Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
The slit-trench and the objective will be located on the smaller slope on the right. I plan to connect the slit trench and the mg-nest with a relatively exposed track. This could be an interesting tactical factor because ammo for the VC will only be available at one of those two positions.
https://ibb.co/c6NT5m
heightmap overview (wip)
https://ibb.co/kYefJ6
As you can see, I'm still working on the heightmap before committing to static meshes and volumes/water.
https://www.youtube.com/watch?v=815RlEJvFkc&feature=youtu.be
---------------------------------------------------------------------------
This video shows the camera jittering problem. I want to slow down players who move through dense foliage or on steep slopes. I created a physics volume and increased ground friction. Now players do get slowed down, but the camera jittering is awefull. :(
https://www.youtube.com/watch?v=yNyFR1ntnQE&feature=youtu.be
I like that fighting aginst the sun will cause real troubles for the marines.
EDIT: Lol I just watched the youtube video and here there is no camera-jittering at all when the player enters the undergrowth. That's interesting .... it's present when playing the game, but the recorded video doesn't show it.
----------------------------------------------------------------------------
I've dropped some grenades close to me to see whether small bumps in the ground do make a difference. And I was pleasantly surprised that they do. So, by adding more little bumps in the ground which stop shrapnel, grenades will be less deadly for prone soldiers.
https://ibb.co/dZqLLm
Note that some water puddles are still missing.On the left the jungle ends, thus the trees' height slowly falls off, trunks are getting thinner but more numbersome. To the right we'll be going deeper into the jungle, with very high, thick but fewer trees. There is a lot of undergrowth of ferns and moss (unforunately, moss only exists as material, not as static meshes) - they don't need a lot of light to grow.
I want a path to lead from the road through the jungle to the river crossing. Using this path will have the advantage that it does not slow down movement like the rest of the jungle. But of course the path will be a priority tactical target (punji traps hello!).
Work in progress jungle:
https://www.youtube.com/watch?v=z5tKjj-rIf0&feature=youtu.be
https://www.youtube.com/watch?v=e0ki3XbkXGc&feature=youtu.be
Almost all trees for the jungle have been placed. The work on the ground (undulations, rocks, roots, plants/undergrowth) is still to come.
The rest of this side of the river has been filled with elephant grass. Not sure yet whether I like it, plus there is a problem: Even at the maximum drawdistance, the grass disappears for Vietcong soldiers who are on the very far extreme on their side of the map. This is a real game-breaker. I will see how I can solve this. I guess I'll have to make due with terrain which is not supposed to provide concealment (and reduce the height of the elephant grass so that no player thinks it provides concealment...).
I can still use elephant grass (perhaps not that tall) on the other side of the river. Here the drawdistance doesn't matter that much because the US doesn't get a very long LOS on the hills anyway (due to the jungle).
Here are two views of the jungle-patch with some bushy area in front of it. As I've stated before, the draw distance of the grass is not ideal even if maxed out, but on the other hand even if a very far away enemy might see people who should be concealed in the grass, he still can't hit them at this long distance. Moreover, the bushes still provide concealment - even at far distances. So I think it will be fine.
https://i.imgur.com/trtVWuT.jpg
https://i.imgur.com/ifQ5vTK.jpg
Also, I've successfully added spawn cameras and ammo crates.
Before I re-start placing the meshes, I might make some adjustments to the layout of the map and also consider carefully the performance-stats of the meshes I use. I'm totally new to mapmaking in a 3d engine, so I don't really know what to watch out for. I'd be very glad for tips.
Complexity of the static meshes (tris, vertices) and LODs
So obviously, the fewer tris and vertices a mesh has, the better it is in terms of performance. Here, some models offer a beter mix of complexity and looks than others. So I try to stick to the well balanced meshes more than I used to. Also, I try getting rid of the practice of reducing the scale of big trees to simulate smaller bushy trees, and of placing trees "under" the map to make their branches simulate bushes. I think the trunks are still there and draw ressources.
Also, I might play around with the Levels of Detail (LODs). As the LOD-settings are set per static-mesh-object, and not per static mesh actor on the map, I suppose I have to copy the models to my map's pack and edit them there. I did a few experiments with some objects, changing their LODDistanceRatio and LODMaxRange and it led to some ugly blurring/popping of forms and textures very quickly. But I still think I can perhaps slightly decrease LOD transition ranges for objects that I place in short-LOS territory like the jungle. Especially considering that the movement speed of the players will be reduced as well, the popping issue should be less extreme.
I think it would be cool if there were seperate nodels for trunks and for foliage (the latter drastically reduced in complexity). I think it's a bit unfortunate to have so many detailed leafs rendered so high up in the air.
Collision models
Obviously some trees and bushes have collision boxes around them (stopping both movement and bullets, I suppose) while others don't. I'm not sure if the presence of a lot of collision boxes this has a large effect on performance, but I guess so. I noticed quite some lag when I fired into an area of lots of small collision boxes. So I might turn some of them off and/or use more smaller branches/bushes than bullet-stopping larger trunks.
Lightning
Not sure how much influence this has on performance. I could try to add some tree models that don't throw any shadow for the interior jungle parts where there is no light anyway. Not sure if I can increase performance in any other way here?
