Latest topics

» Expanded Units MOD Update:
by KurtTank Yesterday at 8:27 pm

» New & Updated Models(Нові Моделі)
by will73 Yesterday at 6:24 pm

» STA 3.3 corrupt file ?
by lockie Yesterday at 6:09 pm

» Video from SF
by geoweb35 Fri Feb 23, 2018 10:35 am

» Ride the Tiger
by woofiedog Wed Feb 21, 2018 9:13 pm

» AnotherBattleEffectMOD(ABEM)for STA3.2
by woofiedog Wed Feb 21, 2018 8:50 pm

» Screenshots funny and serious :)
by woofiedog Wed Feb 21, 2018 8:47 pm

» Варіанти Т-34-76 в житті та в грі (T-34 variants live and game)
by woofiedog Wed Feb 21, 2018 8:18 pm

» Ukraine fights against enemy.
by lockie Tue Feb 20, 2018 3:55 pm

February 2018

MonTueWedThuFriSatSun
   1234
567891011
12131415161718
19202122232425
262728    

Calendar Calendar


    Hello everyone!

    Share

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Hello everyone!

    Post by Stah on Wed Dec 09, 2015 10:37 am

    Hello everyone!

    I'm new to this forum, but not new to Steel Fury. I play this excellent sim and experiment with various aspects of it for a long time. I liked it from the first glance, but I just fell in love with it one day, five minutes after I decided to turn off outside view and aim indicators, and... a GAME instantly became EXPERIENCE - harsh, difficult, convincing and really immersive (just like in my beloved, heavily modded Silent Hunter - in fact, SF became for me kind of a land version of SH Smile. So I would like to express my greatest appreciation for the SF creators and whole community of ingenious modders and mission designers alike.

    Over the time I've found solutions to a lot of things that bugged me in the game (like armoured plants, un-realistic sound scene, tanks balancing on branches and so on), but there is a number of problems for which I couldn't have found solutions myself - and I think I never saw anyone asking for them, either. My question is: should I place them in there - as a newcomer - or rather create some new topics, one for each issue? Sorry for my incompetence, but this is - and I'm not kidding - the first post ever in my life on any forum Embarassed Very Happy

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Hmm...

    Post by Stah on Wed Dec 09, 2015 12:39 pm

    Ok, my question was probably quite stupid - as I can see, almost every newcomer create new topic or writes a post in existing one. I hereby withdraw my question Smile
    avatar
    lockie
    TSF Member
    TSF Member

    Posts : 3764
    Join date : 2014-07-24
    Age : 50
    Location : Ukraine, Kyiv

    Re: Hello everyone!

    Post by lockie on Wed Dec 09, 2015 1:38 pm

    Hi, Stah
    U're welcome!
    Don't hesitate to ask any questions in this topic. It was created specially for beginners(approximately a couple months ago). The purpose to remove often asked questions from the main topic to this supported one.
    If question will be important or repeatable or has some other interesting aspects, then it will go to the FAQ.



    Technopiper

    Posts : 144
    Join date : 2015-07-30

    Re: Hello everyone!

    Post by Technopiper on Wed Dec 09, 2015 2:23 pm

    Hello and welcome. I'm pretty new here too, both the forum and the game. Only discovered Steel Fury this August. I'd have to say the single most valuable assets of this game is right here at this site. The people here are most friendly and helpful. lockie and frinik always had an answer ready for any of my problems and, if they didn't, took the time and trouble to find out for me. This forum is the only reason this game is still kicking ass after 8 years since launch.
    avatar
    frinik
    TSF Member-Golden Feather
    TSF Member-Golden Feather

    Posts : 1372
    Join date : 2014-07-27
    Age : 61
    Location : Bogota-Colombia

    Re: Hello everyone!

    Post by frinik on Wed Dec 09, 2015 2:51 pm

    Hello and welcome Stah.AS for asking questions just look at the recent stickies to see if one matches your query .Otherwise don't be shy and make a new one.

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Wed Dec 09, 2015 2:52 pm

    Hi, Lockie.

