Sniper bullets have unrealistically low aerodynamic drag

[-Gews- 1]

I made this thread because would really like to see these things corrected while the game is still in Alpha so the finished product can be as good as possible.

(Long post, here's a summary: I would like to see air friction values changed for sniper ammunition, air drag on these bullets is far too low at the moment when compared to real life. I suggest values at the bottom of the post, the rest is reasons why.)

First off, I would like to thank Bohemia for taking the time to make these small changes like these! :D

I noticed that the GM6 Lynx now fires APDS ammunition at 1,060 m/s. Previously the config had it fire standard ammunition at an impossible velocity, and I remember wanting to have it changed. Having such a high velocity now makes sense with an APDS projectile. This change is good!

However, there are still one or two small issues, mainly concerning the .408 Cheytac ammunition.

Muzzle velocity and damage are good, it has appropriate values for these, however the airFriction coefficient is too low, even for such a streamlined round.

I calculated airFriction values for many different bullets by comparing velocities to real-life ballistic data... this is the type of value streamlined bullets have:

Typical values for low-drag match bullets

.308 Win = 0.00082

.300 Mag = -0.00073

.338 Lapua = -0.00068

.50 BMG M33 = 0.000615

.50 BMG = -0.00042

^The above values are for match-grade bullets (except M33) which were selected based on availability of very precise data (specifically Sierra MatchKing 190 and 175, Lapua B408, M33 and 750-grain Bullex-N)

There is less data available for .408 Cheytac bullets, however, I was making a configuration for a .408 Cheytac and I spent a good deal of time researching.

I came up with an airFriction value of -0.00042 for this round (409 grain bullet). I found this matched ballistic calculator speeds (from 100 to 1,000 meters) very closely, to the point that I was actually surprised at the coincidence.

Anyways, the current coefficient of -0.00013 is far too low for any sort of bullet to achieve, and the same problem faces the GM6 Lynx ammunition (airFriction = -0.0001).

The 12.7x108 in the Lynx is supposed to be APDS ammunition, but even for that type of round the drag coefficient is too low and it loses almost no velocity at any reasonable range.

For the Lynx, I don't think many people would know an "appropriate" value since it fires APDS ammunition and ballistic data is hard to find. However, I think it would definitely be higher than -0.0001!

I found a discussion on the SOCNET forums (for military members) mentioning ballistic coefficients and 12.7mm APDS (.50 BMG SLAP) ammunition:

This ballistic co-efficient is based on an average decay of velocity from a muzzle velocity of 3950 fps at the muzzle to a 1473 fps reading at 2500 meters. The AVERAGE BC for this distance under standard atmospheric conditions is 1.0562

It's considerably higher initially at the muzzle and this number is NOT based on the penetrators shape or form factor. Its purely based on velocity decay over 2500 meters, which in opinion is much more accurate. I develop ballistic data tables (NOT computer generated) using a scientific calculator and varying BCs every 100 meters, over a 2500 meter spectrum for the 50s. Differing ranges for the other calibers that I have.

Putting that number into a ballistic calculator (with muzzle velocity = 1060 m/s) gives a velocity (ballpark figure) of about 500 m/s at a range of 2,000 meters, which is still very impressive (at 2,000 meters normal .50 caliber M33 is going only half that speed).

So to get that 2,000 meter velocity of about 500 m/s in ArmA, one needs to put an airFriction value of about -0.00037.

Therefore I suggest making the following changes:

.408 CheyTac airFriction = about -0.00042

12.7x108 APDS airFriction = about -0.00035 to -0.00037

A long post perhaps, but I would like to see some of these final improvements done to the sniper ammunition in an effort to make ArmA 3 as close to "real life" as possible, and especially with long-range shooting, these sort of ballistics become quite important in that aim. :bounce3:

Edit: I also noticed "normal" 12.7x108mm ball has an airFriction of -0.00016, compared to 12.7x99mm which is -0.000482.... that should be changed as well (if any weapons plan to use "normal" 12.7x108, right now all the vehicles appear to use 12.7x99).

