Some problematic observations with 2.483

[Volcano]

Thanks for the additional info.

In regards to the sicky A - D keys on your CH setup, have you tried directly mapping the keys to the button on the CH stick in SB Pro (not using a key binding file in the CH manager)? Some who uses CH sticks suggested that as their method of setting up the keys -- maybe that will work. Other than that, we are talking about an external program that establishing key bindings into a file, then loads that key binding file into the memory chips within the CH hardware (joystick and throttle). It could be any number of things wrong with it at any point in the system. One of the key bindings getting interpreted by DX incorrectly or something similar. If you can cut out the middle man (the key mapper) then the less will go wrong between the controls and the sim itself. Have you also tried using a negative type function in the CH mapper software? For example, if a key is "sticking" can't you use a toggle or hold parameter as an experiment to see if it corrects the sticking? I vaguely remember doing stuff like that when I used CH sticks back in the days of DOS.

In regards to the Leo2 issues, you are right, it could very well be an offline only issue since others have not noticed it (since you said you play exclusively offline). That is good to know. One question is, if you take a stock scenario like Deliberate Assault 02 (DA02), what would be the step by step process of seeing this behavior? Can you give an idea on what unit to go to and how to wait? Any sort of guidance on what can be done to reproduce the behavior? I don't mean "start the scenario up and jump around to any tank and play and then you will see it" type instructions.

So it seems so far that the process to reproduce what you are seeing is:

1) Use the DA02 scenario

2) change the tanks to have unlimited ammo

3) play the scenario offline

...are any of those points incorrect? Please state any further instructions as well and we will check it out. Besides that, I can't say it enough -- the more specific instructions the better. The lengthy descriptions of the symptoms are not as important as the description of the procedure as to what causes it, there are a lot of variables involved. One thing is, if you are playing very large scenarios like DA02, there could indeed be a problem where the CPU is getting overloaded. Given your system specs though, I would think that this would be unlikely.

[Skybird03]

I use the Manager software for CH gear. If it does not fail to properly function in transmitting the mapped commands in all other sims I use it in, I fail to see why it should misfunction with just these two commands, A and D, in just this sim, with the command once mapped to buttons on the throttle, and once to the stick.

The manager software is also needed to attach wanted non-linear reaction curves, or - in this case - dead zones to the joystick centre. I use the command manager in all sims I use the hardware in. I do a lot of civilian and combat flying, too. I would expect erratically working joystick axis producing even more noticable effects there. But that is not the case, I also do not have stickied buttons. Note that I said the GPS is also shifting slowly to the bottom even if only the mouse is attached (MS Wireless Intellimouse Explorer 2.0). In SBP, if it is the hardware, I would also expect that what command gets stickied does not depend on the software command (A and D in this case), but on the hardware button it is mapped to. But in fact it is the other way around: it is always the two very same software commands, no matter which hardware slot they are being mapped to. The gun tube issues and the shifting GPS, again, is taking place independently from whether or not the stick/throttle is connected.

There is no reliable way to reproduce the issues I mentioned in this thread at will. That really messes it up, doesn't it. Scenario selection and scenario situation are unimportant as to my best experience so far. I thought until yesterday that unlimited ammo supply maybe had something to do with it, but in fact saw the gun tube locked in extreme up or down position, the tank not shooting while being shot at. But it is true, i run all medium and big size scenario with unlimited ammo, only small ones with limited ammo. I do so since early this year.

I mostly edit abrams out of scnearios and replace them by Leopards (and replace Bradleys with CV-9040s), I have not seen the Abrams behaving like I described for the Leopard-2, but on the few occasions I had abrams in the mission, like this weekend, the Abrams sight gives troubles as well, it does not slowly shift to the bottom, but the turret, if being traversed to any side by mouse or joystick, tends to not only stop, but to jump 0.5-2 cm back, the jump being the bigger the faster the turret was moved. I have occasionalyl noted that on my own rig as well, and in less intensity I have bthat effect with the Leopards, too, occasionally. Before thinking of the joystick again, keep in mind that this effect was ntoicable with two different joysticks, both of which had been mapped with reliable dead zones in the centre of the joystick axis (the effects also can be seen without programmed dead zones, just that then you additonally get noise effects from the potentiometers). The SBP-centre-joystick command does nothing regarding this or the before mentioned issues, it only helps to adjust misaligned joysticks. I do not mean the sight jump you get in the M1 when having lased a target that indicates that lead is being added now, and which stops when pressing the palm button. This GPS jump does not seem to happen with mouse, and in Leopards it also occasionally happens, but not as big in jumps as in the Abrams, on my screen with a 55 cm diagonal, in the Leopard it makes around 0.5 cm at maximum, in the Abrams, as said, it can be jumps of up to 2 cm. The shifting sight that slowly drifts to the bottom, indicating the whole gun tube sinking down, happens both with mouse or joystick. Like almost all problems here, it sometimes happens, and sometimes does not.

again, Volcano, I fully understand your desire to have a reliable method to reproduce these issues in a causal way, and that this probably even is needed to adress bugs. But if there is such a way to achieve that, I have not found out about it. I find the situation as confusing as you do, and hardware getting bogged down and CPU or memory overload (but with 2 GB, even 3 GB under XP?) at least the latter is hard to imagine. Also, one of the systems yesterday was a much stronger CPU than mine.

