Fortnite vs UT mapsize?

[TheDane]

I would realy like it if you mappers would come with your inputs on this.

I'm about to create an Island map simular to the map in Fortnite, which is huuuuuuge! If I where to create a battle ground simular to Fortnite for a player load in UT for max 32 players, what would the ideal map size (in Editors unreal units measurements) be? I've got an idea, but before I post my measures it would be great to get other inputs not affected by my estimates

Please take gameplay into account, I want to have the same balance in the map as in Fortnite's map, just with up to 32 players instead of 100.

hint: Editor ranges from 0 to 65536 UU

[Barbie]

As far as I remember my tests the maximum direct sight of line the engine supports is 2^15 UU. The map can be bigger than (2^15)^3 UU but there should be some opaque obstacles to limit the length of line of sight.

[JackGriffin]

Yeah, view limit looks really ugly if you don't prevent it. Also it's not at all hard to hit that limit. Several large sniper maps have corners that HOM out if you look across the map.

If you are thinking of a parachute in as a way of spawning the players I spent a good deal of time thinking about this problem and how to deal with it. About the best thing I could come up with was having the map be a nighttime one. Then as the player fell I'd just force a restricted cone of view distance around them. Think distance fog but using darkness. Once you hit the ground that would be removed (slowly so it wasn't abrupt). It might be the case that you'd leave the distance limit in place and then you wouldn't need to be so careful of map design.

Speaking of map design I online tested to push it out as far as it would go. You know the map MH-LostInTime? That's the big one with all the dinos and the volcano.After getting permission to modify it I stripped out the monsters and extra shit like triggers and such and then duplicated the map for a total of 5 'layers'. I stacked them and connected them with warp tunnels. Total floor space was 5 nearly edge-to-edge X<-->Y planes. Then I tested it online and I was fully able to travel among all the layers without issue.

The end result is that you can use a large map but the trade-off's start to become serious. You can make a nice big map but your poly/node count will quickly get dangerous if you add a bunch of terrain or try to make towns and built-up areas. A really huge map will just need to be pretty bland and boring unless it's mostly flat.

I'm rambling, I know. I just spent way too much time messing around with the ideas you are interested in and my end result was making BRUT so it just worked with any DM map. The parameters that the engine have are too restrictive to allow any half-decent battle royal playground. Really the only monstrously huge maps that ever worked were the inch-high guys ones, and that's because they were all pretty damn simple when you break them down. Box rooms, a few decorations, and line-of-sight broken up by doorways and halls. It was all suited to the engine. If you want to replicate that in an open-world scenario the results are going to be way less appealing.

Then again, if there was someone to prove me 100% wrong I hope it's you.

[Dr.Flay]

A reminder that UE2 quality terrain is possible in UE1
Realistic terrain + Demo map with funky domed skybox (and 3DS MAX converters)
www.silveribex.com/mt/archives/000043.html

Zoning the map could help performance.
You can use large sheets and divide the map based on the terrain, or put zone boxes around the various groups of buildings or other areas with lots of stuff.
Zoning buildings can be a pain if you put individual sheets in doorways and windows, but you can just put 1 cube in the walls of the building which will often use less sheets (depends on how many openings).
Each zone can have different overrides such as body counts and distance of rendering certain effects.

Stacked maps should be their own zones also (or just sealed normally which will zone them). This can help reduce the sound travel and confusion caused by hearing a fight above or below.

A note on audio ambience
For this sort of game type and weaponry, I would very much recommend using the new audio renderer from OldUnreal and adding EFXZone info to the map, or individual zones if added, so that a natural outdoor reverb and distance of sound can be set (obviously buildings or structures, even woodland/forest can have more suitable reverb preset).
The newer audio system is ray-traced and uses occlusion, so bounces off surfaces and is blocked my meshes such as a breakable window or movable door.
If you are in a building, gunfire outside will sound like it is outside. Likewise if you are outside and people are having a firefight inside, it will sound muffled and contained.
...unless they are inside a hangar perhaps, in which case there is also a hangar preset

It should be noted that Smiftsch has made it so that if the map is loaded by someone using an old audio renderer, it will just ignore the extra info and still work, so compatibility is not a problem.

[Feralidragon]

Technically speaking, you can increase the map size in 2x if you reduce the size of everything else by the same value.
You could probably go up to 4x the size without major issues, especially concerning the compression of certain data structures on replication (namely vectors, replicated as integers), and there's at least 1 total conversion that I know of using this technique and it seems to work quite well (Airfight UT).

Or you could split the map area into different areas altogether, like different areas or biomes, like Apex Legends does (although you can see the whole map from above and go from one area to the other seamlessly), and have the player select the area to spawn from rather than jumping from a bus, and prevent 2 players from spawning at the same area, which is what Ring of Elysium does.

If you go the resizing route, you have a fair bit of work ahead in customizing the whole thing to be x times smaller for a bigger area, and if you go the biome route, you can opt to keep a zone circle like in Fortnite, Apex and pretty much most of the BR games, or be area based like in the Ring of Elysium.

[JackGriffin]

It's a great idea to download MH-LostInTime and playtest the large, open area of the map. I found a quick download here (destination Unreal).

Doing this you can see where the limits come into play for an open world map. It's not nearly as good as you would need it to be. Line of sight has to be constantly occluded by terrain to prevent HOM's and even then there are tricky spots in this map that cause problems. Fly up on top of any of the little hills and take a look around. It's not good.

Flay is correct, there are some good terrains out there but node count is the big limiting issue. There's so much terrain needed that you'll burn through poly count and not have the overhead for an interesting map with buildings and areas to explore.

