Proof of bullet drop in the game

[evilzucchini]

Re: Proof of bullet drop in the game

The bullet would hit the ground in 0.31 seconds.

In his defense, he did say "if you ever hit him."

However, if you ignore the fact that it won't reach the bot because it would hit the ground first or you take away the vertical acceleration, it would take 5 seconds to reach the bot on the horizontal vector (1 m/s * 5 s = 5m).

The question is, if the bullet were travelling 1 m/s, how high above the head of the bot would the gun need to be to get a head shot from 5 ms (horizonatlly) away?

Extra credit: At what angle do you need to fire the gun from level with the bot's head to get a head shot when you are 5m away?

[Kaishu]

Re: Proof of bullet drop in the game

Agreed Lizzle, I too am no PhD, but as far as I can remember from school and reading ballistics related stuff on the web, gravity starts to affect the bullet as soon as it leaves the muzzle and [as long as its parallel with the ground] is not affected by its forward velocity, so it should hit the ground in the same time whether its travelling 1 mph or 1000 mph.

But as I said.. Im no PhD and could be way off the mark.


I also remember reading that because of this, the barrels of guns are tilted slightly in relation to the sights, so that the bullet leaves the muzzel in a slight upwared motion and travels to the target in a parabola.
</geeky gun fact>


PS.. It REALLY ****es me off on this forum when people argue about real life stuff when this is clearly A GAME.. so just to iterate, Im posting this in relation to the physics talk here, not arguing the point of realism

[imported_TEG177]

Re: Proof of bullet drop in the game

As realistic as this game is, wouldn't you doubters figure that EA would be smart enough to make guns stop firing at a certain range? The real question is can we form a chart with all the gun's firing ranges

[Chet Spartan]

Re: Proof of bullet drop in the game

You can even speed up C4 packs to have a velocity of 1000, so you can literally fire them at targets.

Hmmm.... now that would be interesting for sure. :evil:

[LaserCobra]

Re: Proof of bullet drop in the game

Best technical game thread I've seen - EVER!

:salute:

[imported_The 13th Raptor]

Re: Proof of bullet drop in the game

Hmmm.... now that would be interesting for sure. :evil:

Oh damn, imagine the carnage, being able to toss remotely detoned high explosive sticky bombs across the entire map..

[imported_Wasp]

Re: Proof of bullet drop in the game

Wasp and slutmonkey,

I seriously fail to see how this doesn't prove it. If decals are off, they aren't THAT far off. I have never once fired my gun at the horizon only to have the decal appear at my feet.


And bullets drop, naturally. I think that the natural movement of bullets would be just as important as for other objects in this game.

Decals are inaccurate and off by several meters in 400 meters. Anyway it is possible to have a drop ONLY on the snipers out of the regular weapons but it's pointless because of the small range in BF2. With an M16 you have about 140 meter view. How much does an M16 drop irl at that range? Insignificant! Why give the server the added workload? The M24 has a range of about 320 meters in BF2. Irl the bullet would probably drop 4 cm. Why give the server the extra workload when the bullets already have deviation?

When you make a game you always have to consider what is nescessary in order to keep the workload on the CPU down. The CPU has to handle an insane amount of work in a normal game. In modern games they have added more advanced physics to vehicles, players and certain other objects which only adds to the already burdened CPU. I have discussed these issues at length with people who makes 3D engines and physics engines which are an integral part of all modern 3D engines. You also have to consider the speed at which bullets move. In order to give a fairly realistic picture of firing a weapon you have to keep the number of calculations down. Just firing a weapon in 3D takes a lot of calculations without putting gravity into the equation. In BF2 bullets can even penetrate some surfaces.

Like I said earlier I find it unlikely that there is any drop on regular weapons and if you want to convince me otherwise then do a proper test. Turn off deviation, establish that the bullet leaves the sniper crosshair dead center and fire at a player so that you can get proper confirmation. If the bullet leaves the weapon even a little bit off from the center of the sniper crosshair you will notice it at 300 meters.

[evilzucchini]

Re: Proof of bullet drop in the game

Wasp,

You make good points, but the facts do not support them.

Decals are inaccurate and off by several meters in 400 meters.

At 400 meters, 3 meters off from where you are aiming is only .43 degrees, which I believe to be WELL within the "deviation" of any non-sniper rifle weapon and maybe even including the sniper rifle. So are you sure the "decal misplacement" you are seeing isn't just misplacement at all and just the result of the deviation from the weapon? Seriously, what makes you think the decals are off?

Why give the server the added workload?

Not sure. However, considering some of the crazy calculations that the server has to go to, an Ax^2 + Bx, even if it is for each bullet, is rather insignificant. IN fact, it might cost them more to have to define the object as NOT effected by the physics engine than to just define it as a normal object. The fact of the matter is that you do not know AND the evidence in this post is very convincing that there is bullet drop.

Really, I have yet to see someone who claims that there is no bullet drop attempt to prove it. All of the anti-bullet drop arguements are criticisms of the pro-BD proofs or, as you have done, simply unsupported claims that it hurts server performance to do so.

