Jump to content


  • Content Count

  • Joined

  • Last visited

About Captain_Colossus

  • Rank
    Senior Member

Recent Profile Visitors

5,842 profile views
  1. i suppose i could do more, but it would take time and it wasn't my original intention to match all the vehicles. however, as you see in the first screenshot, the stock color schemes already do not match the standard colors, either. not to be pedantic, but even if i wanted to make the same color scheme, it will likely only be approximately the same- similar, but something might seem a little off. if you try to use extract color tool, you can choose a single pixel or a larger sample, but it never quite turns out the same, since color is subjective and depends on many factors to give you the impression of specific color. if you look at the second screenshot, every one of these marked circles will give a different hue of green, some lighter than others, some closer to each other than others, but none are exactly the same as each one is exposed to the light source differently, each one has different dirt or weathering or what have you. across a large sample, you get an average of all the colors, a smaller sample shows you how different even nearby pixels are from one another, some much so it's like comparing black with white
  2. it is within the range of base green that you will find depending on the light source, the time of day, meterological conditions, amount of dirt or water stains or weathering effects. there is no such thing in this universe as a tone of color that exists in a vacuum (in fact we know from physics that color doesn't exist "out there" at all- like sound does not exist in space, it is the brain's interpretation of what it estimates is "out there" based on signals received from the optic nerve. depending on the wavelength of light, say time of day or other atmospheric conditions, the same tone of green will look different to the observer, none are "correct"). the stock paint scheme in steel beasts isn't even 'correct' in the sense that it doesn't match up to reference photos. compare the two images below: both the green and tan contain different values. i chose areas of the reference photo that are approximately most similar, appear to have the least amount of dirt or dust or changes in light reflection or what have you. using a one point sample, the bullseye on the turret is 818375, in steel beasts this approximate same location is 3932c29. the tan color in the reference photo is e8e2d6, while in steel beasts it is 817b66. in both the steel beasts model and the reference photo, these values can differ wildly depending on the part of the tank and the amount of sun reflection or dirt, but within general parameters for these tones. then the next photos show obviously completely different tones- again based on the factors i mention above, and some of which have different paint applied (either at the factory or at the unit level perhaps.
  3. i figured out the workaround- duplicate spawned units in the same location- for example., primary unit has 80 percent chance of spawn with active protection equipped, secondary unit has 20 percent chance of active protection removed. not too bad.
  4. it certainly is modeled that way- however, i request something which permits a user prescribed chance of not function at all through the entire scenario rather than the system having the rare miss (but still functioning)
  5. that's not the point- i am requesting a random chance of these systems not functioning, meaning they still have a random chance to function (rather than non-functioning at all as empty).
  6. i haven't upgraded from 4.161 yet, so if it isn't present in a newer update: add active protection systems and russian afganit to the list of components which can be damaged in the scenario editor (simulating a chance of system failure and/or system has previously spent its bank of countermeasures ).
  7. that's the problem i believe, unless the computer is allowed to cheat, that is, able to sort out an IR beam sourced from a T-72M and a T-80U (or T-62, T-55 and so on), even if the vehicle isn't relatively near to identify or isn't out in the open. then imagine several vehicles throwing beams and the computer bamboozled with information overload if it sees each beam as a projection of the targets everywhere. computer units have some nasty advantages over a human, that is, they can instantly sort out live, dead, civilian thermal signatures when there are lots of killed vehicles or civilian noise and a few live enemy vehicles mixed among them- a player may have difficulty discerning which one is a valid or a priority target in thermal images from all the clutter, the computer on the other hand discerns them immediately- you might really notice the difference when the TC is calling an identified "tank" out of a sea of TIS signatures when the user cannot identify which unit the computer TC refers to from all of the noise- then if you switch views to say, the F8 external view, your computer gunner will immediately sync up with the TC to the valid target. on the other hand, humans have a certain intuition about when to realize when to ignore or prioritize threats that computers do not have based on contextual information, so a computer can conceivably be overwhelmed and scrambled by human players either deliberately or unintentionally using their IR searchlights to project a target everywhere the computer units can see and distract them,, since they essentially see the length of each beam as the projection of a target: and then the computer takes shortcuts to be able to source back to the beam and know immediately the threat level of the target, (even if there is no direct line of sight to the target) the same way the computer as it is now can properly identify any thermal blob and determine target type immediately
  8. one of the issues is programming the computer when to ignore or not to get fixed when the computer detects distant or obscured sources of the beams and sort them out from other threats- you can imagine exploits used to dazzle computer units, that is, detach a vehicle and place it turretl down, have it flip on and off its IR projector and illuminate the sky or environment wherever it can in order to distract the attention of computer units. you would have to program the computer in such a way to prioritize attention to these beams, but at the same time know when to realize that the source can or should be put lower on the priority queue. quickly users may get frustrate by the AI being confused or what have you when beams are projected all over the place and the computer either gets target fixated or is confused by multiple beams projected from multiple sources, some which may be obscured by cover or concealment. likewise, the computer would need instructions to know when it's a good idea or not to illuminate as they lack context- intuition about that this isn't of course to say no one wants more features, but it doesn't seem like something you would accomplish with no less than fundamental changes to everything
