DCS A10 Support? solved with post #24909

[sirnoname]

I have only two planes
The update rate is about 1 second of yaw. Sometimes it increases, but only sometimes.
Speed is calulated from yaw pitch and bank, which is not really possible with this update rates.
For now I do not talk about speed values because this is a second step.
Which plane you suggest in DCS world?
Are there no real flight matrix datas or plane position datas in the world?

[laurivs]

http://www.digitalcombatsimulator.com/e ... ua-export/

There's more info on variables and commands but I think there's more.
Check this out.
http://forums.eagle.ru/showthread.php?t=78529w.

This is how software Tacview gets its data from DCS:
http://lomac.strasoftware.com/download/ ... ortDCS.lua

The problem is that export.lua should need to be universal and work with all aircfafts. Some commands don't work with all planes. (See my recently posted topic under DCS Series)

Update rate of once a second Wow, I have to install Steam and DCS on my other computer and see if I get the results you are getting. I'll get to do that in about 10 hours from *

[laurivs]

Hello,

I'm back home and I'm downloading DCS to my laptop, 40 min remaining. I tested again DCS for the yaw value on my main computer and it does give smooth data with the same two free aircraft that you're using (TF51D and SU25, right?) . Unfortunately the yaw data is completely and utterly wrong and useless in the following case: when you fly upside down, the yaw value gives you the opposite indications than what you should be getting.

Sirnoname, let's forget YAW problems for a moment since getting a functioning yaw acceleration value will be impossible with given data. There is no source data for it at the moment! Helicopter pilots might not stumble on this problem that radically, as choppers quite rarely fly inverted. They would notice the problem when doing vertical turns however.. As nothing would happen with the motion system when it should give you yaw movements.

How a proper yaw acceleration value works, is, when you kick the right pedal (rudder right), you get positive readings, and negative if you give left rudder. This doesn't happen now unless you fly pretty much level. If you bank the aircraft 90 degrees left or right and use the rudder, this should give you some indications but it doesn't. Why? Because this "yaw" value that you see, is more of an indication of Heading, which is fixed to the world, and is not a "local" value that we need for our purposes. These things have no meaning when you drive around a race track in a car, but when flying, this is essential. You can't get responses in the exactly opposite directions all of a sudden.

Could you btw post your export.lua here unless of course if It's exactly the same that I posted here?

Would you have the energy to post a short video how you test the values? I would like to see how the values update, this would be cool for reference. I recommend Open Broadcast software

https://obsproject.com/

Is the bank data always smooth? If we could even get the bank acceleration working and nothing else, this would really be a success in my books!

A few questions and some recommendations that might help:
1)What is your FPS when flying in DCS? (L CTRL + Pause)
2)In "Special" settings -> TF51D, Autorudder -> Uncheck
3)In "Special" settings -> TF51D, TakeOff assistance = 0
4)In Game Settings, Difficulties: Game Flight Mode -> Uncheck
5)Generally try to achieve the settings that most simulator enthusiasts use, I believe that is towards the most realistic options.

I suggest that if you would like to try the "Flaming Cliffs 3" pack? This includes F-15C, A-10A, Su-27, Su-33, MiG-29A, MiG-29S and Su-25. But I believe that the free version aircrafts should work great. It's needless to spend money to payware aircraft if the same problem persists. Let's see in 40 minutes what my laptop says. If I get everything working normally, including the useless yaw value, we could narrow the problem to your system. So, let's not waste any money just yet.

-L

[laurivs]

Okay,

I installed DCS from Steam. Played with my shitty laptop, everything works exactly the same as on my gaming computer, except that fps is poor. I used TF51D and SU25. Instant action missions, free flight stuff.
Even with 30 fps, the motion data is constant and smooth (expect bank accl,pitch accl,yaw accl. of course!) when playing on laptop that gets only 1/5 of the fps compared to a more powerful computer.

So my question is, could you try installing everything on another computer? There is clearly something wrong with your installation or hardware or something

Or even better, could you send me a modified version of the lockon plug that you're working on so that I can test it?

-L

[sirnoname]

Here you have



I use windowed mode and go SU25 ground attack. On User6 is the timediff from last to new time index (seconds).
This value is always different if I start the game. Sometimes 0.6, sometimes 0.2.
Try to roll and pull stick to cahnge yaw. Look at smooth time value to compare.
Change lua script.

[laurivs]

Working, home tomorrow, in 24 hours. I'll test as soon as I can.

[laurivs]

Hi!

Now tested with gaming rig and laptop.

Gaming PC user 6 = 0.01 always
Laptop user6 = 0.03 almost always.

I'm uploading a short video to youtube, ready in 20 minutes. (computer used in video is the gaming one)
User6 value changes if fps drops and also relative to time scale in game. (LCTRL + Z, LALT + Z, L Shift + Z )

(You should not decrease the accuracy of the values given by simulator. This has happened for example to yaw value * Before we had about 628000 possible values, now it is reduced to 360 degrees in 0.3 decimal increments = about 1000 values. It is way too coarse.
I hope that what you're doing is just for the test plugin not how the final version is going to be.)

The YAW update rate is as good as before, but the resolution has decreased significantly.

-L

[laurivs]

Here's what it looks like on my system.

https://www.youtube.com/watch?v=dyIQ-Cs ... e=youtu.be

[sirnoname]

It is very slow, don't you see that the orientation values do jump in 0,4 only all ~200ms or more? This is not possible if the 0.01 seconds are the update rate.
They should change far below the x.xx display instead of after the dot. It must ba a smooth increase in 0.01. If you have at 0ms with 3.00 degree and suddenly at 100ms with 3,40, you have a speed spike that is extreme big. Then you have all 0.01s zero speed, maybe 10 times until the next shock.


ps. the multiplicator is done inside the plugin and will be in the final big. However you cannot remove 0 to 360 if the plugin must know where the 0 to 360 barrier is.

[laurivs]

I agree with you, but I'm afraid that you don't see, that this problem wasn't there before. The problem is in the inaccuracy of conversion from radians to degrees and how you display the result, how many decimals and so on. Just try it yourself with the original export.lua and the earlier LockOnImport.exe. You will notice, that the update rate and accuracy is excellent.

Yaw example:
57.30
57.87
58.44
59.01
These are the smallest increments in yaw with your latest export.lua and LockOnImport.exe.

The difference is always exactly 0,57. Why? Here's why:
radians * factor = degrees.
6.28 * 57,3 =360
Factor = 57,324840764331210191082802547771.
This is a clear evidence, that the changing from radians to degrees is inaccurate, and causes the big gaps in values with new export.lua and LockOnImoprt.exe.

So, let's do a more accurate calculation somehow; I understand that you need 0-360 for getting rid of the spike problem when you go from 360 to 0. But why can't we keep the radians and adjust the math for them instead of 360 degrees? Shouldn't be too hard.

-L