|
Post by hatton on Nov 15, 2011 12:58:42 GMT -5
Anyone know where I can find some official BF3 gun stats? I heard they are going to be posted on this site soon but I can't find anything official from Den.
Also, if someone can link me to the BFBC2 stats that Den produced so I know what to expect for BF3 that would be appreciated.
I'm looking for official stats on the damage, reload time, fire rate, mag size and speed of all the weapons.
Thanks for your time guys,
You are my bro, bro
Hatton
|
|
battleaxerx
True Bro
"You can't take the sky from me."
Posts: 773
|
Post by battleaxerx on Nov 15, 2011 13:15:18 GMT -5
|
|
|
Post by hatton on Nov 16, 2011 7:13:21 GMT -5
Ok thank you. Do you know if the BF3 stats by Den are out yet? Is he going to be doing them?
|
|
|
Post by didjeridu on Nov 16, 2011 16:25:53 GMT -5
They aren't out yet. And no one knows when they're coming out, or if Den's doing them at all. We're all waiting eagerly, and dying a little inside every day.
Anyways, they'll come out eventually. Even if Den suddenly decided that he hates BF, there are other people here who are capable. It's just a matter of time.
|
|
battleaxerx
True Bro
"You can't take the sky from me."
Posts: 773
|
Post by battleaxerx on Nov 16, 2011 16:35:14 GMT -5
We're all happy to take some weight off of the Brotius Maximus' shoulders if it comes to that. lol =D
Not me exactly, I just know how to look at stats... not extract them... I do oddball testing from time to time though. (The jammer in blops has a max radius of 25m, which is the length of a square in nuketown - and the squares in firingrange are slightly larger)
|
|
Den
He's That Guy
Posts: 4,294,967,295
|
Post by Den on Nov 18, 2011 7:00:11 GMT -5
The file encryption is way out of my league, so I'm waiting for those more capable to piece together the 50000 piece jigsaw puzzle that exists in the .cas files. Until then, nothing from me because... Even if Den suddenly decided that he hates BF... I do actually hate Battlefield. For BF3, it is partly due to technical reasons not directly related to BF3. Origin causes my brand new PC to lock up, frozen screen, speakers stuck playing a loud tone (others on the Battlelog forums have the same problem, some refer to it as the BRRRRR). I almost had a heart attack when long after after I had finished playing some BF3 (but left Origin on), my PC froze while I was just browsing the web. I haven't touched BF3 in over a week; I gave Punkbuster a full inning of chances to not put my new box in a coma. I've tried lots of things on my side short of getting a different sound card or turning off my audio to make it not happen, only to suffer more lockups.
|
|
novem
True Bro
Posts: 193
|
Post by novem on Nov 18, 2011 8:12:36 GMT -5
Origin causes my brand new PC to lock up, frozen screen, speakers stuck playing a loud tone (others on the Battlelog forums have the same problem, some refer to it as the BRRRRR). Do you have realtek sound controller on your motherboard and no separate soundcard? If yes this is causing bsods (freeze with bzzzzzz sound and you can do nothing but restart)- realtek really doesnt like Punkbuster. Its issue known from Bad company 2, which hasnt been resolved yet (shame on you Dice/EA). Solutions: 1) install newest realtek drivers and newest version of pb, update it 2) if previous doesnt work (in my case it didnt) you can turn off realtek drivers (sometimes you have to disable it in bios)- but the downside is the game without sound which sucks 3) you can buy additional sound card- this will resolve realtek problems 4) if you dont have much money you can buy very cheap (like 3 dollars or so) usb sound card- it also resolves problem 5) you can use usb headset- it has its own sound interface which will disable realtek (it works for me)6) ultimately you can play on servers without pb Points 4 and 5 are imo the easiest and permament solutions.
|
|
|
Post by didjeridu on Nov 19, 2011 12:14:59 GMT -5
That sucks. I have no problems with Origin personally, but I still think it's a waste of resources (no matter how small) and entirely pointless. The only thing keeping me from using the no Origin crack is the supposed chance of being banned/whatever. Too much of a risk for a game I enjoy so much. Hopefully they fix all these issues soon. I've been lucky, and aside from the servers disconnecting and party joins being wonky, I've had no problems with BF3. But one thing I see everywhere is that people have all kinds of issues with Origin/BF3, and the game is unplayable for a good chunk of the population.