[imported_Wasp]

Re: Proof of bullet drop in the game

Wasp,

You make good points, but the facts do not support them.


At 400 meters, 3 meters off from where you are aiming is only .43 degrees, which I believe to be WELL within the "deviation" of any non-sniper rifle weapon and maybe even including the sniper rifle. So are you sure the "decal misplacement" you are seeing isn't just misplacement at all and just the result of the deviation from the weapon? Seriously, what makes you think the decals are off?

I can hit people in the upper body at 320 meters by aiming at their upper body with the M24.. So are you claiming that for the remaining 80 meters the bullet drops 3 meters? And 3 meter is a mild estimate looking at some of the videos on this forum. I have spent time testing this myself with my clanmates. It's not just something I'm saying for the sake of arguing.

Not sure. However, considering some of the crazy calculations that the server has to go to, an Ax^2 + Bx, even if it is for each bullet, is rather insignificant. IN fact, it might cost them more to have to define the object as NOT effected by the physics engine than to just define it as a normal object. The fact of the matter is that you do not know AND the evidence in this post is very convincing that there is bullet drop.
0
Really, I have yet to see someone who claims that there is no bullet drop attempt to prove it. All of the anti-bullet drop arguements are criticisms of the pro-BD proofs or, as you have done, simply unsupported claims that it hurts server performance to do so.

Not giving these physical attributes to each bullet SAVES calculations not the other way around. If you still can't accept that I suggest you present your post in a 3D programmer forum or talk to one.

As for disproving the bullet drop theory I spent an evening firing sniper rounds at my medic clanmate at the maximum range for each sniper. I managed a very high accuracy on the M24 firing series of 5 shots. I would estimate 4 to 5 out of 5. Once I started firing above my friends head I would start to miss badly. This is something that you or anyone else here can try for themselves. Start your own server and invite a friend to join. Or find an empty unranked server.

[evilzucchini]

Re: Proof of bullet drop in the game

I can hit people in the upper body at 320 meters by aiming at their upper body with the M24.. So are you claiming that for the remaining 80 meters the bullet drops 3 meters? And 3 meter is a mild estimate looking at some of the videos on this forum. I have spent time testing this myself with my clanmates. It's not just something I'm saying for the sake of arguing.

First, this is exactly what I am talking about. You offer no proof, just things that you claim happen. I have never had a problem with decals. They all seem to go where i aim, plus deviation. So, if you are going to argue that the decals are off, you are going to have to prove it.

Not giving these physical attributes to each bullet SAVES calculations not the other way around. If you still can't accept that I suggest you present your post in a 3D programmer forum or talk to one.

This is funny because I have both worked on a physics engine both for 2d and 3d. It was basic and not very good, but I have done this type of work. My best friend is a programmer. He has built his own physics engine AND he agrees with me that bullets are affected by gravity. But you are missing my point. I am not saying that it IS a hit, what I am saying is that we don't know.

I could definitely see how an engine that treats every object equally, as in they are all affected by gravity, is less consuming than one that has to determine whether or not each object that it is dealing with has to be affected by gravity or not. Compares are some of the longest operations and can reduce performance further because they screw up pipelined operations. Multiplication is easy because we have dedicated ALU units in processors and there is no need to flush the pipeline afterwords.

But agian, I don't know what the performance is actually like because I have done no study to show it one way or another, but I could definitely see why it would make a minimal difference or would actually be easier, both from a programming viewpoint AND a server load viewpoint.

As for disproving the bullet drop theory I spent an evening firing sniper rounds at my medic clanmate at the maximum range for each sniper. I managed a very high accuracy on the M24 firing series of 5 shots. I would estimate 4 to 5 out of 5. Once I started firing above my friends head I would start to miss badly. This is something that you or anyone else here can try for themselves. Start your own server and invite a friend to join. Or find an empty unranked server.

I actually have done just this and got the exact opposite results. This is why I want to SEE your proof because it is really unconvincing when you just say that it happened.

[SK]

Re: Proof of bullet drop in the game

Decals are inaccurate and off by several meters in 400 meters.

Note that I didn´t take pictures of the decals, but of the round itself, visible as I added tracers to it.

Anyway it is possible to have a drop ONLY on the snipers out of the regular weapons but it's pointless because of the small range in BF2.

All weapons´ projectiles are equally affected by gravity, you can easily check it yourself (With the exception of the AT rockets).

Why give the server the added workload? The M24 has a range of about 320 meters in BF2. Irl the bullet would probably drop 4 cm. Why give the server the extra workload when the bullets already have deviation?

The "Why?" is in fact pretty irrelevant here. It´s fact that bullets are affected by gravity, period. At distances of 350m+ bullet drop is estimated about as large as an enemies´ head.

