Well, I am not going to argue against the spread values too hard, because it's up to you what you are going to set as accuracy of weapon fired by humans. Some may say people should and were firing more accurate, some other that average soldier was very inaccurate. If one can be satisfied by RO fast fire accuracy, the times when he is - undisturbed - aiming very very precisely and still missing at 100m because of the spread (or at 300m with a sniper rifle, spread of 0.85m in diameter at 300m) can be frustrating.
I would only say that in my personal opinion, the random spread should reflect the weapon dispersion (technical thing), and the human errors should rather be done by tunning the recoil and weapon movement during firing. Which is unfortunately more difficult and much more time consuming than simple setting putting spread value... But I believe it's worth to do in a good FPS. The pure spread + recoil movement type of weapon modeling is from Counter Strike era...
Next thing is, when there is no adjustable rifle sights in RO yet, wouldn't it be better to zero the sights on 200m for example, than for 0m like now ? The bullet patch would be very slightly up on 100m (still smaller than head size), zeroed on 200m, and drop only then. Currently it drops from muzzle on. Does anyone fire rifles with sight set to zero (do rifles have such a setting at all) ? Or rather zero it to 100 or 200m if he have to use constant sight ?
And at last, unlike in tanks, where there is an offset between gun barrel and sight, seem for me that in firearms the sight seem to be exactly in the bore axis. Like I was aiming trough the barrel. If the sight view was shifted up by correct amount, proper zeroing of weapons would give more realistic effects... you know, two distances at which the bullet is crossing the line of ainm, near and far, and relatively flatter trajectory. But this needs coding, and zeroing the weapons on some range instead of at the muzzle, is just matter of setting some fixed addedpitch values for each weapon.
By the way, I'm not attacking or slamming TWI - can you attack someone by just stating the facts ??? (like there is error here, this is clearly wrong)
I also express my personal opinions about some aspects of your product, I say what I don't like (and sometimes what I do like). Is that attacking ? Well someone can feel it as a "slamming" but it's his problem not mine. I also often say that something could be done better (to be honest, EVERYTHING on this world could be done better) but not sure how this could be read as saying that you or TWI are too stupid to do it right. You are far better coder than me, and you could easily code everything better than is is, and better than I did or plan. But for some reasons you did not. My personal opinion about those reasons in some cases could be taken as offence, so I will not express it here. Of course reasons can be different than what I think, few words of explanation would clear this, but companies don't like to explain their reasons, so I have to stay with my opinions.... Believe or not, I have also nothing personal against any TWI member... exept one. And I think I have good reasons to not like him, because I believed and still believe he was saying/writing things that he knew were untrue - and this is something I simply don't like. Yes I know it's uncommon to don't like people because of such details, neverthless I'm untypical and I do. Untill I find that is was me who was wrong, and I misjudged someone, but it doesn't seem so far.
Returning to infantry weapons - shouldn't first few shots of MG burst be rather accurate, and the rest go (semi) randomly ? Or ALL randomly spread from the first bullet? Would it be hard to model different spread for few first rounds in the burst ? Or at least switch fire modes from single to auto only after 2-3 first bullets of a burst left the barrel ?
I surely not require the weapons to have dispersion equal to ballistic tables, but would be nice to have weapons that with calm shooting would pass the test firing/sight calibration test, and meet the required criteria for single shots (of first round in the burst), don;t you think ? Or not more that say 150% of it... Of course the man under battle condtions, supressed, afraid about his life, would usually not aim so carefully as on test firing. But are the soldiers ALWAYS supressed, there are no calm shots fired at all ??? Maybe relate the spread to the supression of the player, as many players ask for, and keep it about what a calm average or good shooter achieve for times when no one is shooting in his direction for some time, and reduce it to current values when a player is shoot at with bullets passing near ? (or even bigger dispersion, if the under MG auto fire... real soldier would even not raise his head at all with MG burst passing over his head, so player fire could be totally ineffective under such conditions, unless he's on drugs...).
(The MG mounted in a tank (coaxial) should be free of human errors and fire with almost pure dispersion, close to the specs (let's sat 1.5 of the specs from tables). But it's rather off-topic in this discussion. I posted a table of Russian DT tank machinegun dispersion from KV or T-34 tank in previous thread.)