  9. question1) : the answer given from the official channel is no, but i think this is premature, because the answer is yes. it's not explicitly baked into the model, that is, for example, the model doesn't automatically suffer performance issues from wear and tear, however, you can simulate these conditions with regard to question 2), the powerful battle and scenario editor allows users to assign component damages and failures, even shots that drift in a specified way- these reliability issues can be attached to random variables and/or explicit conditions for them to occur, but the point of it is that they must be scripted by the user rather than something that occurs automatically. in this regard, some answers to your questions are present but implied through the user defined scripts in the scenario editor rather than the computer keeping track and assigning them; much the same way your question could be answered with regards to unit behavior- the computer does not automatically withdraw units or cause them to surrender (simulating failing a morale check for example), however, the scenario designer can define when these situations occur through the editor; explicitly the computer doesn't do it, implicitly a well designed scenario is capable of some of these behaviors
  10. i wasn't aware of this, but it makes sense- i already rationalized it in a similar way, because infantry would expose themselves in windows and would be too vulnerable if they were any easier to detect or pick off than they already are.
  11. infantry are too easy to kill in buildings because they kneel in windows and stay exposed- even if being fired upon. they don't appear to go prone, the rooms in buildings are bright and aren't shaded, the model isn't sophisticated enough that they duck or spot or fire from concealment or cover, in some ways, this probably balances out the perceived difficulty for the AI to kill them, because a human player in a vehicle can make short work of them. in fact, i tend more to keep them out of buildings if possible, they can find better concealment in broken terrain than in buildings or on rooftops, which fixes their positions and exposes them
  12. back in 1999 or 2000 or so there was interest among the player base for improved graphics, among other things- part of the overall transition to an update. it's not that users would turn down better graphics, it's that the core experience was still good enough at the time to continue playing steel beasts- but there certainly was enthusiasm for improvements. steel beasts was clearly a good basis to keep going, and we saw the value in it with a future upgrade in mind. i think it's somewhat of a myth that steel beasts' graphics were really that awful- given the generations of land focused simulations, armor or otherwise, there weren't many at all to speak of in the late 1990s, and of those that existed, say, panzer elite, panzer commander, m1 tank platoon2, the graphics on those games weren't that stellar either, steel beasts in most cases was either more or less on par with those games (rendering tracers, fire effects, infantry as flat sprites), or it even surpassed them (steel beasts rendered forests consisting of individual trees, even though they were like carboard pictures, and the infantry representation was still better than those other games). i think steel beasts really wasn't as obsolete as it claimed in grahics- panzer commander looked quite artificial and lifeless, had no infantry, and had generally featureless, flat terrain. M1TP2 was better in this regard, but certainly quite flat and barren. pictures on the box gave the impression of much more terrain and urban detail in the game than were actually present. panzer elite looked like a bad impressionist painting, the vehicle models did not look very convincing either. steel beasts had better terrain representation with softer, undulating hills, depressions and gullies. i think out of all the games the putting off was the lower resolution of steel beasts and the lower surface details on the models- but generally it was still on par or better than the competition. computers of that era probably would not have been able to handle much better representation of vast environment details and many actors at the same time, that's why you tended not to see it back then. it was a couple years later when other games started showing more detailed graphics- operation flashpoint, battlefield, and so on- by then steel beasts' graphics were obviously falling behind (as the previously mentioned games also did), hardware moved on, the more consistent player base looked forward to updated graphics. this is also why i don't really understand this move over the last couple years where you see some users looking back and playing steel beasts version one again, and actually in some cases wanting support or updates to it because they prefer the experience of steel beasts 1
  13. i was not aware of sub-events and overpressure as being descriptive anywhere. i am still running 4.161, if it was present then, i have never seen it, i'll look again the next time i can run steel beasts. i am pretty confident by this point that a sufficiently large. non lethal HEAT impact against the front profile of a tank such as the M1 will at minimum always produce the radio and stabilization damage; overpressure or not, you could set your watch to that. more often than not, those are the only two damages occurring unless the hit is directed on the target roof or doesn't manage to hit something vital like the turret ring. to understand the extent of it, i ran a few artillery tests as well in those tests several direct hits to the turret roof did nothing of the sort, and a near hit produced damage to the idler only- by near miss by what look like within a meter. this doesn't mean it will never happen that the radio and stabilization will not be damaged by artillery direct or near hits, but from what i have seen it is obviously much less predictable. as an aside it just happens to be one of these cases frequent use of the software- any game or simulation software starts to show certain predictable behaviors- right or wrong, but you cannot unsee it once that happens. in older DOS based flight simulators players could probably predict the direction or course computer pilots may bank or dodge and this kind of thing as they begin to intuitively see the code behavior running its routines. whenever i see a missile coming in, i know that i am going to lose the radio and stabilization, i am already thinking one move ahead ready to gunnery switch modes to prepare for the loss of stabilization and return fire, i am already thinking what my next move is with radio loss if i need to issue orders to other units or call for artillery (say jumping to another unit to issue orders)- it is so consistent in the SB world that it compares to other rarer but certain events like always running over troops results in a kill without variation (as well as it probably should i reckon), there are some scenarios that are not variable, and this appears to be one of them. i don't have any calculations which would contradict it, if esim believes it is the best possible compromise, that's fine with me- but i do mention it does stick out in a model which largely is more variable than that elsewhere; i hadn't noticed it previously but once i figured it out, it cannot be unseen.
  14. the type 99A seems to be somewhere in between two compromises: could approximate with T-90 (though non playable, type 99A is heavier and probably better protected on the front), leopard 2A4 (there already is a PLA skin included with steel beasts, cannot gun launch missiles, lacks the commander's heavy machine gun) leopard 2A5 or strv 122 (more visually similar turret profile, cannot gun fire missiles) M1A1 (cannot gun fire an ATGM, lacks commander's independent TIS) furthermore, the type 99 appears to have an active protective system or the potential to add one, since there apparently is an IR dazzler similar to the T-90, could theoretically be swapped out for a hard kill active protective system (or for those vehicles in steel beasts which can attach them). given a choice I would go with T-90 as an AI opponent, leopard 2 or strv 122 for a player crewed tank.
  • Create New...