Guest BratZ Posted June 9, 2003 Anyone have a formula for my speedometer? Notice how its not equal all around? Share this post Link to post Share on other sites
Guest BratZ Posted June 12, 2003 Nobody? If I ever figure it out,I will post the formula here Share this post Link to post Share on other sites
Bullz_eye_on_my_back 0 Posted June 13, 2003 I don't know anything about the config files or anything for the speedometers or anything...but if you can use a formula to make the pin rotate around I'm sure a formula could easily be made I'm guessing the speed is represented as the resultant of the x and y velocities devided by 60 min? Also if you can use formulas what would be the output? degrees/radians of rotation? If you can use formulas i could do some tests and fit a curve for ya...would only take a couple min. looks like it would be an exponential curve of some sort Share this post Link to post Share on other sites
Bullz_eye_on_my_back 0 Posted June 13, 2003 oops I'm crazy there is a command for getting the speed... :::hits head:::: whew where is my thinker? but yep if you can manipulate the rotation by feeding it degrees then i'll try and help ya out! Share this post Link to post Share on other sites
Guest BratZ Posted June 13, 2003 Ya its some equation,I'm gonna be working on it.Its on my list of ToDo's. Was hoping someone knew one off their head, rather than me getting braincramps trying to figure it out,lol Thanx for trying Share this post Link to post Share on other sites
Bullz_eye_on_my_back 0 Posted June 13, 2003 lol...no problem... I'm just going to take the picture and figure out how many radians clockwise from zero to each number and then try fitting a log curve to it.... (radians vs. speed) might not be perfect ill try! Share this post Link to post Share on other sites
Guest BratZ Posted June 13, 2003 Hmm your right,I could script that and make it totally accurate.Speed vs. animationphase But the formula for the oem speedometer would be in degrees and I'd rather try to figure that out Heres one from Biscamel that converts to knots,I dunno if its accurate: class IndicatorSpeed { selection="mph"; axis="osa_mph"; angle=-360; min="50 / 2"; max="170 / 2"; }; But I notice that the BisCamel speedo also isnt symetrical 110 knots dead bottom,180 knots top Again tho, whats saying that its tweaked to accuracy, I am probably being too fussy here, but its the only thing left and then I can eleiminate the onscreen graphic for it Its a good finishing touch I think Already have working gas gauge and even rolling wheels This plane will be the kitty cat of ww2 fighters,lol Share this post Link to post Share on other sites
Bullz_eye_on_my_back 0 Posted June 13, 2003 haha...sounds pretty sweet hmmm....can you call commands within a config... like getting the speed? hmmm.....ive been sitting here trying to make an equation ...it will fit part of it but then will be way off near the 600 speed hmmm..i guess there is no way to do a script for it... it would be really easy to do if you could break it down into 3 separate equations... but the way the configering won't allow it. Share this post Link to post Share on other sites
dm 9 Posted June 13, 2003 The dial is probably marked up logarithmically, that way, at the faster speeds the needle moves less and less to indicate a greater increase in speed. How the Game engine handles it tho, I do not know... Share this post Link to post Share on other sites
bn880 5 Posted June 13, 2003 Well yes the proper way to handle this is to create a polynomial through some curve fitting (speed vs Radians or preferrably degrees), however if you don't have experience in this it may be a little difficult to grasp. (Newtons divided differences or cubic splines are a good example http://www.ittc.ku.edu/~rvc/documents/781/report_3.pdf) The other solution is less accurate but quite simple: We will divide the scale into 4 separate ranges [*] 0 - 50 knots = 0 - 10 degress = 10deg/50knots = 0.2 [*] 50 - 250 knots = 10 - 222degrees = 212deg/200knots = 1.06 deg/knot [*] 250 - 300 knots = 222 - 247 degrees = 25deg/50knots = 0.5 deg/knot [*] 300 - 600 knots = 247 - 337 degrees = 90deg/300knots = 0.3 deg/knot Now through a set of if statements you can get a nice output if speed <=50 then angle = 0.2 * speed else if speed <= 250 then angle = 10 + (1.06 * (speed-50)) else if speed <= 300 then angle = 222 + (0.5 * (speed-250)) else angle = 247 + (0.3 * (speed - 300)) end if EDIT: actually the first 50knots needs to be accounted for also, I'll fix that in a few EDIT2: done Share this post Link to post Share on other sites
bn880 5 Posted June 13, 2003 Okay added another scale, that makes 4 different linear equations for the speedo, and it should display pretty nicely. You may need to use a constant multiplier(like Km to Knots) on the speed before you pass it into any of the equations. Share this post Link to post Share on other sites
BraTTy 0 Posted June 14, 2003 Wow thanks but oops. Where my speedo pic go? lol I guess I threw it offtrack showing the Biscamel speedo,but thats the only speedometer offhand I found that even had a formula in it. And its in knots An american F4U speedometer i believe should be mph,but I will settle for the metric kmh,but the bis oem speedo shows mph So really just looking for a formula for the difference in the speedometer pie I can work off what Bn880 has laid out as its going to take calibration anyways Share this post Link to post Share on other sites
Bullz_eye_on_my_back 0 Posted June 14, 2003 oops i was fitting the curve to the one shown on the post. I found all of the degrees of rotation to each number from zero. It turns out to be linear in 3 sections 0 to 50 50 to 250 and 250 to 600 i was trying two different ways. using excel to do a trend line..but it doesn't work then trying to fit a logrithmic equation to it.... Share this post Link to post Share on other sites
