Posts by Jet-Pack (IPACS)

  • F104G Starfighter

    Hi Steve,

    the texture animation basically copies a rectangle section from one source texture to the target texture. The rectangles for the source and target are specified as scaled UV coordinates. For the scale we typically use 1024 or 2048, this means the ...Size and ...Position can be interpreted as pixels.

    In your example we copy a rectangle section with the lower left corner at x=0 and y = 896px (measured from the lower left corner of the source texture) and a height of 64px and a width of 96px from the source texture to the target texture position. The target texture is defined in the texture_animation which we usually call "DisplayTexture. In the "RenderList" for that target texture all the .Render draw calls are executed one after the other and each of the different displays draws either its texture animation or custom draw code like for the glass cockpits displays.

    When the source is a texture with numbers on it you can iterate through several positions and sample the source texture depending on the input. The offset between the numbers is the stride, there is a stride for the source and target. SourceStride means how far did you space the numbers on your source texture and TargetStride defines how far they should be placed on the target texture.

    The integer DotPosition is the number you have to multiply the stride with to get to the pixel coordinates of the decimal dot on the texture.

    The TargetDotOffset shifts the dot on the target texture left and right.

    TargetDecimalStride defines the gap of the decimal dot

    Regarding your texture converting issue: as far as I know you can assign the display source texture to a cube or some other blank object and it should convert just fine, even though your cube or canvas will not be used as a geometry in the tmd file. You could also temporarily assign it to some material to get it to convert.

  • About the Q400 landing

    Some planes auto flare in ground effect and this can even be simulated in aerofly.

    So the pilot may not need to pull any aft elevator to flare the aircraft but of course the pitch still increases. From my experience pilots don't know what they do compared to what the plane does on its own 8o

  • F104G Starfighter

    Hi Steve,

    Yes, these are the three types of autopilots in Aerofly. Only the advanced autopilot currently has a mach hold function.

    If you are really brave you can program an autopilot in the tmd file. You'll need three variables for the target speed, altitude and heading. Then you can use a mixlinear to create the difference of target minus is-value, scale that difference, send it to the actuators. When you also use an integral then you have your own PI controller.

    You could also use the basic autopilot for heading and altitude hold and then just create a feedback loop with the tmd to create a mach hold function. Or leave the mach hold open for later.

  • Pitts / Fixed Wing Flight Model Review

    There is no raw underlying flight model like in other sims. In Aerofly each aircraft defines its own rigidbodies (mass distribution), wings, airfoils, propellers, etc. and the physical behavior is a result of all the components working together. Our simulation technically doesn't even know it is simulating an aircraft, it's simulating individual mechanical parts that work together like a machine and some of these parts happen to create aerodynamical lift. The beauty of this is that we are never limited by a fixed set of parameters that describe an average airplane. Instead we can use the same building blocks to create new aircraft like the helicopter. And there are code parts or modules that act on the aircraft from the outside, this is where we can artificially stabilize the R22 for example.

    We named it profi mode because it is meant to be used by professionals or those that know how to fly the helicopter. The checkbox could have been labeled assistance with the checkbox state being inverted, default checked on.

  • King Air 2nd Navigation display twice applies magnetic variation.

    Quote from Overloaded

    The route displayed in the map shows about 26 degrees of magnetic variation. The outer scale is normal but the displayed route is canted wildly clockwise. This route is 003 degrees true track, 350 degrees magnetic track and the runway magnetic heading is 348 degrees. The track on the right side resembles a chart of an area with about 26 degrees of East Variation. The local variation is 13 east. I was obviously unable to check how it would look in Switzerland which has zero to 1degree East.

    The other glass cockpit planes look OK.

    (2019, iOS FS2 and Android beta FS2 version 2.5.17 AND PC FS2)

    Very good catch!

    I plan to change the EFIS of the C90 eventually anyway so this is just one more thing to check on a long list.

    As you can see the flight plan line goes across the PFD, too, this should never be the case. Also I would love to bring the EFIS to live, I want to have different modes like rose, arc, I want the menus and the other great features :) But one thing at a time.

    The double magnetic deviation should be an easy fix.

  • geoconverter

    The geoconverter doesn't interact with FSET directly...