|
|
|
Post by frankelstner on Nov 19, 2011 19:57:26 GMT -5
|
|
|
Post by rudybojangles on Nov 20, 2011 17:40:24 GMT -5
I am greatly interested, though not technically literate. What's the workload after unencrypted? Seems like the guns had a lot of variables associated with them.
|
|
|
Post by bel on Nov 20, 2011 23:23:36 GMT -5
I'm curious too. I'm fairly tech literate - I use python myself for work, for example - but I've never looked more than casually into the bf3 encryption. What exactly do your scripts produce, and what's required from there to get the gun stats? Have you tried it yourself even just for one or two weapons?
I appreciate that I could just run them myself and see, but it'll be a couple days before I have the space, and so the lazy-man's option is to ask.
EDIT: nevermind, you've explained this in depth in one of the threads you linked. Interesting, but it's also a lot of work and some things still aren't quite there. This will be fun....
Also it's good to finally hear where Den stands on this. The silence had been deafening. Really unfortunate that the game's so unstable for you. I've only had that error once or twice in several weeks.
|
|
|
Post by frankelstner on Nov 22, 2011 8:31:50 GMT -5
One quick note. Do not trust arrays as the script doesn't handle them correctly. Their data is stored at a different place. If you see any section like the one below the values inside are actually from other assignments later on and must be ignored. The 420.0 below are indeed the very same bytes which are read for the projectile velocity. The good thing about this is that I finally know the use of every single section in the file.
The rest of the numbers are not affected by this.
#incorrect values array member FireEffectData Offset Vec3 x 0.0 y 0.0 z 1.0 Rotation Vec3 x 0.0 y 0.0 z 420.0
|
|
|
Post by bel on Nov 22, 2011 18:10:12 GMT -5
Oh, excellent, that makes sense. I'd seen a lot of that kind of data and wondered if it was meant to be nonsensical. The actual projectile velocity is also in an array like that, though, and seems to make perfect sense. Did we get lucky on that one, or are those values actually incorrect?
|
|
|
Post by bel on Nov 22, 2011 18:53:32 GMT -5
Actually while I'm here, frank... how far are we away from being able to actually mod the values in the game? In singleplayer or something, obviously.
I'm starting to feel like the only way to actually see how recoil behaves is to set a bunch of values to zero or 9999 and see what it does. Is that even remotely possible?
Also, have you tested the extraction post-patch? It looks to me like the patch data went into a different folder.
|
|
|
Post by frankelstner on Nov 23, 2011 8:43:13 GMT -5
I am greatly interested, though not technically literate. What's the workload after unencrypted? Seems like the guns had a lot of variables associated with them. Hmm yes, the encryption. Initially I thought the encryption from the beta would make a reappearance, but with the introduction of the nyan cat archives decryption is not necessary for most purposes as almost everything (with a few exceptions) is stored uncompressed and unencrypted in these files. However, it made grabbing decent paths much harder because these paths used to be specified in the encrypted files. And let's not start with the binary XML files now having nine different sections instead of three as there used to be in BC2 (two of them actually being redundant). Oh, excellent, that makes sense. I'd seen a lot of that kind of data and wondered if it was meant to be nonsensical. The actual projectile velocity is also in an array like that, though, and seems to make perfect sense. Did we get lucky on that one, or are those values actually incorrect? My bad at explaining. The incorrect values appear only in "array member" subsections. There are basically two different assignments with subsections, one being identified with 2900 and the other with 4100. I hide these identifiers and merely show the words I get. Coincidentally however I have not spotted an "array" without 4100 or vice versa (I haven't seen an "array" without "member" either). Thus the word "array" is a pretty solid indicator that the values in the following subsection are not to be trusted. The other sections are unaffected. Actually while I'm here, frank... how far are we away from being able to actually mod the values in the game? In singleplayer or something, obviously. I'm starting to feel like the only way to actually see how recoil behaves is to set a bunch of values to zero or 9999 and see what it does. Is that even remotely possible? Also, have you tested the extraction post-patch? It looks to me like the patch data went into a different folder. I had modded the game two weeks ago by twiddling with bytes: imgur.com/3RxgrIn fact I had 10000 bullets, it's merely the HUD showing 1000 instead. BC2 supported 5 digits, weird. I suppose I will make a tool to pack up these things eventually. But I'm not done going from binary to XML yet. I think I will finish the tool I am working on right now in about a week. If I were to take a guess regarding a date for a tool going the other way: It took me about 70-80 hours of analyzing in the past 25 days to get this far, so I think Christmas or some time around New Year's Eve is a good estimate. I'm glad you actually want to test these things. For now I recommend modding BC2 as the tools are already available: www.bfeditor.org/forums/index.php?showtopic=15517You'll need to get quickbms and its fbrb script from elsewhere though. Sadly BC2 doesn't have horizontal recoil specified yet. On the other hand, you'll surely find something to test considering the similarities. And BC2 loads much faster than BF3. I assume the patch works the way it did in BC2. The game probably first loads the usual files, then loads the patched files overwriting outdated values. It's basically just the load order. Well, actually BC2 was even more extreme because maps contained objects too. I think the game loaded up the common files first, then the individual map files (which might overwrite existing things), then the patched common files and finally the patched map files.