    I asked this question precisely because I thought this place to be meant to ask fundamental questions by newbies. As I am new to the forum, but not to the game, and the era of simple questions is way behind me Very Happy I wasn't sure what to do. Currently I created new topic in "game editing" in which I touch the periscope speed issue, and I intend to add some other topics concerning issues which I never saw discussed before (or only touched).

    Hi, Technopiper

    Yes, I could't agree more. I've been sort of passive user of this forum for months Wink and the amount of knowledge I amassed from here is really huge. And I'm, like you, impressed by the kindness and helpfulness of people here...
    avatar
    Pointer

    Posts : 163
    Join date : 2015-04-18
    Location : Poland

    Re: Hello everyone!

    Post by Pointer on Thu Dec 10, 2015 2:47 am

    Stah wrote:Over the time I've found solutions to a lot of things that bugged me in the game (like armoured plants, un-realistic sound scene, tanks balancing on branches and so on)
    Hi,
    I think it would be very interesting to all, if you present those solutions. Especially those balancing on lying trees.
    avatar
    kapulA
    TSF Member
    TSF Member

    Posts : 206
    Join date : 2014-07-29
    Location : Split, Cro

    Re: Hello everyone!

    Post by kapulA on Thu Dec 10, 2015 2:56 am

    Agreed with Pointer. And I'd like to re-check my bushes, they seem to have a rather effective armor atm. Smile

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Thu Dec 10, 2015 1:47 pm

    frinik wrote: Increasing the weight of the vehicle in the tech_cfg file is one possible solution.

    Of course. This is exactly what I did. Nothing fancy, really, that's why I didn't make a big discovery of it Smile
    In addition I increased (proportionally) engine power. And spring stiffness (3rd wheels parameter), but this one - not proportionally (this parameter shold be set up individually for each vehicle, so as not to cause too rough a ride).
    My idea of weighing up came up some months ago, when I started driving ultra-light German tanks (first two Panzers especially, but also Fiat and (35)t), which are just undriveable in plants covered area. I started by multiplying weight and power exactly by two, then by three, than by four, just as an experiment (in "four" mode they were sitting on ground, unable to move, so I came up with the spring stiffing up scenario). If I remebmer correctly, 2x proved to be satisfying multiplier (maybe 2.5 - on the lightest machines). This is just a matter of taste, of how you feel the machine when driving. Of course I didn't want to mke them uber-powered-tree-crushing machines - these changes should be made with caution, to preserve overall character of each vehicle.
    As an afterthougt I experimented with uneven spring stiffness (different for middle and outer wheels), and with the... (what's the English word??) - of how much up and down wheel goes (12th and 13th parameter). Interesting stuff - but just as geek fun, without much impact on playing style (ok, overall stiffness of vehicle has great impact - it affects the gunsight bouncing speed/ratio after a hard stop...)
    Now I gradually create my humble, little mod, adding to it edited version of tech_cfg file of every tank I happen to play (adding to "notrace" list the vehicle main gun, but this is another story). Not every tank needs such a significant amendment, of course.

    As for bushes - of course I edited armor map, this was my first "mod", long time ago. For every plant and for this light village fence too, practical mg-fire-in-the-village killer Laughing (you know which one I mean, I just lost the English word for its material - it's weaved). And I produced amended version of every weather ("seasonal" rather) mod I use (they include some plant armor maps too). I used full black (0,0,0) after some consideration - of course, my bullets (even mg bullets) are going through tree trunks, but it only sounds unacceptable - in a combat you really don't see it (nobody aims the enemy with the tree in the sight line 50m from the gunsight) - after all, this is much lesser evil than heavy AP shell ricocheting on a trunk of a young birch Very Happy I play with it for more than a year and I never encounterd situation which I could descibe unrealistic and immersion-breaking.