As it is, you play, and if you are lucky and look at the right place/unit at the right time, you see it happen - or not. Not all Leopards misbehave simultaneously, the one may do, and when you jump to the number two in theb platoon just ten meters beside it, you see it functioning fine, look at number one, and see it doing strage things with the gun. The chance for an "observation" increases the longer you play, maybe also the more units and places you visit over a mission.

That it is some kind of hardware overload, since I tend to play bigger scenarios mostly, is just a thesis out of the blue that I asked for in search of finding an answer to what causes this behavior. I now think that if this is a valid trace, it more leads towards scenario sizes in general - independently from the hardware it is run on.

This is really all I can tell on these issues, I only could repeat myself from here on, and actually already have.

P.S. I must correct one mistake I reported, I wrote "quadcore 2.8 GHz" yesterday, which is wrong. It was a duocore 2.8 GHz.

[Ssnake]

I mostly edit abrams out of scnearios and replace them by Leopards (and replace Bradleys with CV-9040s), ... but on the few occasions I had abrams in the mission, like this weekend, the Abrams sight gives troubles as well, it does not slowly shift to the bottom, but the turret, if being traversed to any side by mouse or joystick, tends to not only stop, but to jump 0.5-2 cm back, the jump being the bigger the faster the turret was moved.

...which happens to be a feature of the fire control system - well, a consequence of the single axis stabilization of the sight and the automatic lead application. If you let go of the turret control you also disable the lead calculation, hence the turret (and with it the sight) goes back a little in the opposite direction.

[Ssnake]

I use the Manager software for CH gear. If it does not fail to properly function in transmitting the mapped commands in all other sims I use it in, I fail to see why it should misfunction with just these two commands...

But couldn't you just give it a try instead of giving reasons why this can't be?

Then we will all KNOW that it doesn't contribute to the problem, instead of taking your word for it. It's an essential step in the analysis.

[Skybird03]

...which happens to be a feature of the fire control system - well, a consequence of the single axis stabilization of the sight and the automatic lead application. If you let go of the turret control you also disable the lead calculation, hence the turret (and with it the sight) goes back a little in the opposite direction.

As said, I do not mean the jump the sight does when you lase the target - not before the Abrams adds lead, or do I get that wrong?

1. You qickly traverse the turret without lasing (hold mouse/joystick to an side), and then stop the turret movement control (centre control), and the sight jumps back, as if swinging back - 2. then you hit laser, and again the sight jumps a second time.

3. Or you swing around the turret, aim at a moving target, lase it with the turret on the move, and the sight jumps.

Only the second and third jump I understand as regular behavior of the Abrams, which already has been like that since 2006. The first jump, after just stopping to move the turret without lasing, I do not understand. How can the turret calculate or compensate for lead, if the laser did not mark anything to which lead should be added?

This has been shown with two different joysticks, my CH, and his Sidewinder.

[Ssnake]

Can you please contact me by email?

I wonder how far away you're living. Maybe I should stop by with a video camera to document the whole stuff.

[Volcano]

As said, I do not mean the jump the sight does when you lase the target - not before the Abrams adds lead, or do I get that wrong?

1. You qickly traverse the turret without lasing (hold mouse/joystick to an side), and then stop the turret movement control (centre control), and the sight jumps back, as if swinging back - 2. then you hit laser, and again the sight jumps a second time.

3. Or you swing around the turret, aim at a moving target, lase it with the turret on the move, and the sight jumps.

Only the second and third jump I understand as regular behavior of the Abrams, which already has been like that since 2006. The first jump, after just stopping to move the turret without lasing, I do not understand. How can the turret calculate or compensate for lead, if the laser did not mark anything to which lead should be added?

This has been shown with two different joysticks, my CH, and his Sidewinder.

Yes, all of these are correct behaviors of the M1, M1A1 FCS. The M1, M1A1 FCS has come a long way and was very problematic to get "just right" over the past 10 years or so, and I aim to keep it so no modifications are made to it so that we do not enter the cycle of breaking things with it again. The best way to describe the real (and SB) M1 and M1A1's FCS is "unique" and "quirky", to put it in terms that are polite and respectful to the real tank. If someone played in the Leo2 all the time then it would definitely seem like an alien experience, or like something was broken, but I assure you that it is as the real FCS behaves.

I would say that we (I) play the M1 tank every week and I haven't noticed anything abnormal with the operation of the FCS.

[Froggy]

Skybird, you might use Fraps to record the behaviour you are describing, because I have played DA2 the way you mod it (leo instead of M1 and unlimited ammo) and i have not seen gun at max or min level. I have a scenario with 45 leo, and no problem too, solo or multi.

I am not part of SB team, but with 507rcc guys we have played many scenario to reproduce your problem, and we have (still) not seen it.