A bit confused about how LOS works sometimes...

[noxnoctum]

So I was playing the "In Search of a Ghost" campaign by Webwings and in the first mission you have to take out 2 heavy MG positions (and try to minimize casualties as much as possible as you don't get replacements). To make a long story short, I was going to have my main attack group (2 regular squads, and one command squad with 8 people, all with basic M16s) flank from the left, I had an mg and its commander ready to pop up over the hill directly in front of the building (about 150 meters away), and I had a squad and an MG on the right flank hidden behind a hill who I planned to have pop out and suppress the building so that the main group could assault it.

Now, one great feature I love of CMx2 is that you can do LOS spotting from movement markers, so you know exactly whether or not that place you're going to move up to will have LOS to an important area. Thing is, in this situation, it was completely incorrect. ALL 3 groups (I simultaneously had them come behind their cover to open fire) were able to spot the enemy held building at a point further away than the game had told me they could spot it (once I launched the assault). This resulted in my covering elements unnecessarily exposing themselves (because I'd ordered them to move to a position that the game said they would DEFINITELY be able to spot from) and undoubtedly taking more casualties than they would have otherwise.

Now, I'm only talking a difference of about 5 meters here, but the thing is, if the game had told me, correctly, that I'd be able to spot from the further away position, they would have been MUCH less exposed (in the case of the two support elements I would not have moved them over their respective hills as much). What's up with this?

[noxnoctum]

Ok here's a completely different situation, and one that's also very odd. As you can see in the picture all 3 soldiers can clearly see down into the ground below yet they get a magenta/blue blocked LOS! Also note they're in perfect C2, and not suppressed at all (they haven't been fired at once the entire game). AND it's a sniper team :confused:

[Jonny(FGM)]

I'm not sure about the first example, but I believe in the second example the fact that the units are prone could have something to do with it.

Quite often I'll plot a move to behind a wall, or in a field or the like, check the LOS there and once the units get there and go prone they lose that LOS as the LOS is now calculated from the prone position but when you check it on the waypoint it is calculating it (presumably) from the stood position.

Having said that it does look like those guys could clearly see that ground you are pointing too so that may be a little bug.

[Thomm]

Having said that it does look like those guys could clearly see that ground you are pointing too so that may be a little bug.

The image looks very familiar to me and some tests I made a long time ago. I think the conclusion back then was that it was not a bug in the strict sense, but rather a shortcoming of the engine and the LOS calculations.

That being said, it would not harm if you had a save game for a CM:SF tester to pick up.

Best regards,

Thomm

[Jonny(FGM)]

Where is LOS calculated from on a soldiers body?

[noxnoctum]

The image looks very familiar to me and some tests I made a long time ago. I think the conclusion back then was that it was not a bug in the strict sense, but rather a shortcoming of the engine and the LOS calculations.

That being said, it would not harm if you had a save game for a CM:SF tester to pick up.

Best regards,

Thomm

Actually I saved it right after I took the SS at that exact moment. Who should I submit it to?

[noxnoctum]

Where is LOS calculated from on a soldiers body?

From their feet apparently

[hoolaman]

Where is LOS calculated from on a soldiers body?

LOS is based on the action spots and the height of the soldiers (prone, kneeling, standing). It is not based on individual soldiers.

In your first situation, I believe the waypoint LOS hieght is based on the current height of your soldiers, so if they are all prone, your waypoint LOS trace will be done at prone height. When you actually arrive there by a fast or quick order, the LOS height is quite a bit higher and your troops are more exposed. Try using SLOW for the last bit if you want to sneak up to a waypoint.

In the second, the LOS is traced from centre tile to centre tile, so unfortunately this is an issue in some of these situations looking down into a valley. It's not very good.

[noxnoctum]

In the second, the LOS is traced from centre tile to centre tile, so unfortunately this is an issue in some of these situations looking down into a valley. It's not very good.

Ah that explains it. Well I gotta say that really sucks. If you're gonna do 1-1 then at least make it actually 1-1! It's absurd because in that situation it would mean I would have to move my sniper team OUT of cover (behind the edge of the hill but where they could still CLEARLY fire over it as you can see in the screenie) and down a hex, completely exposed on the side of the hill. Pretty retarded.

I hope that gets changed. Probably won't be in CMN. But by the time CMx2:Ostfront rolls around I should hope so.

[Jonny(FGM)]

Might be worth putting some enemy soldiers in that spot where it looks like they should be able to see but according to the LOS tool can't.

Ive seen many a time where a unit can spot an enemy unit, and fire upon it even tho the LOS tool says they can't see there

[Thomm]

It is not based on individual soldiers.

I think the final stages of the LOS checking process are based on individual soldiers, however ...

In the second, the LOS is traced from centre tile to centre tile, so unfortunately this is an issue in some of these situations looking down into a valley.

This is correct, I think. So the tile-to-tile check fails first and then the LOS from the soldiers is not checked either, accordingly.