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Thu Dec 10, 2015 3:17 pm

    But this whole weighing up issue leads me to another, more universal issue, that I wanted to start anyway. Experimenting with the physical parameters (shy in the beginning, mad at the end Very Happy ) led me to conclusion about how to treat in-game physical values. I think they are surely not physical values per se, rather than some equivalents of real-world physical values. What I exactly mean is this: if the game engine, physical model, or however you name it, depicted real world directly, assigning 50 000 kilograms of weigt to certain vehicle wouldn't cause it to balance on broken birch or (we all saw these) flipping over after driving on some wooden structure, just like plastic toy pontoon on a wave. I saw a lot of discussions here and on other forums about historical speeds, ratios, taus, masses, coeficents and on how to implement them in a game (this or that). This approach, I think, won't work in many (even most maybe) game aspects - in-game values should be treated, I think, like sort of fine-tuning tools, mostly abstract, only vaguely conntected to their real-world counterparts. Another example, even stronger: I found out that increasing bullet weight increases its penetration capability, but does nothing to its ballistic curve - doesn't flatten the arch and doesn't shorten shot distance Shocked This is little off-topic (or rather over-topic Smile ), but I think this is strictly connected with the tank weight issue: setting realistic value causes non-realistic effects - and vice versa.

    Coming to the point at last Very Happy - less than year ago I started a huge project of making the SF world as realistic (in all mechanical aspects) as it is possible, because I found a lot of strange things in it (i know, this is not new thing on this forum). So I started gathering lots of data concerning weapons and vehicles and building them up in a huuuge table (two tables: German an Soviet: half a square meter each, packed tightly with 4-6p. print Shocked Very Happy ), including everything about each vehicle: engine parameters, crew, armour parameters, visiors, turret speeds, angles, dimensions, speeds, additional armement, months they were in duty, etc, etc. - vehicles grouped under their main guns, and every guns descibed in all aspects, up to average loading speed (very importand parameter during battle, as we all know) - and under each gun all shells used by it, with their parameters, kinds, speeds, weights, notes aboud period of usage, and penetration values for four distances with angles...

    I also started some preliminary penetration/ricochetting test on a Stug that I armor-map-prepared for the task (I did't have any kind of reference piece of steel, and Stug's flat vertical area between tracks seemed convenient) and managed to fine-tune T-34/85 AP projectiles, so that they ideally matched historical data (no huge difference: they were sligtly underpowered, as I remember). But then I started reading really serious stuff about the penetration issue and I came to this conclusion: this is not he way. First, the real-worl historical penetration fact is not what game understands as penetration - secondly, there are these differences between German and Soviet methods, thirdly, we don't know exacly what steel was used as a target, and if the same steel for every tests, etc, etc., etc. - I think you know it as good as I or better than me. I think we just can't fine-tune the SF armour universe by picking each parameter and carefully measuring it, and then setting up the armour maps and parameters and being sure that it will work. I'm afraid it just won't, beacuse - again - SF parameters are not real-world parameters. Beside this, every tank is in gun/armour correlation to all other, and if you will be forced by real-combat tests to tweak any value (which is unavoidable), the whole equilibrium of painstakingly picked parameters is destroyed...

    I still think we can make, as someone somewhere wrote "strong tanks strong, and weak weak" - in a real-life propotions - but not by measurements... God, I got to post it, it's too long now Very Happy

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Thu Dec 10, 2015 7:05 pm

    ...ok - to make things as short as possible - I think of changing point of view from physical modelling to combat situation modelling - just to start with the central tank of this theatre (any of T-34's), making it a point of reference, and put it in series of combat situations with most important enemy counterparts (1 on 1 duels), tweaking armor coefficent (if necessary) and shell's penetration/ricochet/etc coefficent - not to get penetrations or avoid them at certain distance/angle/kind of steel plate, but to just get the outcome of every duel in accordance with historical sources (with some error margin, of course). And then another tank...

    Ok, this plan may seem like an overkill (after all, it's only game Smile) but after I got penned frontally in Jagdpanther by bunch of T-34's from, like, 800m, and after winning several duels in a row with Tiger commanding a platoon of T-70 cyclops , I became a bit suspicious, to say the least scratch - and I'm just curious what may come out of such an experiment.

    Donken

    Posts : 183
    Join date : 2014-07-25
    Age : 32
    Location : Sweden

    Re: Hello everyone!

