I'm currently working on a little project - making something like Ace Combat with Shiva3D. I think I understand the basics of arcade flight simulators - thrust, pitch, roll, and yaw. I'm using a physics-based approach by applying forces on a dynamics object, and I couldn't figure out how to make it work perfectly.

As of now I'm simulating lift by applying a force equal to the dynamic's mass times gravity upwards to the positive direction of the local Y axis and applying a yaw according to the horizontal component of the altered lift when it makes a banked turn, but I couldn't emulate the yaw completely.

Some time later I decided to look at one of the Vector Thrust's airplane objects and opened its .ini file. There are many variables concerning the mobility, including the acceleration rate at sea level, acceleration rate at the service ceiling (the service ceiling height itself is also depicted), maximum velocity at sea level and ceiling level, etc.

These variables make me wonder whether I should just change into the ordinary translation/rotation model.

So, which approach would be better? Should I stick to the physics-based approach with corrections, or should I use the normal translation/rotation model with complex variables?

EDIT: To clarify, I am aiming for a mostly arcade experience, but with plausible speed (both linear and angular) changes.


closed as primarily opinion-based by Philipp, Engineer, Gnemlock, DMGregory, Alexandre Vaillancourt Jul 14 '17 at 15:57

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ Your idea to use corrections is something I saw in AAA code a while ago (won't divulge the company name, but let's just say they've implemented a complicated wheel friction model and then applied so-called "rotation helpers" to motorcycles because these were very unstable and the player could not perform wheelies without capsizing - now their Physics is completely unrealistic and all the complex Physics is overshadowed by the "helper" action). If it's a game you're after, don't bother with forces. There are many aspects, including integration errors, solver issues which are time conusming. \$\endgroup\$ – teodron Jun 30 '17 at 10:54
  • 1
    \$\begingroup\$ BTW, there will be debate as to whether your question is opinion-based or not, so be advised. You may want to rephrase it such that it has a quantifiable solution. Perhaps emphasizing that you're aiming for a fun and not necessarily realistic game (the "arcade" word in the title suggests that) should settle it: ditch cumbersome Physics-based approaches and, at most, work with velocities instead of accelerations. \$\endgroup\$ – teodron Jun 30 '17 at 10:58
  • \$\begingroup\$ By definition an arcade flight game would be one that isn't a simulation flight game, and so wouldn't simulate the physics, I would have thought. \$\endgroup\$ – Miles Rout Jul 4 '17 at 20:58

Browse other questions tagged or ask your own question.