Seperating Ground Cover from Trees.

[Peter Cairns]

In CM1 we get basically three types of trees, Pine forest, Deciduous, and Scattered. Plus scrub.

However there is afr more variaety than this and it is quite restrictive. What is more depending on the type and age of wood there can be huge differences in ground cover.

An alternative to wood types would be layers, where a wood was built up in 1m stages. Thus a forest cartpetted with ferms could have the first 1m very dense, and then the next three to five metres light (trunks) and then the next 3 to 5 metres dense again (canopy).

This system would mean that depending on height visibility would vary as the LOS might pass through canopy or not.

This system would also allow for the simulation of first contact where units would have limited cover, being partcally in the 1 to 2 m level, but would then be far less exposed when they dropped down in to the dense !m zone of ferns, heather or wheat.

In a way this would add a vertical element in to LOS and spotting in the same way as we currently have horizontal one.

How much the game could be designed to let scenario designers mix and match tree typres and gound cover to create more realistic and complex vegetation is open to discussion.

Peter.

[Rodent]

Do they got trees in Syria?

They only have got about 1% trees in Syria.

I for one would like to see pistachio forests modelled.

There are two notable pistachio forests in Syria, one in the middle and one on the right.

There is a small forest near Damascus.

[oren_m]

Originally posted by Rodent:

Do they got trees in Syria?

They only have got about 1% trees in Syria.

I for one would like to see pistachio forests modelled.

There are two notable pistachio forests in Syria, one in the middle and one on the right.

There is a small forest near Damascus.

Actually there are great numbers of trees and vegation in the northern parts of Syria.

Most of it are man made, in the shape of fields and plantions, but there are some woodland areas which are natural.

And also, don't foget the areas around the river Euphartes, i believe that during a war, a major part of an advancing army will fight in the areas near the river in order to clear passages to cross the river.

Oren_m

[Rodent]

Your map is better than mine.

This has nothing to do with trees, but what kind of dirts and sands and soils do they have in Syria?

I read about something like 'sabka'[?] in Iraq, that is like a tar pit covered in a layer of sand capable of bearing the weight of a man but not a vehicle...or something.

[c3k]

Hmmm,

Obviously, we need to differentiate the COVER aspects of ferns (little to none) from the CONCEALMENT aspect (very high). This being the obverse of the aforementioned "trunk zone" extending from 1 to 3-5 meters above the ground; high COVER, low CONCEALMENT.

Modelling the various calibers' penetration of various species of trees would be quite important. How does an M-4/M-16's 5.56 round penetrate a pine? How does that compare to, say, a pistachio trunk? What of the various Syrian rounds?

How do we apply the uniform rows of a man-made orchard? If the firer and target are in the same orchard lane, it would be wide open. If they are separated by a few lanes, then angular differences would be critical. One lane apart, at a range of 1,000 trees, would seem to be total COVER and total CONCEALMENT. Yet, one lane apart at a range of 10 trees would be much less. Now, what of 10 lanes apart, but firer and target are at 90 degrees to the axis of the lanes? Hmmm, has anyone thought of irrigation systems?

Regards,

Ken

[Peter Cairns]

c3k ,

Given that CM:SF has an 8x8 grid with the likes of walls able to be placed at I mtr intervals, we could have three possibilities for orchards.

One would be to build 8x8 gaps, which although 26ft gaps is a bit much would still be okay and give a good represntation of an orchard for game purposes.

Second we could creat Tree walls which would be set on a predetermined ground cover and would be a bit like wire fences ( just the posts no wire) for sighting and protection. these can be set out in lines so would be an easy way to make regular lines of trees.

Thirdly the last option is "empty walls" were a lane could be made through a forest, but inserting a "wall" made of thin air. This would be a way to add paths, trails or roads through dnese ground cover or trees.

On a general note I wasn't particularly thinking about Syria, as opposed to best way for the new engine to deal with ground cover and trees of various types as opposed to the old CM1 three types.

Peter.

[mav1]

Hopefully LOS and cover will be more realistic in cmx2. I would also like to see more complex abd realistic vegetation than the limited choice in cmx1. I know that urban areas will be much more complex in cmx2, I wonder if vegetation will match it. It all comes down to number crunching, what will be possible with computers at the moment.

Also much small rises in height will add depth to cover, by creating small dips in land to hide behind.

I have to admit, I would like to see the vegetation (amazing trees) ftom Elder Scrolls Oblivion in CMx2, but that's me dreaming again.

[Cpl Steiner]

Does anyone know if trees will be individual objects in CMx2 or will they just be placed randomly within the designated "woods" terrain tiles as in CMx1?

My feeling is that 1:1 representation of soldiers will not extend to precise positioning of soldiers behind real physical objects but I could be wrong. I imagine that when a man comes under fire in a wood in CMx2 he will get the cover benefit whether or not it looks like the line of fire crosses an individual tree, just as in CMx1.

[mav1]

I doub't that trees will be individual objects in cmx2 and i guess you won't be able to precisely position soldiers behind real physical objects. All this would be very welcome and would add to the realism. But I don't think computer's are good enough yet to handle such complex calculations, but I could be wrong.