    Post by Donken on Thu Dec 10, 2015 7:43 pm

    Not to burst your bubble but its not going to work well. Almost all tanks have weaknesses due to the game engine. So even if a shell should not penetrate or give spalling, you are getting that. That is mostly why you get killed in lots of awkward ways that you should survive. I think you have taken a look at my thread i started when i made a practice target? You find lots of info in there.
    And there are also 10 other things you need to take in account. Like armor strength, armor spalling, the armor map, armor quality and so on. A tip is to use the practice target, and set the thickness on it instead Very Happy

    What you want to do is something i have wanted for many many years. But i always got downvoted/yelled at so i didnt bother anymore =)

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Thu Dec 10, 2015 8:17 pm

    Donken wrote: Almost all tanks have weaknesses due to the game engine.

    I know, you're talking about these empty edges or whatever they're called - I've heard of them and even experienced them myself, being killed in PzIII with a soviet hand grenade Very Happy .

    I know there is a lot coefficents to take into account, almost none of which corresponds directly to any real physical value (even SF concept of penetration doesn't make sense in real life, because penetration never was boolean Smile  - it's gradual, like soviet 75% or German 50% penetration) - so this is exactly my point: set the overall ratios between combat capabilities of vehicles ("who wins with who") rather than trying to do the impossible: mapping detailed real-world facts on a model that has its own, different logic (like 0-1 penetration, for example).

    But ok, maybe this make no sense - if I were all-knowing I wouldn't have put my thoughts to discussion... Btw, Firnik made a mod like we're talking about - but it doesn't work with my configuration ("Frinik Realistic Update" I believe it was called). Have you tried it? He remarked it made things much more realistic for him.

    Aha, I can see now - I remarked only shell's coeficients, so you thought I didn't take armor maps, armost_str, qual etc under consideration... No, no - I've made lots of testing with them, I am familiar with SF shell/armour modelling quite well.

    Donken

    Posts : 183
    Join date : 2014-07-25
    Age : 32
    Location : Sweden

    Re: Hello everyone!

    Post by Donken on Thu Dec 10, 2015 9:37 pm

    Hehe its open edges Very Happy but its alot more then that. Lots of tanks also have two sides polys and two objects touching eachother. That screws up the calculation by either ignore the armor map or its using the lowest value. That means full penetration without doubt. A few examples would be the fenders on lots of tanks. Also mantlets and corners when the corner polys are to close.
    Penetration is not boolean or linear. If you have seen all the configs already you know its all random to a certain degree. It is possible to make like 75% of the same shell fail to penetrate, or only 10% if you want that. So pretty much everything is calculated regarding the shells. I only know of three things that are not taken into consideration and that is barrelwear, airhumidity and temperature. So it is possible to make it exactly as you want. The problem here is not within the shell, barrel or armor configs. I know that probably 60-70 percent of most weird stuff are because of the 3D models.

    No i have not tested hes mod. But i will do sometime. Thanks for the tip!

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Fri Dec 11, 2015 10:15 am

    Ok, the 3D models issues are unfortunately new to me - and I frankly don't know much about 3D modelling. Your revelations here are really depressing, I must say :/

    By "boolean" I meant that in SF penetration is - or isn't. By "linear" (ok, it was stupid) I meant that in real life armour could be penetrated only partially, and I'm afraid what you're referring to is the probability of penetration by certain projectile, set by one of common_res coefficents, am I right? You used word "random", so I think this is what you had in mind. I didn't see any coefficent responsible for degree of partial penetration...

    But these are details, really. The problem is, I was misunderstood as to my point. The main point of my philosophising Smile , long-as-hell Smile post wasn not "hey guys, there's something fundamentally wrong with your sim and I plan to fix it" - he he he, it really wasn't what I was meaning. I wouldn't think that way, 'cause there are a lot of real tanksim pros here, and I'm just an aspiring SF fan.

    My point was precisely this: "tweaking with this and that I came to conclusion that IF one wants to change something in SF, one shouldn't always rely on real-world values and tweak the game by inserting them there and assuming they will make SF behave like in real life - but rather treat them as somewhat abstract coefficents and tweak them as long as one gets desired effect in the game, no matter what their value".
    The story about me drowning in an ocean of detailed data was just an explanation of how I came to this conclusion. Now I think that data have secondary importance, and what unit wins what unit is the only importand question - and imho this seems to be true precisely because of what you're talking about: the errors in some aspects of models, or any other possible errors. Of course, I may be wrong - I was just curious if any of SF-tweaking guys had similar thoughts.

