experience yourself because you got (probably) a cloud CPU-rendering system and dumb terminals. Not I, and probably not the majority of the community have access to such luxuries. I get to about 500 surfaces and the Engine slows down, and everything and anything the community can say on this topic is just 'a workaround' and certainly not a solution:

    "you need 16 GB or RAM" - I got it.
    "you need a high-end CPU" - I got an over-clocked water-cooled behemoth.
    "you need a more powerful GPU" - I can run Outpost-23 on Epic @ 120 frames...and the editor doesn't use the GPU, does it?
    "you have to convert the BSP to a static mesh, FBX it into blender, dance a jig, cure a disease, scale Mount Everest...(three days of insane suggestions later)...and import it back in UTE4" and if I want to change the map during testing? Ah, thought so.

    I got a lot of time and <3 for the community, you can really tell it's not anywhere near as 'toxic' as some of the gaming forums I'm sure we could all name, and this is partially what makes the state of the editing system all the more worse. We want to build maps, but the methodology is a 'parts bin' of interfacing systems borrowed/stolen/ported from the big brother which must be returned by midnight.

    Digress...

    So yah, we want simplicity of design and a much better editing system to get these designs to the community faster.

    PS: sorry for the waffle this is a great thread, hope I haven't tainted it.
    Attached Files

    Leave a comment:


  • replied
    Ideally it seems like you'd want to aim for something that's easy to learn quickly, but has a lot more options once you know the map a little better. As someone else pointed out, Rankin does a good job of this, with a lot of clear easy pathways and distinct rooms that make learning it relatively simple, but with interesting movement options that become apparent once map knowledge and game mastery progress.

    I've been trying to think about what makes a map easy to learn vs hard to learn, and two things come to mind.

    First is fairly simple, and doable with just an art pass: different areas of the map need to have distinguishing features, whether they're large gameplay areas, a set piece, distinctive lighting, it can be anything, but it needs to have something. Secondly, maps with distinct "rooms" tend to be easier to navigate than maps that rely on winding ramps and carved out geometry. Compare Rankin, which has both rooms and distinct visual design to DM-1on1-Roughinery, a map that was well-regarded and fun to play, but horrible to learn. Roughinery by contrast has a visual sameness that pervades the whole map with no visual landmarks and while the gameplay is varied there are no real distinct clues to most areas that make it immediately visually obvious. It's the same thing with all the other UT200x duel maps like DM-Trite, DM-Serpentine, DM-Squander, DM-Spirit, etc.

    That's not to say that that complicated layered geometry is the only way to make a map confusing to play when you're new, though. While UT200x wasn't so guilty of this one, UT99 had some real winners in this category, like DM-Zeto and DM-Conveyor. Even though both of these maps have large, visually distinct rooms with varied gameplay, they also have sections that are confusing to traverse and overall don't have the greatest flow. An ideal map with good flow will end up with players naturally exploring the whole arena during the course of a game, whereas one with poor flow will see players not using certain areas often, leading to their mental maps of that segment being disjointed, and leave players with a lower desire to go to that area (or, in some cases, a desire but just the inability to find their way back there; I think we've all been in that situation on a new map).

    Leave a comment:


  • replied
    I think maps like DM-Radioactive is a good example of fairly simple in design and easy to navigate and I would like to see DM-Antalus as well as far as simple/fun/easy maps for beginners!

    I'm not a weapons guy but I love Chill and Deck as those two were pretty easy to remember where pickups are placed for me!

    Leave a comment:


  • replied
    Oh god TL!!! I hate that thing with a passion lol.

    But yeah sorry wasn't trying to get off topic here. So please continue. This great topic.

    Leave a comment:


  • replied
    Originally posted by -=}WoLvErInE{=- View Post
    ...in weapons I just happen to respawn with enforce and you come into my base fully jucied I stand no chance regardless our skills. And being its a complex map, you have a better chance of getting away from me before I can pick decent enough weapon to track you down...
    This is the whole reason CTF is played with Translocator

    But like I said, I like iCTF a lot as well, and you're right it is largely preference.

    And to keep this on track, having a bit more complex map also prevents the flag runner from making a B-line back to flag and gives the opponent a chance to catch up and cut him off. But this is more a component of competitive play than casual play. Which, again, I think that's the point.

    Leave a comment:

Working...
X