But I think the grid for cmx2 could be smaller, 8x8 is an improvement but i reckon it's still too big to create very detailed enviroments.

On a different topic, the developers were thinking of doing Napolonic or civil war etc.

But I wonder how big the armies would be. Are computers powerfull enough at the moment to create large armies, or will they be division size at most.

[mav1]

Does anyone know where I can find a website with topography maps, which will allow me to print a map for free. I need the info to create historic maps.

[Dogface]

topo maps of where? US? Europe? Middle East? Africa?

I'd try Google to start. Though its hard to find free maps that have any good info on them.

May be your local library will have something. (Published AAR's W/ maps)

[MikeyD]

BFC's been saying that we'll be 'painting' the terrain features onto the maps. So at least for forests they'll probably be brush-applied and the computer figures out where to place 'em. Maybe small brushes will do individual trees? In Drop Team I believe at least some trees on the map are individual - able to push 'em down & blow 'em up - instead of CMx1 abstract representations. I rather suspect a forested mountainside of individual trees could slow the game a bit!

[Cpl Steiner]

I believe the LOS system used in CMx2 will be the same, in principle, as that used in CMx1, i.e. a LOS map will be loaded into memory such that checking LOS from one point to another is merely a case of looking up the result for the two points.

If this is the system used (and I believe it is, as BFC have said that mobile objects like vehicles will not block LOS because they won't be on the LOS table) then the next question is, how big is each "point" on the LOS map?

In theory, you could go down to smaller and smaller points and generate bigger and bigger LOS maps. The only downside would be the amount of time taken to generate the LOS map and the amount of memory available to store it. Having said that, computers have much faster processors and much more memory than they used to.

I am pushing the boat out here and saying it should be possible to go down to a metre or less for each point on the LOS map. If BFC manage to achieve this level of precision, then the holy grail of individual soldiers gaining cover from individual trees might just be possible.

Just to reiterate, this is pure speculation given the way LOS is done in CMx1 and advances in computing power.

[MikeyD]

I'm all confused. Which did they say vehicles now would block in the new engine? Line-of-sight but not arms fire, or arms fire but not line-of-sight? I recall people were hoping to be able to huddle behind accompanying tanks while advancing for protection.

[Peter Cairns]

If the game was to be designed so that the LOS mapping time was usefully employed it might not be such a problem.

two possibles spring to mind. A flyover?through battlefield preview, covering objectives etc, to fill the time, or,

Have it so that you can start deploying forces or organising the exact TOE options that you want after you have created the map. ie You would br making force selections( which shouldn't be that CPU intensive) while the computer was doing the difficult CPU intensive LOS map in the background.

I am generally not sure if hiding behind individual trees etc is the wauy to go, as it would probably lead people too micromanage their squads too much, and CM is more about giving orders and seeing them realistically carried out than an over glorified FPS.

Besides of all the bullet stoppers nothing beats mother earth, and as I recall a 12" sandbag is better protection than a 12" log, because of the way that it disperses energy.

The point about being able to simulate dips and bumps in the ground, and things like ditches by the side of the road, are for me far more important than individual trees.

Peter.

[Cpl Steiner]

Originally posted by MikeyD:

I'm all confused. Which did they say vehicles now would block in the new engine? Line-of-sight but not arms fire, or arms fire but not line-of-sight? I recall people were hoping to be able to huddle behind accompanying tanks while advancing for protection.

To save you the trouble of a search, here's what Steve said.

The example in the linked thread involves moving units. Can they block LOS? No. Can they block LOF? Yes. The question is... if the system knows that the vehicle is in a specific spot, why can't it treat it like a piece of blocking terrain and therefore not see what is behind it? After all, the system knows that it can't shoot through that vehicle, so why can't it also know it can't look through it? Remember what I said above about the very clever shortcuts? Remember that I said they are essential for managing the hit to the hardware for all those LOS checks? That's where the answer lies.

One of the most important shortcuts is establishing a "LOS map" of the entire battlefield which is, basically, a precomputed LOS check. Units don't scan the terrain map directly, they scan the LOS map. So it doesn't matter what is on the terrain map, it matters what is in the LOS map. Since the LOS map is precomputed, it can't possibly know about things that move around dynamically since that would require constant recomputing the LOS map data. So much so that it would probably negate the reasons for having the LOS map in the first place. Which brings us back to the point about this shortcut being necessary for the game to run at all.

-Steve

The point is, LOS is precomputed but LOF is done dynamically. All this means, however, is that units that can't be shot can still sometimes be seen (which is not ideal, for sure).

One thing I don't understand though is how the game knows that a house blocks LOS when it is standing but doesn't block LOS when it has been reduced to rubble. Obviously my understanding of all this is somewhat incomplete.

Even if the game could somehow do LOS and LOF checks for individual soldiers behind trees, logs, sandbags and the like, I agree that it wouldn't be a good idea to allow the player to position soldiers individually to take advantage of various items of cover. Hopefully, if such things were possible, the AI would also be well capable of positioning soldiers optimally.

[Peter Cairns]

Cpl Steiner,

The LOS problem for buildings, might be the same as moving vehicles in a way.