    The examples of strange combat outcomes (this poor Tiger etc) were really aside, just as digression. My fault - I always talk chaotically, touching few issues at a time, in real life too Mad Rolling Eyes

    Donken

    Posts : 183
    Join date : 2014-07-25
    Age : 32
    Location : Sweden

    Re: Hello everyone!

    Post by Donken on Fri Dec 11, 2015 12:00 pm

    I get exactly what you say. My rant about the other things are just to tell that what you want is hard if not impossible. But of course, its worth a try. What you do cant be worse only better. So dont take it, that it is impossible or that we dont want it. Im just telling about the difficulties about it and hopefully get you thinking about what you can do about it and hopefully give you some workaround ideas Very Happy

    And about my tries to make everything using real world data (of what we can see and do from it) was to try to get it as real as possible.

    Ok, partial penetration, i dont think it is in the game. But the logic says there is no real difference on the "outcome" from a partial penetration and a big shell hitting and spalling a plate with no penetration. So one way to try to fix it is by change the spalling value. I do know there are a big difference in the technical point of view with armor weakness, spalling, bending, loosening of welds etc. But the result in the end is pretty much identical for the enduser., its a kill on a crewmember or knocked out tank!

    The main problem by making this tank win over this tank and this one will loose over that! because it did that in real, its a harder approach. You need to balance 50 other vehicles to get the reality of what you want Very Happy And that is what most of the STA guys have been doing. That is why you can see some very wierd configs like a ridicilous amount of armor strength on some vehicles etc, or very overpowered AP values on some shells etc, just to get it behave close to the real deal. Problem is by the law of physics, what you gain in some way, you loose in another. Some tanks are very overpowered against one type of tank, and very bad against another. Its just a bad circle that has started.

    My main idea from the beginning was to reset the configs on everything. And from there only put in real world values and see where it goes. I think that the technical bit in the sim is good enough to actually get the exact result that you want, IF all the 3d-models where correct. I do have some test files somewhere with a Panther G, t-34 and Pz4 with correct values and it works as it should. They are like 98% historical. The problem is they are way lot weaker then the STA original ones. But against eachother they are perfect! Now i only need to fix 50 other tanks, and that im not doing, took to much time to just fix those 3 (around 20h each with testing, remodelling, testing again, looking for facts and pictures and so on)

    I really do hope you are going to work on this, i will gladly help you with anything and share my experiance and ideas. So feel free to ask away Very Happy

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Fri Dec 11, 2015 1:45 pm

    Ok, I basically think the same way. Maybe my idea of fine-tuning everything through battle was far-fetched, I just felt tired with all the maths, tables etc and wanted to try radically different approach. In fact, this change is my thinking is only 4-5 days old - all the time before I was strict fan of real-life-values, physics and so on.

    So - if I got it right - there are three main uber-topics concerning the projectile/armour issue:

    1. errors in models 3D design (solutions in their construction that create - to simplify - "holes" in the armor) - and they can be fixed only with 3D modelling soft (modelling and then implementing the models again to the game with the Objects Editor)

    2. on the ballistic side: shell coefficents in common_res file: penetration force, projectile mass, probability of spalling, deflection (ricochet) angle and some others (to say only about AP's - HE, SH, smoke etc are different issue of course, I put them aside as for now)

    3. on the armour side: configuration of armour simulation system, e.g. finding correct values for and relations between: RGB value of arm_map and armour values in tech_cfg file (overall armour thickness multiplier and three coeficcients defining features of steel plate)

    As I understand:
    number 1 - it'is such a hell of work that this solution is out of question and we just have to live with the "holes" in armour (which seems to render all futher discussion obsolete Very Happy )
    number 2 - there seem to be some unrealistic settings (judging by some topics on this forum) but through tests on a reliable model of steel plate (just like yours) that can be done (time-consuming, but not a lot of controversy here, I suppose)
    number 3 - armour maps and values in tech files seems to me mostly correct, besides, they are probably easiest part of the equation - of course, there are always these differences in historical sources, or just lack if info, or just problems with numerous versions and semi-versions of certain vehicles (especially concerning soviet side), and I've seen numerous traces of this controversies in tech_cfg armour values.
    There are also gun characteritics (in common_res, I mean), but to me they seem to not have any affect on the outcome of hitting armour.

