I just did some testing with my spider and cicada to get a 'feel' for how long it takes to chew something with 4 machine guns versus 4 small lasers. With the spider I put in 4 machine guns and moved up to point blank on each of a commando a cicada and an atlas and stopwatch timed how long it took for the target to go down with damage *only* to their front centre torso.
The interesting part is I used a stopwatch but also noted the ammo counts before and after. All machine guns were in fire group 1, with the trigger held down the whole time.
Machine Guns x4:
Commando> 9 seconds, 340 rounds (ammo use suggests it was shorter, than 9 seconds)
Cicada> 15 seconds, 476 rounds (ammo use suggests it was only 11.9 seconds)
Atlas> 51 seconds, 1588 rounds (ammo use suggests it was only 39.7 seconds).
This suggests the machine gun is only firing at 7.78 rounds per second, not 10.
Almost as though someone modded the fire rate so the buff from 0.08 to 0.10 would be completely offset>
edit> Just tested 4 group fired machine guns on the stopwatch seeing how long it took them to drink 4000 rounds of ammo. 119 seconds on the nose for a fire rate of 8.4.
So yeah, machine guns are firing 20-25% slower than advertised.
I would but I don't have any screen cap software installed - If I asked real nicely would one of you take a machine gun out for a test and post back? Maybe if we get independent confirmation that will also motivate a response.
This suggests the machine gun is only firing at 7.78 rounds per second, not 10.
Almost as though someone modded the fire rate so the buff from 0.08 to 0.10 would be completely offset>
edit> Just tested 4 group fired machine guns on the stopwatch seeing how long it took them to drink 4000 rounds of ammo. 119 seconds on the nose for a fire rate of 8.4.
So yeah, machine guns are firing 20-25% slower than advertised.
Cubivorre sent me here. I've noticed machine guns have been firing at odd rates for a while now. Mine fires as fast as 13 bullets per second, but it bounces from 8 to 13 and it takes going through second after second of video to find out the rates as they shift about.
It's been something that seems to shift from PC to PC, much like Combat Arms' minigun firing rate. I could fire it for about 15 seconds and be out of ammo. Some can fire it non-stop for over 45 to get through the same amount of ammo.
This was before any buffs if it helps for a comparison.
3988 at the start of the video.
1888 at the end of the first firing.
2,100 bullets fired. (Each doing 0.04 damage per bullet) for a total of 84 damage.
Starts at 52 seconds in. Firing stops at 1:38. 60 seconds + 38 is 98 - 52 = 46 seconds of firing.
2100 bullets divided by 6 = 350 bullets fired by a single machine gun.
350 to 46 seconds. Ooo. This suggests an overall firing rate of a single machine gun to be 7.6 rounded down per second before ANY buffs. But if you go second by second, there's times when it comes out to 12 point something bullets in a second per mg.
2100 bullets for 84 damage at 0.04 damage per bullet within 46 seconds at an average firing rate of 7.6 shots per second per machine gun on client-side before any buffs.
(for fun the same 2100 bullets now would deal 210 damage in 46 seconds at the same firing rate.)
Thank you Koniving for that video, and cubivore for bringing him here.
I'm tempted to try and figure out why it is that the gun can show such variance from second to second. I wonder what rate the server is updated at and if the game is trying to send information faster than the 'nyquist' rate of this update process.
Anyone know if the client updates the server faster than 20x a second?
Anyone know if the client updates the server faster than 20x a second?
I suggest you read Karl Berg's posts with respect to the network (there should be a command chair post explaining the frequency). He specifies this info there.
That post describing the 20 Hz state update makes me more curious, and wonder if gun fire states are delta compressed like the other states. If they are, then the 20 Hz update rate shouldn't cause any trouble for the guns fire rate, but if it is not delta compressed then a 20 Hz state update is not sufficient for a 10 Hz gun update.
The 'nyquist zone' of a 20 Hz state update process is less than but not equal to 10 Hz, and that is assuming you have no aperature jitter or sample loss.
I can confirm this issue both on testing grounds and live server.
If it's a network issue, why does it occur on testing grounds, where there is no client/server communication?
I don't know.... I wasn't aware the testing grounds were serverless
Even if they are, if it's a locally simulated server with a 20Hz update rate depending on how dodgily the 20 Hz comms is implemented sampling the 10 Hz firing command you would still see aliasing/missed shots.
0