Also there is .50-caliber SLAP which uses airFriction = 0.0001, and that should be changed (see velocity values above which are from 50 SLAP).

Edited June 1, 2013 by Goose

[roberthammer 582]

If you can , create a ticket about this - i will upvote it :)

[-Gews- 1]

If you can , create a ticket about this - i will upvote it :)

Here she is: http://feedback.arma3.com/view.php?id=8985 :D

[kklownboy 43]

Awesome stuff! Which Calculator do you use?

[-Gews- 1]

Awesome stuff! Which Calculator do you use?

QuickTARGET Unlimited for real life calculations, for the ArmA calculations one I made myself, it's fairly accurate.

Example:

Range-----In-game v-----Calculated v

000--------900-------------900.0

300--------678-------------678.6

800--------423-------------423.9

1000-------350-------------351.1

1400-------242-------------241.0

1753-------174-------------172.9

[Insanatrix 0]

I kind of thought they were a little fast, The .408 especially.

[bigpickle 0]

upvoted too

[gpha5e 1]

voted up, a very good post!

[dmarkwick 261]

QuickTARGET Unlimited for real life calculations, for the ArmA calculations one I made myself, it's fairly accurate.

Example:

Range-----In-game v-----Calculated v

000--------900-------------900.0

300--------678-------------678.6

800--------423-------------423.9

1000-------350-------------351.1

1400-------242-------------241.0

1753-------174-------------172.9

I'm not sure I understand or see the problem: if that table represents ingame and RL velocities at those ranges, then they seem pretty close to me?

[-Gews- 1]

I'm not sure I understand or see the problem: if that table represents ingame and RL velocities at those ranges, then they seem pretty close to me?

The table is just an example calculation with 7.62x51mm to show the calculator I made is fairly accurate (greater than 99%). It shows in-game and calculated velocities (the in-game velocities were rounded to the nearest meter-per-second).

[Covert_Death 11]

love the research you went into so far but i was not aware the sniper pack was released yet... how are you testing sniper physics in game?

[-Gews- 1]

love the research you went into so far but i was not aware the sniper pack was released yet... how are you testing sniper physics in game?

Don't need to test it in-game, can simply look at the config and run calculations. Did plenty of in-game testing in ArmA 2. As for the snipers, opt into the dev-build.

[pettka 694]

This step away from real data has been done on purpose for gameplay sake - AI snipers weren't able to shoot at distances longer than 800 meters reliably, they didn't take zeroing that far into account. This would make player extremely overpowered at distances around a kilometre, where player is safe and still deadly with the rifle thanks to better zeroing skills. And nobody wants AI just sitting like ducks and waiting to be shot :icon_twisted:

[NeMeSiS 11]

This step away from real data has been done on purpose for gameplay sake - AI snipers weren't able to shoot at distances longer than 800 meters reliably, they didn't take zeroing that far into account. This would make player extremely overpowered at distances around a kilometre, where player is safe and still deadly with the rifle thanks to better zeroing skills. And nobody wants AI just sitting like ducks and waiting to be shot :icon_twisted:

Couldnt you just handle the bullets differently for AI and players? :)

[k0rd 3]

That's all we need, a bunch of AI littering the forum and tracker with "HUMANS ARE OVERPOWERED!!!!" and "PLEAZ NERF HUMANZZZ" posts :D

on topic, thanks for sharing your research OP, I hope that a middle ground can be found between realism and the need to balance player vs AI

[kuIoodporny 45]

Goose, please check your inbox. ;)

[-Gews- 1]

Hmm... I would rather have AI looked at than realism sacrificed.