    Does it look that way? It is all there is, or did I missed something? I would probably have some detailed questions for you, but I need a solid base - a map of the terrain, so to say Smile

    Donken

    Posts : 183
    Join date : 2014-07-25
    Age : 32
    Location : Sweden

    Re: Hello everyone!

    Post by Donken on Fri Dec 11, 2015 2:00 pm

    1. Yes pretty much, it is a tremendous work. Remember that this mod is old, very old and for every new tank we have all learned new things about how to model it. That is not a problem really, the problem comes to the old models, many models are made +6 years ago and the authors have gone awol or doesnt have the original model file anymore. So its impossible to change them because we dont have the model, one solution would be to ripp them and redo them but that is as much work as to make it new from scratch anyway.
    2. Yes again, it is possible but require tons of work and probably a total rewrite in the configs.
    3. Same here, this is probably the easiest aproach. You dont always get the realistic result on all vehicles but atleast you can make a change and hopefully get a result that is realistic, for the moment or on just the vehicles used in that specific mission!

    Shoot away, im always here everyother week or daily by PM. If you tell more exactly what you need as a solid base i can hopefully help you with it, to get you started Very Happy

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Fri Dec 11, 2015 2:30 pm

    To be honest, I am really disappointed with this 3D models issue. First, this is only one thing here beyond my reach, second, I didn't think this is so seroius. I always thought that it is some minor 3D issue concerning few models, and some geniuses for which it is a breeze, piece of cake and walk in the park will fix it in no time Very Happy  Seems now like a most serius problem. As I understand, it concerns not only a player (ok, I could live with it - let's assume I always drive some beaten-off tank with lots of holes in in  Smile ), but AI vehicles also?

    If it is true, all this armour/shell tweaking looks like fixing a drink cooler on Titanic after the rendez-vous with an iceberg pirat

    Ok, let's assume there's no problem with 3D. So:

    1. if in reality there is no such thing as yes/no penetration (it can be partial), than how can we set the qualities of russian shells versus german starting with historical penetration data, if soviets defined penetration as 75% pieces of projectile going through the plate, and Germans - 50% (if I remeber correctly, that's not the point now). How to convert one data type into the other (rhethorical question, I'm afraid).

    1a. Moreover, going futher in this partial penetraton subject: let's assume that 1% penetration takes x kinetic energy, and 99% penetration takes x + 20% x kinetic energy. What part of this continuum (what amount of energy) shoud one set up (by tweaking shell pen value, or mass, or whatever) as a correct SF penetration value? If we assume, that penetration = death, then it's not 1% for sure, and probably not 10%, and maybe not 20%... (I mean percent of this continnum: percent of the difference between the amount of energy causing 1% pen and the one causing 99%)

    2. both sides used different steel plates as their targets (that's for sure).

    These were precisely the questions that made me thinking about taking battle reports as a reference, not the historical tests (of course, with the test values as a starting point). How do you see it?

    Donken

    Posts : 183
    Join date : 2014-07-25
    Age : 32
    Location : Sweden

    Re: Hello everyone!

    Post by Donken on Fri Dec 11, 2015 3:32 pm

    1. As i mentioned a few posts above, no need to make "partial penetration" The end result is still the same as a non penetration but with spalling instead. Its more or less the same shit but with different name if we remove the technical bit. So one way to make the shells behave almost as in real life would be to adjust the shell failure/quality parameter instead. That is how you change how often a shell penetrates the same plate. My tip is to not think about partial penetration that much. Its not something that is interesting (for the players) or game changing anyhow. Use spalling instead to "fake" the same effect.