The better solution in my eyes was to just use whatever DM map and make your adjustments like I did in BRUT. At least that way the maps that played well would tend to bubble up to the top. It's not the same experience but I wonder if the limits in this old engine are not really useful in some situation like this. The hacks to get it to work make the end result unappealing.

[TheDane]

Thank you so much for your feedback. However I might have puttet the question incorrect? I was looking for an answer like e.g. "Hey D, you should build your map in 32768x32768 UU in order to get the gameplay you are looking for".

I know lot's of the limits you all adress here, call me the KING OF HOM'S!!! I started my mapping back in the days making big boxes for sniper servers with custom high zoom rifles, so trust me, I've been there a lot! Zoning will definatly be the key to achieve a playable map, and the line of sight needs constant focus. I plan drop entry from the players, meaning full top view of the entire map.

My thoughts where to create a map spanding 32768x32768 UU, with water arround the island it provides a gaming area of about 30000x30000 UU, which is just on the border of HOM's, so correct zoning and small hacks should do the trick. My main concern is the polycount, the map need to run smooth or gameplay dies. I've got to be honest, I don't think as a developer you must look at the specs for a PC back when UT was released, in order to make it run on those rigs. Well knowing that the renderers will not make full support of moderne PC's, I do expect to gain a little extra from the increased CPU speeds on avergare PC's today, compared to the P2-333 Mhz back in '99.

So, what do you think, will I be able to deliver gameplay Fortnite'ish on a 30000x30000 UU island hosting (up to) 32 players?

[Feralidragon]

The response to that question highly depends on what exactly do you mean by "Fortnite'ish", because no matter what you do you won't be able to come even close to the overall experience Fortnite offers overall on this engine, that's a game on a tier of its own even among popular battle royale games, and that's why it's still very popular even with the competition of more recent games such as Apex Legends, and why I still play it myself a lot (and do some content times to times too).

What I mean is that Fortnite is not only about a big island with 100 players, it's also about the building, the variety of weapons and items, each biome, each POI, the forever ongoing story and events, different gametypes (LTMs), vehicles, emotes, skins, the overall happy, comical and non-serious tone of everything, etc, etc... and you probably have only a very specific set of things from there you would want to experience in UT.

If you just want a big map for a battle royale in UT, that's really no different from any other battle royale game, and if you just want it to "look" close to the Fortnite one, then the geometry is not very complicated actually, and most of it could be done in UT without issues (including instant killing HOMs, since Fortnite has some of those times to times: ).

[TheDane]

Well, it will not be a Fortnite clone, it would be Fortnite with the things I don't like removed. So, no, it won't be Fortnite as one of the things I won't be making is the IMO crappy building part. I don't find it fun or even remotely realistic when I encounter an oponent and 1 second after he/she is standing on top of a 7 storie high construction shooting a rocket at me.... I'm more the dogfight kind of guy So no, Fortnite gameplay at 100% is not what I aim for. But the variety of weapons and its well balance will be my top priority. First it will be Solo mode, but later if I still have the will to continue Duo, Squad and the likes will be added, but that will ONLY happen if someone actualy will play the game, otherwise it will be pointless. Vehicles will also be present in the end product.

The money machine of Fortnite will be opted out simply due to the fact that zero income must come from UT (ut99 EULA), but also because I hate that stuff. However that does not leave out the fun of completing challenges and recieving non advantage objects from it Just don't expect me to throw in a million daily challenges

[Dr.Flay]

@JackGriffin How big was that map DM-((B-A))Mountains ?
I seem to be missing files for it so can't open it.

[JackGriffin]

I cant find that map in my stuff Flay so I have no idea. Can't remember ever playing it either.

Dane, don't expect anyone will play it let alone it get 32 players on the same server. I barely got a handful of people to partially test BRUT and aside from HoF playing it for a night it got zero interest. That doesn't bother me, I enjoyed making the thing and that's the fun part. I just want you to be aware that you are heading down a bad road if you hang your idea of success on whether or not it gets played.

If this were me I'd very much consider doing it in 2k4 where the engine is much better suited to the limits you're going to need to have.

[TheDane]

Dane, don't expect anyone will play it let alone it get 32 players on the same server. I barely got a handful of people to partially test BRUT and aside from HoF playing it for a night it got zero interest. That doesn't bother me, I enjoyed making the thing and that's the fun part. I just want you to be aware that you are heading down a bad road if you hang your idea of success on whether or not it gets played.

If this were me I'd very much consider doing it in 2k4 where the engine is much better suited to the limits you're going to need to have.

I Know my friend I want to do a solo gamemode for the fun of it. Then only if a group of players ask me to make a team mode i'll consider it, it's not that much work once the solo mode is made. I totaly hear you, it won't be played so its just for fun.

And.... ut99 FTW! Limits are ment to be broken

[JackGriffin]

Well, it goes without saying if I can help at all I'm at your service.

[TheDane]

Well, it goes without saying if I can help at all I'm at your service.

I know my friend

Not much codewise has been done yet, only the HUD as I showed you has begun along with a few weapons to test it, it's a slow project as my sparetime for this is very limited. But... UT is close to 20 years old now, so if I can finish this in say 10 years it should still be arround and playable

Right now I'm building up the map, the terrain is a killer, I've never mastered that side of mapping, so it's very time consuming - but educational

[Dr.Flay]

Hmmm I'm sure it was you that gave me a copy of the map.
It has a lake with a wooden storage hut nearby, and a couple of towers either side of the map, and a few alpine houses dotted about the hills.
In 1 place there is a tunnel through a hill.
I remember wanting to edit it into something suited to a Day-Z type game.
It is coming back to me now, it is a very large sniper map, but not as big as it could be.

I will try and salvage it if I can, I remember I added lots of nice trees and plants including reactive grass.