|
|
|
Post by bel on Nov 23, 2011 9:07:41 GMT -5
Thanks, frank. As to modding BC2, I'm not so sure... the values in BC2 accurately described what was going on (or I think so, anyway). The values we have so far for BF3 clearly don't.
I don't know if you pay much attention to the other threads, but my big issue at the moment is that the recoil and spread recovery rates are absurdly high compared to the spread added per shot. So high that the guns shouldn't recoil or increase spread at all.
I'm at the point where I have to believe that either DecreasePerSecond doesn't actually occur per second, or there's other variables modifying it which were not present in BC2. I've found a couple of extra vars for recoil, but nothing so far that gives me theory matching practice. There's also a few other quirks, like some guns which have equal horizontal min-max (-0.2 +0.2) but have a clear sideways drift when fired.
All up, there's a lot of variables that made sense in BC2 and don't make sense anymore. I'm not sure modding BC2 will help me there, but I'll have a play.
|
|
|
Post by frankelstner on Nov 23, 2011 9:51:10 GMT -5
Thanks, frank. As to modding BC2, I'm not so sure... the values in BC2 accurately described what was going on (or I think so, anyway). The values we have so far for BF3 clearly don't. I don't know if you pay much attention to the other threads, but my big issue at the moment is that the recoil and spread recovery rates are absurdly high compared to the spread added per shot. So high that the guns shouldn't recoil or increase spread at all. I'm at the point where I have to believe that either DecreasePerSecond doesn't actually occur per second, or there's other variables modifying it which were not present in BC2. I've found a couple of extra vars for recoil, but nothing so far that gives me theory matching practice. There's also a few other quirks, like some guns which have equal horizontal min-max (-0.2 +0.2) but have a clear sideways drift when fired. All up, there's a lot of variables that made sense in BC2 and don't make sense anymore. I'm not sure modding BC2 will help me there, but I'll have a play. Yeah I read all the stuff and have a laugh. With my experience with BF2 and BC2 these things seem quite natural to me. I've never had any issues regarding a theory behind DecreasePerSecond but I won't spoil your fun though at figuring it out. Now, I haven't tested my theory either, but let's just say that the system heavily reminds me of BF2. I wrote a few lines about this matter in May 22, 2011 in these forums if you're still at a loss. Thankfully even my post back then is a bit cryptic to make it more challenging. As for BC2 modding and stats, Den only added the most crucial parts to his chart (I'd like to add, he did a pretty thorough job though with virtually all values being correct). There are plenty of unknown variables in BC2 because no one ever bothered to mod it and test it as the tool to pack files back into fbrb archives was released by dhwang just at the end of last month. Understanding either format definitely helps with the other, especially as BC2 files give you a good idea of which things aren't in my converted files yet.
|
|
|
Post by frankelstner on Nov 24, 2011 5:38:17 GMT -5
Alright I've tested it, my theory above is correct. Take a weapon with a ROF of 15 (sic) and the crosshairs will start to retract four seconds after the shot was fired.
|
|
|
Post by bel on Nov 24, 2011 7:05:25 GMT -5
You are correct. The recoil caps out within a few hundred ms of where I would expect it to if there were no recovery while firing at the maximum rate. I've tested with half a dozen guns in BF3, and with the hbar. Spread seems to behave the same, but it is much harder to tell and only the LMGs have enough capacity to hit their limits anyway. I didn't expect it to be so simple, and I also wasn't aware that BF2 was known to operate that way.