    Reasons why it could take so long to do things:

    - target area in the tsc file was specified to large,

    - check lon-lat vs. lat-lon

    - check specified area overlaps with your downloaded tiles

  • Pitts / Fixed Wing Flight Model Review

    The "professional flight model" vs. "easy flight model" option is a myth. The flight model is always the same highly realistic and fine tuned one, we just activate an assistance when you don't select the profi mode. The assistance stabilizes the vehicle, just like some real world helicopters that actually have a flight controller. It does not alter the physics in the slightest, it still is the exact same flight model, there are no two versions, just assisted and non assisted. The option to switch flight models has been invented by users...

    Having said that, there won't be a different flight model for the pitts either. We can indeed turn up the assistance for the rudder on a taildragger. When the user doesn't have pedals we can turn on an assistance that helps to keep the pitts straight. Our goal is to always get the aircraft as realistic as possible and then bridge the gap with some other means for beginners. The R22 profi mode option is the perfect example for that.

  • F104G Starfighter

    Hi Steve and happy new year,

    the output sends a value from the physics thread to the graphics and sound threads. These have to be declared since only a minimum of values from the physics are actually observable in the graphics, most physical states just change how it behaves, not so much how it looks in a single graphics frame.

    Code
                <[output][ThisNameIsImportant][]
                <[string8][Input][ThisCanBeAny.OutputThatCouldBeNonStandardOutput]>
            >

    It would be used whenever you want to keep the original name of the value like for a lever or for a discrete knob etc.

    The output_free does the exact same but does not alter the name of the variable.

    Code
                <[output_free][ThisNameIsIrrelevant][]
                <[string8][Input][ThisNameIsTakenInstead.Output]>
            >

    I only use this if I have to, e.g. when one object has multiple outputs or non standard outputs. E.g. a jet engine would have .N1, N2, FuelFlow, etc.

    The sender is a completely different story. It sends values to the outside of the aircraft, e.g. to the external dll, to saitek panels, to external autopilot displays, to the simulated copilot and so forth.

  • Pitts / Fixed Wing Flight Model Review

    Thank you for your feedback on the Pitts flight model. This is the first feedback in a long time that I'm receiving from an actual pitts pilot (or soon to be) :)

    I find that student pilots often still have a better attention to detail. Each time I flew a new glider I immediately noticed a lot of things different to other gliders but after some time you just get used to them and it becomes normal.

    When I set up the Pitts flight model I wasn't too too happy with the result and I didn't have a real pilot to work with, like I had with the Extra. Our Extra 330 LX flight model is much better I think, it should also handle differently on the ground. In our pitts the spring for the tailwheel might be too high, this way probably done intentionally to help users takeoff and land this aircraft with all sorts of devices. Imagine taking off the real pitts but all you have is the aileron stick, no rudder. That's at least 50% if not 75% of our clients right there and they all want to takeoff and not crash every single time. (The reason why it currently takes off too easy may well be that it was intended to do so)

    My biggest issues with the setup were the lack of data, videos and feedback. When I find videos of the pitts they are usually filming an aerobatic show and I don't see any of the pilots controls. Also we don't have much data on the Pitts, e.g. I would like to know how heavy the prop actually is, how quickly the tail comes up, e.g. if you hold the brakes and go full power, at what elevator deflection does the tail start to lift up. But I also didn't have a value for where the center of gravity is located, forwards/aft as well as height above ground? Does anyone know what the wing mass is? How heavy is the upper wing or the lower wing(s)? Whats the empty weight for just the fuselage (wings detached, fuel empty)? That is the sort of data I need to create an authentic flight model....

    The nose going into the wrong direction when you push the nose... that would indicate a different propeller spin direction?

    The p-factor as such is not a physical thing. There are multiple reasons for the aircraft to turn left or right on takeoff and they often cancel each other and change with airspeed and rotation speed of the propeller. Effects that are modeled in aerofly include: propeller torque, propeller wash on tail including twist, propeller gyroscopic forces (procession), wheel friction based on normal force and a few more (I need to prepare for new years eve, I'll skip the rest)

  • F104G Starfighter

    Hi Matt,

    If you need help crossing the bridge I'll be happy to help. To me the TMD is a playground. If you have never developed an aircraft for any of the other simulators then it is probably way easier to get into it. Otherwise you first have to unlearn how you think it works. The tmd itself is actually a very clean construct, it doesn't assume anything, it just does what you tell it to do. That can be overwhelming but also very empowering. To say it with the voice of today's children: its super OP.