If the LOS map only takes in to account the physical contours of the ground, then if two units have a direct line between them then it will check for LOS/LOF and only then. At this point it will determine if a building blocks it.

Thus if a building was on a hill between to units it wouldn't check as they couldn't see each other whether their was a building or not. If however it was a building between two units on hills, then it would check for "potential" sighting, and would therefore determine if the building "as is" blocks LOS. This would depend on the condition of the building it's height etc.

however if this was the case, it wouldn't explain the fact that Vehicles can't do the same.

I can only come up with the idea that when the CBU crunches the LOS map it divides it in to two class "never" and "Check" and that building and other deformable features are " Checkable" but vehicles aren't because they had a fixed position that could be "recorded" at the time the LOS Map was constructed.

Thus Hills are "Fixed Never", Buildings are "Fixed Check" and Vehicles are " Mobile Can't".

Question, do wrecked vehicles put in by the game designer, or dug in Vehicles block LOS because they were fixed at the point of map creation, and what does that do for my idea of doing the TOE while the computer id mapping LOS.

AND SHOULD GROUND COVER AND TREE TYPE BE SEPARATE......

Peter.

[Cpl Steiner]

Peter,

This is really puzzling me now. If, as you say, the LOS map says paths though buildings are clear but the game engine does an extra check to see if the building is standing or rubbled, then it should be possible to do this for buildings and wrecks. In fact, it should be easier, as there are usually many more buildings than vehicles in a scenario. What is so different about buildings?

Okay, for starters the game knows they will never move. Therefore, they can be factored into the LOS map somehow. The question is how?

Maybe the LOS map is a bit smarter in what it returns than just giving a "clear/not clear" result. In theory it could return an array of blocking points along the LOS check path. Each point could then be examined to determine if it is part of a building, and if so, if the building is still standing.

Come to think of it, this must be how it works, as the LOS tool is shaded in different colours to show where LOS is blocked, and there is no delay. It must just take the start point and the end point, do a look-up on the LOS map to return an array of blocking points, and examine them to see if any are buildings that have been rubbled.

[ April 14, 2006, 03:18 AM: Message edited by: Cpl Steiner ]

[Peter Cairns]

Cpl Steiner,

pretty much how I see it too, only I see two classes. Always blocked, ie that hill will always be their and "checkable" ie, it has buildings in the way that might be destroyed.

By using two classes a huge number of possibles can be eliminated right way with a flat "Never" result. The way to test this theory would be too build a series identically sized maps, with varaitions in the number and size of buildings and hills and see how long the CPU took to calculate the LOS map.

Again as vehicles aren't known or fixed at the time the LOS map is constructed they can't be built in. Although Wrecks could.

Which leaves us with the issue of "dug in Buildings and bunkers, which in theory should block sight, but might not if they are added after the LOS map is made.

Peter.

[Colin]

This thread makes you appreciate just how talented Charles is.

SCREENS?!?!!??

[Peter Cairns]

Colin,

Crawler.... Anyway how talented can he be if I've worked it out....

Peter.

[Cpl Steiner]

Here's a suggestion for how you might go about having the LOS map take account of wrecked vehicles mid-game.

Whenever a vehicle is wrecked, the game could initiate some sort of slow, low-resource-hit, background process that started to generate an entirely new LOS map - one which took account of the wreck. This would only need to recalculate LOS crossing the wreck for any two points for which LOS wasn't already blocked, so it would already be doing substantially less than for the original LOS map. The new LOS map would be placed in a spare block of memory and not used by the game. Once the background process had finished, it would tell the game that the new LOS map was available, and the game could then instantly switch over to the new map. This might mean a wreck didn't block LOS for a few seconds but I for one could easily live with that.

Before anyone jumps all over me and tells me I shouldn't be second-guessing Charles' work, I agree. Charles will no doubt have already thought of the best way to do this if it was possible. I'm just speculating idly until we get some screenshots!

[ April 17, 2006, 03:19 AM: Message edited by: Cpl Steiner ]

[cassh]

On a similar vein to LOS checks I heard talk of real-world handheld GPS/digi-map units that allows one to check intervisability between any two given points and therefore the ability to quickly and easily indicate deadground and approach marches when advancing on an enemy position in unfamiliar terrain?

Does anyone know whether is in use yet?

[Cpl Steiner]

Originally posted by cassh:

On a similar vein to LOS checks I heard talk of real-world handheld GPS/digi-map units that allows one to check intervisability between any two given points and therefore the ability to quickly and easily indicate deadground and approach marches when advancing on an enemy position in unfamiliar terrain?

Does anyone know whether is in use yet?

What will they think of next! GPS seems to pop up everywhere on the modern battlefield. I wonder how long it will be before the terrorists and insurgents are using GPS guided mortar rounds and the like.

If this is in widespread use by US forces, perhaps in CM:SF the US player should be able to shade parts of the map between two points to indicate dead ground.

[Peter Cairns]

Why don't we go for the Hollywood option, Wrecked vehicles don't block sight because even if hit by only three rounds from a 9mm Colt, they explode in a fireball and fly in to a million pieces,

Peter.