Trees on roads block LOS

[Rokko]

Something odd I noticed recently.

I opened an original CMBN map (Pleasantly Shaded Woodland) which features a sunken lane with trees placed both ON the street and next to street in the editor.

The game is smart enough though to not place them on the road in 3D mode but left or right next to it.

Strangely though, vehicles do not have LOS "through" those trees.

The effect is that vehicles can't spot each other despite having unobstructed view and even on shortest ranges. With infantry the problem doesn't really exist I've noticed.

[Thomm]

This is clearly undesireable.

What you are seeing may be a consequence of the fact that the engine treats the foliage as a cylinder that reaches all the way to the ground (I would be happy if this is actually not the case, but as far as I recall it is set up that way).

The "foliage cylinders" may be close enough to touch or overlap in your case, thus blocking the view even though the 3D graphics show a gap between ground and foliage.

Best regards,

Thomm

[dieseltaylor]

Yikes. Sure going to make scenario design checks important.

[womble]

What you are seeing may be a consequence of the fact that the engine treats the foliage as a cylinder that reaches all the way to the ground (I would be happy if this is actually not the case, but as far as I recall it is set up that way).

If it were generally that way, infantry would have the same problem, and you'd never get some of the LOS I've seen through wooded areas with a dense 'canopy'. It might be an issue with some specific tree models, I guess.

[Rokko]

Well its Tree Model A which is the Most Common One in the Game.

[Rokko]

And I've noticed another thing:

LOS past a single tree on the road is possible, but as soon as another tree is added - the distance to the first one doesn't matter here, it could be 1 km away - the LOS is blocked.

Edit:

Okay, some furher testing brings to me the conclusion that the tree type does have some influence on the matter.

Trees B & E (The very high ones) don't obstruct LOS at all.

Tree A does obstruct it heavily (the 1km-away-thingee), I also tried it with a Kübelwagen (lower profile than the StuH), it also didn't have LOS

Trees C and D also obstruct LOS but not as bad as Tree A, I have no idea why, since Tree D is actually lower than Tree A.

[c3k]

Rokko,

Thanks for the detailed posts with pictures, and the followup.

Ken

[Childress]

And I've noticed another thing:

Thanks a lot, Rokko. You may have delayed the patch another week.

[Rokko]

Thanks a lot, Rokko. You may have delayed the patch another week.

I sure hope I did not

Honestly, this is really a minor issue. I mean how many maps have trees on roads? I just thought I had found a way to make some super-duper bocage road.

If it turns out that there is something to fix, it can wait till the next patch

[Battlefront.com]

It is interesting, indeed. I checked with Charles and he said there is nothing in the code that should be producing this effect. That means one of two things:

1. For whatever reason it's fixed for v1.01 even though he doesn't remember, or even know, that he fixed it

2. There's a bug deeper down in.

Either way, we're looking into it.

Steve

[Dingchavez]

It is soooo nasty of you, teasing us like that about 1.01 is mean :-p

[PhantomApple]

Nice catch.

[Javaslinger]

Release the patch.... or I will release the Kraken!

[Rokko]

It is interesting, indeed. I checked with Charles and he said there is nothing in the code that should be producing this effect. That means one of two things:

1. For whatever reason it's fixed for v1.01 even though he doesn't remember, or even know, that he fixed it

2. There's a bug deeper down in.

Either way, we're looking into it.

Steve

How about you give me the patch and I check for you if the problem's still there

[Rokko]

I just checked with 1.01 and the issue still seems to remain.

Also with infantry, but it appears that the lower the silhouette the better

the LOS is.

I checked trees A and C on gravel road with one tree every two tiles of the road.

LOS also improves if tree density is further reduced.

[GreenAsJade]

Release the patch.... or I will release the Kraken!

http://fledderjohn.bandcamp.com/track/the-kraken-wakes

(I recently saw this performed live, it wierded me out. You went and reminded me!

"When I awake, I brush my teeth, eat toast, watch the news...")

[Rokko]

Has anything been done about this issue for 1.10?

I mean it's not a lets say technique done often by mappers (for obvious reasons, though the effect looks rather nice), but for instance "Pleasantly Shaded Woodland" has trees on roads.

Yet I'd say this problems might reveal a more fundamental with foliage and the way LOS is calculated in the CMx2 engine that would need someone looking into.

[IanL]

We don't know the list of changes in 1.1 yet but Moon did throw us a bone saying that the list of changes would be out this weekend:

http://www.battlefront.com/community/showpost.php?p=1344604&postcount=3

OK more like he waved a bone under our noses for us to smell and beg for:)

[usgubgub]

I ran into this problem in another scenario. I edited the trees out in the map. Putting the right kind of trees in the hedge tiles adjacent and parallel to a road produces a very good effect, without the need to put trees in the road tile.

One could imagine that putting trees in the road tile represents a lane where foliage is so thick as to limit LOS along the road axis. This makes sense. If the scenario designer or the end users prefer to simulate a road with clear LOS along its axis, take the trees out of the road tiles.

[Praetori]

Maybe a stupid question. But does it matter if you place the tree in the tile before placing the road or vice versa?

[Bulletpoint]

On 7/15/2011 at 0:55 AM, Battlefront.com said:

 

It is interesting, indeed. I checked with Charles and he said there is nothing in the code that should be producing this effect. That means one of two things:

 

1. For whatever reason it's fixed for v1.01 even though he doesn't remember, or even know, that he fixed it

 

2. There's a bug deeper down in.

 

Either way, we're looking into it.

 

So here we are, six years later, and it seems this problem is still around. At least as of version 3.12. Was it fixed in 4.0 or did you give up trying to solve it ?