But what bothered me much more, was the infantry weapons round ballistics. I always thought, that tank modeling is as is, because it's only addition to the primary game that is infantry fight, and the infantry weapons realism is way higher. So I was very surprised to see that even the key weapon parameters like muzzle velocity, velocity profiles and ballistic drop were not correct. Unless I'm wrong (possible) the ballistics are wrong for two easily recognisable reasons - wrong muzzle velocities because of wrong factor used during calculation of Unreal speeds, and wrong velocity profiles because of wrong balistic coefficients - at least for rifle bullets that I have checked. It's not very hard to find ballistic data or even velocity profiles for German 7.92 or Russian 7.62 rifle bullets. On average, the ballistic drop is about twice of what should be, and what is written in tactical and historical guide about the weapons.
I could make a mistake in tests, with a factor of 2, and I will check again (can post results and pictures here, no problem). Can also release this test mut, and everyone can check it on their own, are you against this maybe ? But isn't it unlikely that the ballistic drop would be correct, with incorrect muzzle velocities ? At the beginning I was not sure what the "official" scale of RO is, if a meter is equal 60.352uu or 52.48uu, because the second value was used sometimes. Now I know that official scale is 60.352 (with almost everything scaled with this, including terrain, tank models and tank shell speeds) and 52.48 is used only for one purpose (tank shell ranges), probably because of error. To my greatest amazement I saw that muzzle speeds of infantry weapons are all calculated by 52.48 factor. Now, if the terrain ranges are by 60.352, how can you get correct ballistic drop from that (without changing the gravity) ??
Kar98 bullet:
Speed=37808 // 2363 fps
DP28 bullet:
Speed=44082 // 2760 fps
2363fps is 720m/s and you mean this as Kar98 muzzle velocity. OK. Don't know if it's proper value, I seen different, but you are the expert here.
2760fps is 841m/s for DP28 velocity. OK.
But 37808uu is really just 626m/s in RO, and 44082uu is just 730m/s ! Speed is definied by distance and time. Distances in RO are by 60.352, right ? So speeds have to be too, or you get too long times. And longer time is higher ballistic drop.
I'm not so sure about the second error - wrong ballistic coefficients, as about simple physics above. But from ballistic data of 7.92 and 7.62 bullets I found on the net (some with velocity drop tables, some saying at what range bullet went subsonic) it seems the BCs used in RO are not correct and lead to excessive velocity drops.
And when checked in game, the drop seem about two times too big. I will check again, though and make exact measurements.
Maybe you projected the ballistics correctly (only had different BCs... but they can't be too different from what I found) but you don't know about the scale conversion error (60.352 vs 52.48) ? Check the code please and say what you think about bullet speed conversion to uu. Or try to disable the spread, and check if the drop is what you are expecting. This have nothing to do with native ballistic calculations, which seem to be just OK, at least for the game. It's bad muzzle velocity and BC, so bad input data.
For comparison in tank shells, the muzzle velocities were ok (calculated by 60.35), and most BCs were too low (but BC and velocity profiles of tank shells are harder to find than those of rifle bullets).
***
----> astat
Panther should be DEFINITELY easy target for 76mm from the side. It is designed to be so (in the mod). It has now 55mm of side hull armor simulated (was 60mm) so it should be easy to penetrate from up to 1000m. But 30deg of additional angle would increase protection to almost 70mm, to 95 mm at 45deg, and you can't penetrate it at 60deg additional angle (so firing 30deg from the front) because it becomes over 160mm... The side turret factor is 47mm.
In the future, the Russian shells will have lower slope modifiers so it wouln't be that bad. Well the whole armor modeling will be more advanced, currently this is only "armor factor" so trated as vertical plate with protection of 55mm RHA, and generous generic slope modifier of 1.6
But if you have problems penetrating (or do you mean killing?) a Panther directly from the side (with no additional angle >20deg) from T-34/76, then something is definitely not working as it should. Personally never had problems destroying panthers from the side.
The mod is meant to be improved further. The armor and penetration values are not yet fine tuned, it's only a first approximation, just values that I would not be ashamed to use in beta version
. Currently the mod development is suspended for some time, for reasons I already described. Hopefully I will return to it at some day in near future. Now I only use computer doing historical research and developing my penetration model from time to time...