There are still a few things that are strange, like the M27IAR's horizontal recoil inverting with a bipod, but this accounts for the bulk of the confusion. I'm also not sure if the first shot multiplier does anything; it would appear not.
I appreciate the work you've put into this and am always happy to defer to those more experienced than I am, as here, but I don't particularly appreciate being told to go play with my toys while the adults wink and laugh. I'm not doing this for fun, I'm doing it because I want to know. If there's more to it that you're aware of, you may as well say; I can't help test a theory you won't explain in full.
EDIT: apparently I didn't refresh since your last post. Good to see it confirmed to that level of detail for spread as well. What weapon fires at 15 rpm? Or did you mod that into BC2?
|
|
|
Post by frankelstner on Nov 24, 2011 8:52:46 GMT -5
You are correct. The recoil caps out within a few hundred ms of where I would expect it to if there were no recovery while firing at the maximum rate. I've tested with half a dozen guns in BF3, and with the hbar. Spread seems to behave the same, but it is much harder to tell and only the LMGs have enough capacity to hit their limits anyway. I didn't expect it to be so simple, and I also wasn't aware that BF2 was known to operate that way. There are still a few things that are strange, like the M27IAR's horizontal recoil inverting with a bipod, but this accounts for the bulk of the confusion. I'm also not sure if the first shot multiplier does anything; it would appear not. I modded BF3, it takes me just a minute or two to change the floats I want to change. BC2 would not work in this case as it is an exception with the reduction being applied even while firing. I can give you the scripts I use if you don't mind changing a few bytes to mod the game on your own. The inverted recoil sounds truly interesting. As of now, I can't think of any reasonable explanation for that either. I appreciate the work you've put into this and am always happy to defer to those more experienced than I am, as here, but I don't particularly appreciate being told to go play with my toys while the adults wink and laugh. I'm not doing this for fun, I'm doing it because I want to know. If there's more to it that you're aware of, you may as well say; I can't help test a theory you won't explain in full. Yeah sorry, maybe I'm being a bit too cynical these days. I probably shouldn't direct my frustration at you. It's just that even after giving simple instructions and tools which require no programming knowledge but simply some copy-pasting, in the end there are still merely four or five people actually using my stuff. Yet at the same time virtually everyone gets excited when seeing some stats magically appearing out of nowhere. Additionally there are plenty of false ideas floating around as well some of which I will rectifiy now. E.g. Answer: -there's no need for the client to know weapon damage. Hitpoint calculations are done by the server. I think it is quite fortunate however that most of the time the values are indeed patched for the clients too. -projectiles, explosions and targets (and almost everything else) have materials assigned to them, you can look up the appropriate damage modifier in the material table (which is stored in the map, similar to BC2). For example the M95 rifle in BF2 dealt 190 damage, but the modifier against inf was 0.5 against the body and 1.5 against the head. This was done because the bullet proof glass material had a penetration variable (I can't recall the actual name) set to 100 which reduced the damage of all incoming projectiles by 100 and let them pass with reduced damage. Thus only the M95 made it past the glass with 95 damage left, enough for a headshot kill. That's what you kind of implied here: denkirson.proboards.com/index.cgi?action=display&board=general2&thread=3307&page=1#53540Sorry for going after you again; please take it as a compliment that you are one of the few people who actually care. Answer: -projectiles cannot be looked up with the current version of my tool. I had said the following before you made your statement: "Things missing are GUIDs mainly. You cannot see yet which projectile belongs to which gun. Or the materials used." Of course this leaves the question what "WeaponClass wc545x45mmWP" is good for. I'm fairly sure it is there for stats tracking or similar purposes. This variable most likely doesn't have any effect on actual gameplay. Another thing people might wonder about is the "AmmoBagPickupAmount 4294967295". Now, this number happens to be 2^32 minus one or FFFFFFFF. The number was stored as an integer just like some values before, so there are two options. Either the game does something different when given exactly this number or it simply treats it like any other. I'm pretty sure it does the latter. The devs wanted this variable to be infinite, but this would require adding another bool just for that. A much simpler option is using an insanely high value which cannot be possibly reached in a normal match.
