Bullet velocity - inaccurate ?

Il be brief, I had made a similar past ages ago and didn't get many replies. Firing at a target 400 / 450 m away , it appears, from dust appearing that it takes around a second for the round to travel this far. Although not schooled in such things, with a muzzle velocity of 900m/s from a m16 , even accounting for drag , surely this is a little slow? Arma2 vanilla 1.05 if it makes a difference.

Try to make more accurate measurements... I doubt it actually takes an entire 1.00 seconds at 450m. Add that to the fact the bullet doesn't keep flying at 900m/s (for example, an M16A4 will drop down to ~800m/s at ~150m), and you get that the bullet speeds in Arma 2, even if not perfect, are at least not bad.

At 900m/s, one would expect that it would take a minimum of 1/2 second for a bullet to travel 450m. I'd imagine that since there was no scientific measurement taken, that this is probably about the case you're seeing in ArmA 2. In order to isolate an actual problem, one would have to script some kind of distance and time measurement.

Well nothing beyond the old Mississippi 1 mississippi 2. Any sort of 'air' or 'drag' in the arma 2 world ?

mission.sqm:

hmm ... some time ago i checked bullets speed ingame. From what i remember calculated speed was around 3270 Kmh, just around 900m/sec.

Well, that information should be in the initspeed entry of the cfg, unless that is in doubt.

Ok so all this is based on that you fire at a target which stops the bullet prematurely (before it would hit the ground due to gravity).

http://en.wikipedia.org/wiki/Drag_(physics)#Drag_at_high_velocity

If you do some calculations using the drag formula above you get this expression for bullet flight time:

t = e^(K * (s - C)) - 1/(K*v0)

where K = Cd * rho * A * 1/2 * 1/m and C = -1/K*ln(1/(K*v0)).

Parameters:

http://en.wikipedia.org/wiki/External_ballistics#Doppler_radar-measurements

http://en.wikipedia.org/wiki/Density_of_air

Using

m = 0.02 kg

rho = 1.2 kg/m^3

Cd = 0.25

v0 = 900 m/s

r = 0.009/2 m (edit: this was too high before)

A = pi * r^2 m^2

s = 450 m

you get about 0.56 seconds with drag. Without drag it would just be about 0.5 secs (a little bit more because you have to fire above the target, v0 in the direction of the target is less than 900). But depending on the parameters of the bullet you can have a flight time of one second or more at that distance. (depends very much on the calibre)

v(t) = 1/K * (1/(t+1/(v0*K))) with t=0.56 gives about 726 m/s (impact velocity). Velocity Range 900-726 m/s. In reality Cd depends on velocity. To make it simple I used a constant value of Cd.

And remember that the speed of bullets in Arma2 is decided not by the gun, but by the magazine. This means, that despite the M16 having a higher initial velocity in real life, in-game it fires at the same speed as every other weapon that uses the same magazine. M4, XM8, even the SAW (which can take STANAG mags).

ACE2 fixes this, BTW. It's changed so that there's a value in the weapon's config which modifies the magazine velocity.

He is so right, rifle bullets are slow and they were in ArmA 1 too. I think the initial design was afected by lag and FPSs.

It is so obvious, a bullet medium speed is around 800m/s while in game it takes a second to reach like 300m, so less than half of real speed.

I noticed that too and I wanted to increase the speed but I supposed that affect damage, penetration and others so I gave up.

Well, does the issue seem any better in ACE?, tbh i'm not wanting accurate bullet speed if 7.62 is even slower!. Place 1 unit in the editor , fire at a roof of a building at 400/ 450 m and watch carefully for the impact. Easiest with the XM sharpshooter . Shooting horizontally, it's difficult to judge just how far the bullet has gone due to drop.

I'm playing this game mainly because of its simulation capabilities, and ballistics are a very important part of that. If it really takes a second there might be something wrong with the ballistics simulation. There's too much drag or something. We could create a ticket about it and make them look into it.

But maybe it takes a few hundred millisecs to calculate impact and display that dust animation? 300-500 ms is enough for getting 1 sec in that case.

Good point ^^ and in that case it would be fine. However , reactions of AI enemy to wounds is pretty much the same , the 'flash' of impact when firing certain weapons (m240 comes to mind), upon impact with certain structures reveals the same slow bullet speed.

Similarly to spokesperson above, when a issue like this is present it does play on the mind , not too mention screwing up realistic lead angles.

Its not like I've taken any measurements but I kind of feel in my bones that the bullets do travel too slowly.

With a stopwatch and taking about .15 off for reaction - on average 1 second precisely for a target (450m) away. Done in the editor with no other units at about 50 odd fps. Its about 1/3 too slow. Any chance of a ACE gamer trying a few different weapons ? - great game all the same :)

50 cal is faster , average .80 for same distance.

If anyone is interrested in testing the speed of a bullet instead of guessing I would like to point you to post number 5 in this thread which contains a copied mission.sqm text.

One way to use it is to start the editor, place a unit, save mission, go to mission folder, open mission.sqm, replace text, save changes, load mission, play mission.

For everyone who likes guessing, dont let me interrupt ;)

the tip is never trust your eyes and hands, unless there is a scientific way to measure bullet speed, we are going nowhere.

guys, test with tracers, is more visible.

If someone could test inqwipers script that would be great - assuming it measures flight time and not just speed. - no net at mo .

If someone could test inqwipers script that would be great - assuming it measures flight time and not just speed. - no net at mo .

Yeah, it gives the following results:

Bullet fired at: # seconds from mission start (t0)

Bullet hit at: # seconds from missions start (t1)

Distance: # meters from player to target

Average bullet speed: # distance / (t1-t0)

You could try slowing down the game time and seeing what should happen given vastly increased CPU time (which is what time deceleration represents): make a radio trigger with the code:

setAccTime 0.01;

which will slow time down to 1/100 when you trip the trigger. Then you can repeat your stopwatch tests with an increased margin for error/reaction.

You could try slowing down the game time and seeing what should happen given vastly increased CPU time (which is what time deceleration represents): make a radio trigger with the code:

setAccTime 0.01;

which will slow time down to 1/100 when you trip the trigger. Then you can repeat your stopwatch tests with an increased margin for error/reaction.

Thats a good idea. I tried it in my mission but for some reason it returns the same time for bullet fired and bullet hit when running at 0.01. Ofcourse this is not a problem if you want to use the stopwatch method (which I dont recommend because you dont know how long it takes for the smoke puff to appear after the hit). What did work was setAcctime 0.05 which is probably slow enough and seems to return the same exact value every time unlike running the mission at normal time.

I vaguely remember that someone found that there was a significant delay in the dust/smoke effect reaction in Armed Assault.

This is weird but i used to lead distant targets incredibly well, even behind a scope with the a.i. warping in the distance (online), then after BIS changed the FOV in some late patch for the first Arma i started having a hard time leading moving targets and now in Arma 2 i still have the same problem.

Heatseeker - I was also good at moving targets in ofp - never played arma1 and am experiencing similar issues to yourself, perhaps due to slow rounds. Great game but this is an issue that can't be ignored.