---------- Post added at 19:30 ---------- Previous post was at 19:29 ----------

Goose, please check your inbox. ;)

I believe I replied, however I may have not sent it? :D

[mr.g-c 6]

This step away from real data has been done on purpose for gameplay sake - AI snipers weren't able to shoot at distances longer than 800 meters reliably, they didn't take zeroing that far into account. This would make player extremely overpowered at distances around a kilometre, where player is safe and still deadly with the rifle thanks to better zeroing skills. And nobody wants AI just sitting like ducks and waiting to be shot :icon_twisted:

Errr ok... Well then change the AIs behaviors and routines ;-) I really hate to read that Realism values are sacrificed for the AI. This "juggling" with values to have a "fair" and so called "balanced" game, gives me the slight feeling (which i hate aswell) that there wont be changed much in the field of AI for A3 final. :-(

Also keep in mind, in RL its not at all easy to fire and hit a target at such ranges. Its not like in the current ArmA, where you going prone, do the zeroing, fire and perfectly hit in under 15seconds. Thats arcadish. So even if the AI takes a long time for a shot at that distance, let it happen - its realistic.

Just add some more realism for players when firing with a hand-held gun at those ranges, so for instance your player needs more time until the waepon sway is down to an appropriate level, having to hold your breath for some longer seconds than currently, implementing wind drag for bullets, etc etc.

It could be soooo good if done right:-)

Just my 2 cents

[twisted 128]

Errr ok... Well then change the AIs behaviors and routines ;-) I really hate to read that Realism values are sacrificed for the AI. This "juggling" with values to have a "fair" and so called "balanced" game, gives me the slight feeling (which i hate aswell) that there wont be changed much in the field of AI for A3 final. :-(

Also keep in mind, in RL its not at all easy to fire and hit a target at such ranges. Its not like in the current ArmA, where you going prone, do the zeroing, fire and perfectly hit in under 15seconds. Thats arcadish. So even if the AI takes a long time for a shot at that distance, let it happen - its realistic.

Just add some more realism for players when firing with a hand-held gun at those ranges, so for instance your player needs more time until the waepon sway is down to an appropriate level, having to hold your breath for some longer seconds than currently, implementing wind drag for bullets, etc etc.

It could be soooo good if done right:-)

Just my 2 cents

good thinking. and itd help the ai seem more human in that it needs time to take a shot. very important in making ai seem a bit more 'human'. and also in keeping true to the functionality of the weapon. please.

Edited June 5, 2013 by twisted
spelling

[blu3sman 11]

All this recent "balancing" talk starts to bother me.

Like here feedback.arma3.com/view.php?id=8656

Why limit fire rate of semi-auto rifle to be the same as bolt action one? Why then bother to have different weapon models if they have identical specs? More logical would be to make realistic characteristics, which ARE balanced: GM-6 has low recoil, high fire rate, but worse accuracy and effective range.

On the issue with ballistics. Taking aside sci-fi projectiles, while it makes AI to hit targets, this also makes it SUPER EASY for humans to hit targets. Especially moving targets are easy, because bullet flight time is just ridiculous. All you did with this balancing is just shift both AI and human effectiveness to higher %.

Edited June 6, 2013 by Blu3sman

[byku 13]

I have to say that it also irritates me a bit. Artificial balancing is horrible. I'm personally disappointed, especially the comment on GM6 Lynx was a huge let down :/. I thought by recreating something in Arma devs would make it as close to reality as possible... and now we've got GM6, and I suppose only the visual aspect is real...

Edited June 5, 2013 by Byku

[pathetic_berserker 4]

In summary, if the community sees it as a bug its not balanced

[.kju 3245]

"Balancing" is the wrong approach. Also this does not lead to better gameplay.

AI tweaking/fixing and more realism, like bullets to be affected by wind, would be the right approach.

[byku 13]

FYI the Comanche also "has been balanced", the original has 500 rounds for main gun, Arma's 1000. http://www.army-technology.com/projects/comanche/

[Masharra 10]

I would say something about sounding like battlefield adding "realistic" sniper flare on the scopes but that would be ....