PS: CULL DISTANCE VOLUMES
PS: Usefull: https://docs.unrealengine.com/latest/INT/Engine/Performance/Guidelines/index.html
https://www.youtube.com/watch?v=Q1a0ruRTYVg&feature=youtu.be
https://i.imgur.com/xzSwHDc.jpg
The main problem I encountered was that one needs very sharp slopes (90°) because the players need to be able to step as close to the edge of the trench - and the loophole respectively - as possible. Every little inch of distance between the player and the loophole drastically decreases the angle of the field of fire. Unfortunately, you can't create very sharp (90°) slopes for trenches by terraforming the height-map (whose verticaes are lined up on a grid). So I decided to use the BSP-brush and create a box, which has a 90° angle. I've also turned off the colision boxes of the wooden supporting logs to make sure they don't get between the player and the loophole. I had to keep the collision-boxes for the sandbags though because when I turned them off, the player always tried to set up his MG on the BSP-box. The end-result is a fire position that provides an okay field of fire (the trench needs to be placed somewhere else, so that there is less dead ground) and forces your enemies to hit a small window of perhaps 30x30cm. I don't know yet how resiliant the position is against suppression fire. At the ranges at which the mg will be firing (350-500 meters), suppression can reduce it's effectiveness drasticall, while at the same time, it should be quite hard to hit the loophole (the map has no slots for snipers).
The overall position of the trench in the video is not final. There is lots of dead ground in front of it right now and the mg can only reach the far bank with plunging fire. I could turn it into a really mean position by adjusting the slope all the way down to the river so that the MG can deliver grazing fire (with only very few dead spots).
Also, you wouldn't want that cliff directly behind you (anything hitting there will rain shrapnel on you from above/behind). Moreover, the trench here is still lacking exits, some logs laid out on the floor, the supporting logs on the backside and some kind of roof (camo-net vs. aerial recon [recon will not be allowed on the map anyways], something that creates shadows so that enemies can't see if the position is occupied or not). I've just tried to nail down the tricky game-play part, not the aesthetical aspect. (I'll also get rid of the gap between my BSP-box and the heightmap-terrain).
I also need to point out that the trench is broader than I'd like it to be. But I can't make it any slimmer (unless I increase the tesselation of the terrain) because players would have difficulties to move in it. On the other hand, the map will not allow arty, so that is not too much of a problem (still grenades can hit the trench more easily though).
The other VC-position on the other hill will not be that delicate. I only plan some quickly shoveled foxholes, not a sophisticated trench with sandbag-cover.
-------------------------------
I've also restarted the jungle, carefull to use unlit models that will not cast shadows wherever it is reasonable to save resources.
-------------------------------
PS: Interesting instruction-video on the use of heavy MGs in the attack, 1937 and 1944: https://www.youtube.com/watch?v=DnWuT_ArzeE
https://www.youtube.com/watch?v=9_HYmcm9A2o
------------------------------
PS2: Screenshots of an updated trench
https://i.imgur.com/hnLDYoJ.jpg
https://i.imgur.com/3gwl1Bk.jpg
The only thing still missing is an option to leave the bunker (except for the tunnel...) and more concealment (some stuff on the roof, primarily).
I'd really like to see and test how much suppression you can get on the bunker (with one, two ordinary riflemen, with the grenadier...)
https://www.youtube.com/watch?v=5dG9LtyxQ48&feature=youtu.be
Right now I'm working on the VietCong positions.
------------------------------
For anyone interested in bunker-realism/positioning, I can recommend one of the videos of one of my favourite youtubers:
https://www.youtube.com/watch?v=l5PnGgUvJiQ
------------------------
Another interesting read I've discovered today: observations on infantry operations in the Korean War:
http://cgsc.contentdm.oclc.org/cdm/ref/collection/p4013coll11/id/1350
---------------------
I'm working on my grassy hill, placing bushes and grass-groups, and working on a trail that curves up the hill.
---------------------
I also toy with the idea of implemeting a first objective called "rally and organize" that has a miniimum cap-time of 3 minutes or so and will be placed almost immediately at the spawn-zone of the attackers. The idea behind this is that players should get a little bit of time (3 minutes?) to prepare and organize. The problem is that even when working with map-boundaries (triggered by capturing objectives) it doesn't prevent players from avancing? Perhaps I can create blocking volumes that disappear once the first objective is taken? I will see what's possible. I just don't like that there is no preparation time at all. Players should at least have enough time to find their proper squad and to distribute themselves over the front.
This first objective could also be handy as I want to give the defenders two spawn-options in the beginning, but only 1 once the game has started.
------------------
Here is a list of some of the points I keep in mind for creating a realism-oriented map:
PS: Military maps on Vietnam are very handy here: http://www.lib.utexas.edu/maps/topo/vietnam/
--------------------
* It's woth noting though that concealment of fortified positions is impossible in a game. After playing a map for the first time, you know where everything is.
------------------
PS: Doesn't look as if you can survive arty, which for the americans fires airburst ammo. So that's out of the game. This is just too easy: you know where the VC positions are - just drop arty on them, game won.
-----------------
My new map will be smaller. I will keep the distance high (ca. 250 meters) but scale down the width of the battlefield. The scenario: A patrol of US. forces has dug in for the night in the vicinity of some montagnard-huts. As it is but a temporary position, the troopers have only dug some foxholes, no sophisticated trenches or bunkers, maybe they have also cleared some important firing lanes. The northern forces, "streaming" out of the jungle, attack this night-camp through the dense brushwood/grass that separates a trail from the huts. (maybe it would be interesting to have a night or dawn scenario)
It will still be an asymmetrical scenario:
Note to self: 1 unreal unit = 2 cm