    1a i almost answered in question 1, there is a parameter to change the quality on the shells, like parameter 19 that is the chance of the projectile to shatter upon impact (like those pesky apcr shells have a tendency to do) and parameter 44 that is the chance of failure to penetrate. By playing around with those two you can try to make like 7 out of 10 penetrates, 1 out of 10 fails and the other two shatter. But i think (not tested) that parameter 44 is how big the chance is for the explosives in the german ap shells to detonate or something like that. Also by changing parameter 8, at what angle the shells bounce you can further make the shell more "random" so you have not the exact science behind this, but you still get the correct result with like 1% chance a shell will fail. Think less scientific and dont involve such complex math. Instead think bigger, like how many shells will penetrate this plate. 2 out of 10 or 9 out of 10. No need to calculate angles vs kinetic energy vs windforce and so on.

    2 yes they did, and there you have the armor parameters Very Happy You do get different results if you choose armor quality 1 and armor strength 2000 vs armor quality 0.8 and armor strength 2100. I dont know the exact math behind this. But it is possible to simulate cast armor, face hardened armor, rolled armor and so on. One downside thou, it is not possible to change to different armor on hull or the turret, like on the soviet tanks where hull is rolled armor and turret are cast armor. What you can do is to fake the armor map instead. Like making the configs for hull armor and then change the turrets armor map instead to try to get it behave like cast armor, its rhetorical but it should work if you know how to convert cast armor thickness to rolled armor thickness Very Happy

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Fri Dec 11, 2015 3:51 pm

    Ok, after a refreshing walk in a wood with a beer - one more idea concerning this drink cooler on Titanic Smile :

    there's this issue of p_ elements for the crew (hitboxes, if i's the correct term). Someone somewhere here wrote that they're usually unrealistically big, causing death too often (apart from the holes in model issue) and suggested adding them to "notrace" section of tech_cfg file, arguing that this will cause crew member to die only after hitting the head, just like a regular soldier in SF. Don't you think that doing this for ALL tech_cfg files would minimize the unrealistic influence of 3D model holes? After all, the 3D model errors are probably in 99% cases responsible for letting the pieces of projectiles (I think of them as a kind of spray) to get inside the tank, not whole AP projectiles, aren't they? So such significant minimizing the area that the fragments must reach to cause death should fix this 3D problem in, say, 90%, am I right?
    Besides, not all models have the p_ elements for crew defined, if I'm not wrong (someone wrote about this, and I think I saw one or two such models...) - so: either all or no one of them.

    Ok, I know, this is serious tampering with this really well designed game model, but if it fixes some other, bigger problem, why not try it?

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Fri Dec 11, 2015 4:20 pm

    Donken wrote:parameter 44 is how big the chance is for the explosives in the german ap shells
    to detonate or something like that.

    I tested it with soviet BR shell - it acted like as simple probability of penetration multiplier.


    Donken wrote:Think less scientific and dont involve such complex
    math.

    Ok, so you suggest to assume that one should treat 50% and 75% penetration as the same event, because both would cause the same effect on a SF crew thanks to shattering being modelled in the game, did I get it right? If that's what you mean, ok to that. Better be too scientific and then simplify then the other way Smile


    Donken wrote:yes they did, and there you have the armor parameters Very Happy You do get
    different results if you choose armor quality 1 and armor strength 2000
    vs armor quality 0.8 and armor strength 2100. I dont know the exact math
    behind this.

    Precisely my point! I know about these coefficents, as you saw already, and I know what they're supposed to mean, but still I can see no sure way to use them to convert German reference steel plate to soviet one (if only soviets used the same in all their tests - you know, legendary soviet order Very Happy ). So what the point of calibrating soviet projectile against, say, 80 mm plate, if you don't know, if you shold apply quality 1 or 0.7 - and strength 2000 or 2100? Ok, I know that 2000 is the game default - but it doesn't mean that soviet tests were performed on a steel behaving in such a way...

    Really, it's not being picky or too theoretical - it's just trying to set possibly strong set of rules for a really herculean task. Do you imagine spending a year performing these tests and then discovering, that it's all rubbish, beacuse you didn't pinpoint some factor in the beginnig? affraid

    Donken

    Posts : 183
    Join date : 2014-07-25
    Age : 32
    Location : Sweden

    Re: Hello everyone!