|
|
|
Post by bel on Nov 24, 2011 23:06:16 GMT -5
I modded BF3, it takes me just a minute or two to change the floats I want to change. BC2 would not work in this case as it is an exception with the reduction being applied even while firing. I can give you the scripts I use if you don't mind changing a few bytes to mod the game on your own. I'll have a go if you're willing to send them. I'm a complete noob, but I might be able to muddle through. Also I had a look for BC2 tools, and I couldn't find any uploads that still worked. I can get your xml-fbrb and dhwang's fbrb repacker from the code posted on bfeditor, but the original unpackers all seem to be dead links. Yeah sorry, maybe I'm being a bit too cynical these days. I probably shouldn't direct my frustration at you. It's just that even after giving simple instructions and tools which require no programming knowledge but simply some copy-pasting, in the end there are still merely four or five people actually using my stuff. Yet at the same time virtually everyone gets excited when seeing some stats magically appearing out of nowhere. I did find it a little interesting that ten minutes' worth of downloading a script and pressing go gave me access to a world that just about everyone was fascinated by, but nobody seemed to want to visit themselves. I'm certain that playing the returning tourist gave me plenty of space to propagate false ideas. As an example, I'd seen you post about the materials before, but I guess I assumed you were putting your time into writing the tools and weren't so interested in digging through every line of the actual output, and so might have missed the weaponclass line. If you'd made it obvious at the time that you knew more about that as well, I'd have shut up much sooner and everyone might be less confused. For the record, I did realise it wasn't the actual projectile type when I noticed that there were projectiles labelled X_Carbine, X_Bolt but only WeaponClass wcX for each of those guns. I also saw the ammo bag thing. I assumed it was a highest unsigned int, but it's still a little strange. You definitely get ammo from support packs in increments of like 50. The next thing I want to do is see if I can use something like AutoHotkey to get precisely-timed clicks every firetime +n and abuse the spread recovery... all in the name of science, of course. I also want to have a look at the horizontal recoil stuff in more detail; in addition the bipod, it seems like a few guns have directional recoil in x when they shouldn't.
|
|
|
Post by frankelstner on Nov 25, 2011 4:36:13 GMT -5
|
|
|
Post by balberoth on Nov 25, 2011 6:41:43 GMT -5
The recoil decrease is multiplied by "ShootingRecoilDecreaseScale", which for automatics is mostly 0.1, so the actual recoil decrease per second is 1.8, not 18. In semi automatic the "FirstShotRecoilMultiplier" applies every time, so on the F2000 for example the recoil is not actually 0.2 when firing single shots, it's 0.6. I've worked out how to calculate the minimum time required to allow recoil to reset 18 degrees per second * 0.1 = 1.8 degrees per second, so it takes 0.5555555sec to recover 1 degree, which is 555ms, this can be used to calculate optimum ROF. Formula for assault rifles w/HBar/Silencer for semi-automatic allowing recoil to reset between shots is as follows:
555*(recoil*firstshot*hbar*silencer)
HBar value is 1.2, Silencer value is 0.9.
AEK with no attachments: 555*(0.2*3*0*0) = 333ms
M416 with HBar: 555*(0.28*2*1.2*0) = 372ms
F2000 with Silencer: 555*(0.2*3*0*0.9) = 299ms
G3A3 with HBar: 555*(0.55*1.2*1.2*0) = 439ms
Edit: To convert ms to RPM use this: 1000/xxxms*60 = xxxRPM So F2000 with Silencer: 1000/299*60 = 200RPM
As you can see, the rate of accurate fire in semi is very slow, I'm sure this can also be used for the semi-auto snipers, incidentally for me it points towards HBar not being very useful for semi (which is supposed to be it's primary use) because it magnifies the recoil, and as a result increases the time needed between shots, so although it decreases spread it means that you can put less rounds down in the time the target is exposed. At long ranges, when you factor in gun sway and imperfect distance gauging/drop compensation it's probably better to have a bit of spread but be shooting faster, as you're more likely to get enough hits than if you were firing more accurately but slower.