When you make a game you always have to consider what is nescessary in order to keep the workload on the CPU down. The CPU has to handle an insane amount of work in a normal game. In modern games they have added more advanced physics to vehicles, players and certain other objects which only adds to the already burdened CPU. I have discussed these issues at length with people who makes 3D engines and physics engines which are an integral part of all modern 3D engines. You also have to consider the speed at which bullets move. In order to give a fairly realistic picture of firing a weapon you have to keep the number of calculations down.

I haven´t coded for a game yet, but I have done quite a bit of 3D coding in the last few years. Let me assure you: Adding a single code line in a loop for affecting the Z-axis of all projectiles with a simple multiplication is pretty much irrelevant compared to the transformations the CPU/GPU has to do to transfom the coordinates of even a single 3D model with a reasonable number of polygons.

Like I said earlier I find it unlikely that there is any drop on regular weapons and if you want to convince me otherwise then do a proper test. Turn off deviation, establish that the bullet leaves the sniper crosshair dead center and fire at a player so that you can get proper confirmation. If the bullet leaves the weapon even a little bit off from the center of the sniper crosshair you will notice it at 300 meters.

Well I think I did quite reasonable testing, I also fired at bots and it hit exactly where the tracer went. You can do some tests and try to prove me wrong, but I won´t take some more hours of testing just because you think DICE should´ve coded the game in a different way.

I can hit people in the upper body at 320 meters by aiming at their upper body with the M24.. So are you claiming that for the remaining 80 meters the bullet drops 3 meters? And 3 meter is a mild estimate looking at some of the videos on this forum. I have spent time testing this myself with my clanmates. It's not just something I'm saying for the sake of arguing.

Remember that there is the cone of fire by deviation. This means that you might get hits that look like there is no bullet drop, even if you just randomly hit your target. I made this graphic to explain what I mean:

Red: Your crosshairs.
Black Circle: The cone in which the bullets will land, bullet drop considered.
Blue: The target you intend to hit.
Even if you don´t consider bullet drop you have a chance to hit your target, but the chance is much greater if you take bullet drop into account, as the area of the target inside your cone of fire is much larger.

As we don´t have hard data on the cone, this is only speculation though.

Not giving these physical attributes to each bullet SAVES calculations not the other way around. If you still can't accept that I suggest you present your post in a 3D programmer forum or talk to one.

On the other hand a unified physics engine, where all objects are treated the same, makes you more flexible. It´s much easier to alter an object´s properties when it´s instance of a generic object class that is affected by physics like anything else. As mentioned, adding a single line with a multiplication isn´t that demanding.

I would estimate 4 to 5 out of 5. Once I started firing above my friends head I would start to miss badly. This is something that you or anyone else here can try for themselves. Start your own server and invite a friend to join. Or find an empty unranked server.

My experience differs, but I might do this kind of test if I find the time.

[imported_Wasp]

Re: Proof of bullet drop in the game

But agian, I don't know what the performance is actually like because I have done no study to show it one way or another, but I could definitely see why it would make a minimal difference or would actually be easier, both from a programming viewpoint AND a server load viewpoint.


I actually have done just this and got the exact opposite results. This is why I want to SEE your proof because it is really unconvincing when you just say that it happened.

I have a friend who has been programming for more than 15 years. He was very good with assembler but went over to C++. He made his own 3D and physics engine and did a lot of research. He also spent a lot of time optimizing the code. One of the sources he used was John Carmack. I don't code anymore I have just done some work in Maya. So I often discuss issues with 3D engines with him. I have also discussed the "bullet drop theories" on these forums and my own findings with him.

If you need proof then I'll simply make a fraps video and post here.

[Dogness]

Re: Proof of bullet drop in the game

I guess I don't understand this debate. The "bullet drop" is in there. It has been in there since BF1942. The bullets you fire from your gun are no different from the shells fired by your tank.

I remember playing Charlie Sector in FH. If you had the sniper rifle on the beach and you wanted to hit the MG42 gunners in the bunkers (40 ft up and at the edge of the fog), you had to use the lowest tick mark in the sight to hit them. It took me forever to figure that one out.

In BF2, whenever I am shooting at someone from a distance (with any gun), I always aim a little above them. If I miss low, I keep raising my aim until I hit them. This isn't rocket science. Well, maybe it is. :)

In any case, what is the purpose of this debate? If the no-bullet-drop folks think there is no bullet drop, let them think that. Why are people trying to prove to them that it does exist?

Oh, wait a minute, is this what people are talking about when they complain that the game is broken because they can't hit what they are aiming at? I never understood that one either.

[Ubermensch]

Re: Proof of bullet drop in the game

wasnt this topic like already proven some time ago?... how many debates are there going to be about it?

[Fea]

Re: Proof of bullet drop in the game

Some guys in here "need" to start thinking about projectile paths. Will you actually land a hit a bit too high if you aim straight between the eyes at say 50-60 m? The bullet is still rising at that range. If I'm not confusing the numbers (IRL that is).
I would look in to it myself if I had the time and if it wasn't 2½ years since I had anything to do with physics in school.

That there is bullet drop is proven until someone produces results of the same quality saying otherwise.