    Post by Donken on Fri Dec 11, 2015 4:26 pm

    It was i that wrote that a while ago when testing the Panther G, I defenitely got better results when i removed the hitboxes for the crew. Now they are not as sensitive and also its more realistic becase as of now when a shell hits and doesnt penetrate, usually all crew members die. When i removed the hitboxes, only the crew member closest to the hit died. I also did some experimenting with the firewalls inside the tanks, i added a firewall to the panther, both between the driver/radiomans compartment and turret and the enginecompartment and turret. It also gave good results in less wierd behavior regarding the spalling problem. By just removing all the hitboxes should remove lots of headache. Problem with the 3d-models are still quite big. I will explain one of the problems.

    A picture with the firewalls:



    Take a look here:



    This is the Panther G hull with all stuff removed. And the red line is the path of a shell. This hit count as a penetration automaticly from any gun (i found out when i hit it there with a t-34) and with that hit my driver died (also very wierd), it did travel further thou and finally it hit the front of the turret with no damage.

    My thought is the physics calculation isnt fast enough for this type of hit when two opposite polys are to close (i have the same problem in unity) and therefore send it as a full penetration. I also found out after this why it got penetrated on the mantlet in the middle. The front and rear polys are to close and the shell will travel throu it (when it hits the sides in bounced of the turret front). On the mantlet it is a bigger problem because there is an open space behind it and it will directly hit the crew members, the barrel mechanism and also the ammunition. This is the same for lots and lots of tanks. But many tanks are made with no hole behind the mantlet (thank god for that) and that is ok as there is no interiors made for them.

    There is a fix for it, by seperating the front and rear polys to two seperate objects, that fixed all problems. So in this version of the Panther, it is immune to the t-34 at the front.

    So in the end, this means that it is possible to fix lots of the problems by tweaking but there is still some problems that is always gonna be there, no matter what you do =)

    Donken

    Posts : 183
    Join date : 2014-07-25
    Age : 32
    Location : Sweden

    Re: Hello everyone!

    Post by Donken on Fri Dec 11, 2015 4:37 pm

    Its better to change the value of projectile failure by testing it against the target then by putting in the numbers first. Use the real world probability for German test that 75% shells needed to penetrate the target, then setup the target and shoot at it, adjust the failure parameter to reflect so that 75% of the shells do penetrate! Its as simple as that. I did it like this with my targets. I changed the fail value to 1 and that means 100% (0 is 0% success rate and 1 is 100% succes rate) after that i adjusted the value so it penetrated exactly the thickness i wanted, That way i got a consistant value of how much it penetrates, and after that i adjusted so a certain percent would fail.

    If you use a converter you can adjust the armor quality. Its not precise but in non scientific numbers i can for example say that a 100mm face hardened plate is approximately the same as 130mm cast armor (numbers just taken from air but you get the point). If you know this exactly you can adjust the armor from it by remove 30mm armor on a turret made with cast metall. But, this means you need to use the same all over. And make your own "standard" throuout all the tanks, so 100mm rolled armor are exactly the standard parameters. and if you choose face hardened add 10mm on those places or change armor strength by 10,and for cast, remove 30 in armor strength or directly manipulate the armor map by removing the same percent.

    Stah

    Posts : 108
    Join date : 2015-12-05
    Age : 44
    Location : Poland

    Re: Hello everyone!

    Post by Stah on Fri Dec 11, 2015 4:52 pm

    Donken wrote: Use the real world probability for German test that 75% shells needed to penetrate the target, then setup the target and shoot at it, adjust the failure parameter to reflect so that 75% of the shells do penetrate! Its as simple as that.

    This is exactly how I performed my tests. I draw a table in which i counted penetrations and I aimed at desired percentage (number of penetrations for 10 shots). But then I red more about ballictcs and realised that the % values referred to the quality of projectile that get through the armour, so I thought I was terribly wrong and all my tests were rubbish. But, as you say, of course one could - for the game reality - translate one value into another directly, if there is no other way. Statistically, in battle, both values (probability od full penetration and degree of partial penetration) probably have the same meaning.

    Sponsored content

    Re: Hello everyone!

    Post by Sponsored content


      Current date/time is Sun Feb 25, 2018 6:25 pm