|
|
|
Post by bel on Nov 25, 2011 8:37:17 GMT -5
Uh, regarding the tools, I've just had the awesome idea of creating a cas_11 file containing just a single weapon. The way I've done it until now involved changing the existing cas files. In the end it'll hopefully take just a small tweak in the cat file while the weapon can be freely modded. Good luck. lol. I'd got to that several times and figured it looked an awful lot like the "this file can't be found" error. Turns out it actually says this file can be found. I need to work on my Italian. The recoil decrease is multiplied by "ShootingRecoilDecreaseScale", which for automatics is mostly 0.1, so the actual recoil decrease per second is 1.8, not 18. In semi automatic the "FirstShotRecoilMultiplier" applies every time, so on the F2000 for example the recoil is not actually 0.2 when firing single shots, it's 0.6. Have you tested those in-game?
|
|
|
Post by frankelstner on Nov 25, 2011 10:15:18 GMT -5
Alright, here is a modded cat file and a cas_11 containing just the AEK main file: www.gamefront.com/files/21023744/cascat.rarHere is the analyzer tool: www.gamefront.com/files/21023745/%233x4bytesanalzyer+2.0.pyUse the analyzer on the AEK in the "safe" folder and it will create "AEK971 analyze.txt" showing the absolute offset for everything in the cas_11 file. E.g. "MagazineCapacity fdc0 0 7368" means ammo is stored at offset 0x7368 in the cas file (fdc0 means integer, so there are 4 bytes to consider). The weapon is already slightly modified with mag capacity on 3100 (if I recall correctly) and a ROF of 2000. The issue with arrays still remains. If they are handled like anything else their offsets will overlap with offsets later on. In fact an entire array containing lots of things is given just 4 bytes in the section with most of the data. The array stuff is then saved separately at the end of the file. So once again, ignore arrays. The AEK can be found in the last Russian mission, "Kaffarov" I believe. I chose this gun because it comes with a laser. Might make it easier to experiment with attachments in the future.
|
|
|
Post by balberoth on Nov 25, 2011 13:19:25 GMT -5
bel: Yes, I've tested them in-game, I've set up a script to click the mouse every 333ms and confirmed that in semi-auto with the AEK with no attachments the bullet impact point does not rise as a result of recoil, if I make it click any faster the impact point climbs. I've not tested individual pieces of data that i have used to arrive at this figure as i play on XBox, so I can't rip the files myself as I don't have them on PC, but the experiment performed with the timed mouse clicks gives the result I would expect if my calculations were correct, and if i deviate from the timings the result changes in the expected direction, so it's scientific enough to satisfy me, although I'd be happy if someone could test it more precisely.
|
|
sleep
True Bro
Posts: 10,189
|
Post by sleep on Nov 25, 2011 16:38:37 GMT -5
(numbers) As you can see, the rate of accurate fire in semi is very slow, I'm sure this can also be used for the semi-auto snipers, incidentally for me it points towards HBar not being very useful for semi (which is supposed to be it's primary use) because it magnifies the recoil, and as a result increases the time needed between shots, so although it decreases spread it means that you can put less rounds down in the time the target is exposed. At long ranges, when you factor in gun sway and imperfect distance gauging/drop compensation it's probably better to have a bit of spread but be shooting faster, as you're more likely to get enough hits than if you were firing more accurately but slower. great number crunching, testing and analysis (thanks, much appreciated!), but i disagree with your conclusion about the HBar. since its multiplier is 1.2 we're only talking about a 20% increase in optimum time between shots; for the M416 it's 310ms vs 372ms, or a 16th of a second difference. personally i don't even think i could consistently adjust my rate of fire by that small of an increment if i tried. meanwhile it's basically doubling your accuracy / halving your cone of fire, which at longer distances is a huge deal. personally i'd rather shoot 1/16th of a second slower on semi and have 6 out of 8 shots hit instead of have to waste twice as much ammo and almost twice as much time to kill a guy, especially considering that gives him a lot more time to find cover and survive.