Best regards,

Thomm

[Thomm]

Well I gotta say that really sucks. If you're gonna do 1-1 then at least make it actually 1-1! It's absurd because in that situation it would mean I would have to move my sniper team OUT of cover (behind the edge of the hill but where they could still CLEARLY fire over it as you can see in the screenie) and down a hex, completely exposed on the side of the hill. Pretty retarded.

Good summary, I'd say.

Best regards,

Thomm

[jnt62006]

Even more frustrating are the vehicles. Center line for a Bradley is higher than center line for a BMP or even a T72. Fortunately all of its optics are just inches from the top of the turret, making it a great firing platform when hull down. But the game mechanics negate this advantage. In order for a Brad to have LOS to an enemy it has to expose most of the vehicle. Making its height a huge disadvantage.

[akd]

This can often pose problems, and I prefer not to play on maps with steep, broken terrain because the advantages of seeking high ground will often be negated.

[Guest Guest]

deleted per user request

[noxnoctum]

By "explanatory posts" do you mean saying "Yes we're working on correcting that and it will take a while" (fine by me, I know they're a tiny team), or trying to explain it away through some contorted logic?

Don't ban me please

[Frankster65]

Well I got to admit...this is disconcerting. So much for advantage of the high-ground.

[dan/california]

Has anybody checked if making the ridge line less than absolutely knife sharp helps.? It may be simple matter of slightly better map making. Not to say the engine shouldn't be able too handle it better, but it would define the scope of the problem.

[Gpig]

Once those dudes on the "kinfe edge" take a kneeling position, they'll have LOS/LOF. But, it'd be nice to be able to remain prone and have a go . . .

Gpig

[stikkypixie]

It's probably possible to eliminate these kinds of situations, but then we'd all have to have a high performance clusters to calculate the turns

[Alan8325]

By "explanatory posts" do you mean saying "Yes we're working on correcting that and it will take a while" (fine by me, I know they're a tiny team), or trying to explain it away through some contorted logic?

Don't ban me please

I believe that there was something in those old explanatory posts that said that LOS fidelity would improve over time as computing power improves. Perhaps it would happen at around the same time as multi-core support or something. I definitely am one who hopes it gets better!

[Guest Guest]

deleted per user request

[noxnoctum]

Ya I just don't get his argument of how it relates to processor power. All they have to do is change the LOS point from the hex/action square/whatever it's called, to the individual soldiers' heads. They're die hard on this 1-1 thing, and that's fine, but if they're going to go that route, they need to do it right. As it is, high ground's basically useless.

And this isn't about LOF which he seemed to be going on about. This is pure LOS.

And what are you saying about their "cover matches with the lack of LOS". This has not been fixed. This is 1.30 I'm playing and it's not fixed.

[Battlefront.com]

Ya I just don't get his argument of how it relates to processor power.

That's probably because you're not a game programmer I don't understand why we can't have cars that get 100mpg and are as safe to drive as the ones that get only 30mpg. But apparently there are reasons for it, beyond the obvious oil-car industry conspiracy

All they have to do is change the LOS point from the hex/action square/whatever it's called, to the individual soldiers' heads.

And all car manufactures have to do is burn fuel more efficiently. It's really that simple!

Seriously, you have no idea what you're suggesting. This post of mine, in Adam's linked thread, explains things in more detail:

http://www.battlefront.com/community/showpost.php?p=1144528&postcount=54

This is not a "simple" thing to address and still have CPU cycles left to run the game at an adequate speed. Ask any programmer of any game what they think of LOS checks and hits to performance and they'll all tell you similar tales of woe.

They're die hard on this 1-1 thing, and that's fine, but if they're going to go that route, they need to do it right.

Bad logic. By this logic we shouldn't make games at all because it is impossible to make a simulation of real combat that doesn't have abstractions, flaws, or other undesirable features.

As it is, high ground's basically useless.

Under certain circumstances, I'm sure that can happen. But it's not a for-sure-given.

And this isn't about LOF which he seemed to be going on about. This is pure LOS.

LOS and LOF are identical in this specific context.

And what are you saying about their "cover matches with the lack of LOS". This has not been fixed. This is 1.30 I'm playing and it's not fixed.

We will, of course, look into this more. But note that I never said it was "fixed" in the sense that there will never be any cause for complaint. What I said it was improved, which means it was better than it was before. There's a difference.

Steve

[ChrisND]

Under certain circumstances, I'm sure that can happen. But it's not a for-sure-given.

I do run into this specific problem quite a bit, Steve (as in, like the screenshot on the second post). Not every scenario mind you, but it is sure to show up if there are tall hills around. It's pretty frustrating when an ATGM on the high ground has very clear LOS to the human eye, but the game decides that there is a ten foot wall in front of them. OK, that was a bit of hyperbole, but you get my point.

The problem doesn't crop up for me if they are kneeling instead of prone.