|
|
|
Post by balberoth on Nov 25, 2011 17:19:23 GMT -5
sleep: I see what you're saying, but that isn't what happens in practice. In practice you fire in semi as fast as you can while still keeping your sights on the target, people mostly run away when being engaged at long range, and they fairly often manage to hide, so you want to put as many rounds down as possible while still maintaining accuracy, given the gun sway on all weapons, no matter how slight, and the fact that you don't fire fast enough to have the spread increase over the base figure, the chance of the slight cone of fire decrease helping you is very small, what's hindering you is gun sway, as your reticule moves slightly off the target, and you can't compensate for it to any real degree, so what you're trying to do is put the rounds down quickly enough that your statistical chance of getting enough hits is as high as possible., basically filling that cone with rounds until enough hit. Heavy Barrel makes it more difficult to do this, the fact that you're putting less rounds down means you get less hits, and the decrease in spread (which isn't significant as the value is so small already) isn't enough to make up for it with the smaller cone. Try it, the F2000 has a base spread of 0.4, fire it at a wall slowly on semi-auto without moving the mouse at all, then do the same with the AEK (base spread 0.2) you won't see any difference in the pattern of the shots, I've tried it, Spread doesn't seem to be significant here, I'm sure it is with LMGs firing sustained bursts of automatic, but on semi the spread is low enough that it's less of a problem than gun sway, and certainly less of a priority than fire rate.
|
|
sleep
True Bro
Posts: 10,189
|
Post by sleep on Nov 25, 2011 17:47:04 GMT -5
well, the way you talk about this gun sway that can't be compensated for, i can only assume you're using a rifle scope or something. i experience no noticeable/significant sway using a RDS, holo, or 3.4x iirc, and i see no use for any higher powered optics than that unless you're using a recon rifle. without sway, i find a 50% reduction in cone of fire to be anything but "slight"! but to each his own thanks again for the numbers etc.
|
|
|
Post by bel on Nov 25, 2011 20:52:19 GMT -5
Alright, here is a modded cat file and a cas_11 containing just the AEK main file: www.gamefront.com/files/21023744/cascat.rarHere is the analyzer tool: www.gamefront.com/files/21023745/%233x4bytesanalzyer+2.0.pyUse the analyzer on the AEK in the "safe" folder and it will create "AEK971 analyze.txt" showing the absolute offset for everything in the cas_11 file. E.g. "MagazineCapacity fdc0 0 7368" means ammo is stored at offset 0x7368 in the cas file (fdc0 means integer, so there are 4 bytes to consider). The weapon is already slightly modified with mag capacity on 3100 (if I recall correctly) and a ROF of 2000. The issue with arrays still remains. If they are handled like anything else their offsets will overlap with offsets later on. In fact an entire array containing lots of things is given just 4 bytes in the section with most of the data. The array stuff is then saved separately at the end of the file. So once again, ignore arrays. The AEK can be found in the last Russian mission, "Kaffarov" I believe. I chose this gun because it comes with a laser. Might make it easier to experiment with attachments in the future. Fantastic, thanks... except I haven't played the campaign at all and had no real desire to, so it'll be a few days before I get there. For science! Any chance you could change an earlier full-auto gun and save me the time? bel: Yes, I've tested them in-game, I've set up a script to click the mouse every 333ms and confirmed that in semi-auto with the AEK with no attachments the bullet impact point does not rise as a result of recoil, if I make it click any faster the impact point climbs. I've not tested individual pieces of data that i have used to arrive at this figure as i play on XBox, so I can't rip the files myself as I don't have them on PC, but the experiment performed with the timed mouse clicks gives the result I would expect if my calculations were correct, and if i deviate from the timings the result changes in the expected direction, so it's scientific enough to satisfy me, although I'd be happy if someone could test it more precisely. Alright, cool. That's interesting. I'd tried to do that on PC, but the PC version seems to block keyboard hooks; AutoHotkey and its ilk become so unreliable I couldn't do it. If anyone knows an alternative, it'd be appreciated. well, the way you talk about this gun sway that can't be compensated for, i can only assume you're using a rifle scope or something. i experience no noticeable/significant sway using a RDS, holo, or 3.4x iirc, and i see no use for any higher powered optics than that unless you're using a recon rifle. without sway, i find a 50% reduction in cone of fire to be anything but "slight"! but to each his own thanks again for the numbers etc. Keep in mind he's on console. Some/a lot of things are going to be different for recoil and sway. MinAngle is probably under the xbox aim assist most of the time as well.
|
|