VW Scirocco race simulator

[nicolas]

Complete textures overhaul! Bye bye cartoon looks, hi nice and warm late summer afternoon look. To achieve this effect I also boosted the gamma while running TV3D (in the TV.Init3DFullscreen code). Is this gamma boosting a good approach or not? I'll have to take care to make my menus not too bright, because with a boosted gamma you'd need sunglasses otherwise. 

[nicolas]

I've just found that TV3D even defaults to a boosted gamma in full screen, so that should be OK. I've reduced it a bit compared to the screenshots above, as I got a bit too bright spots here and there.

Screenshots tend to be dark this way, unless made with Fraps and the like (which capture only after the backbuffer). Screenshots above were made with Fraps, so what you see is what I see minus monitor differences.

[nicolas]

Added Directsound and (Windows API) joystick support. No concrete functions yet, but the functionality is there.  atm I use the Windows API for both keyboard and joystick. Only the mouse currently goes via TV3D, but I don't use the mouse atm. I kept it through TV3D because I just might need it to click things in the 3D world or something like that (menu's?) in the future.

[nicolas]

A few questions regarding physics:

-I can't use radians as system, as that completely messes up vehicles (wheels all wrong). So I assume that Newton uses degrees internally, hence also in the output of getrotationvelocity (deg/s)? Or does it use rad/s in this output?

-I assume that I need to use physics.Simulate TV.TimeElapsed * 0.001. However, this appears like slomo, moon like simulation. I need to set physics.Simulate TV.TimeElapsed * 0.002 to get it realistic (traveled distance/time = returned linear velocity). but WHY? I would have thought I just had to transfer milliseconds into seconds. Why the factor 2? Is the timeElapsed function (or accurateTimeelapsed) somewhere making a mistake or something (due to rendering happening in two steps or something? just guessing)? I'm working completely in SI units (metric) apart from the degrees instead of radians.

[nicolas]

Nobody?

[ZaPPZion]

i noticed a lot of physics engines are "slow-motion" when using default settings, they usually look good when on double speed.

[nicolas]

It's not only "looking good", it's simply giving mathematically downright wrong results when used at normal milliseconds - seconds conversion, and mathematically correct when used at double speed conversion.

Also the radians thing is bothering me. Apparently Newton uses degrees internally, so using the radians call of TV3D is useless when you're using Newton (vehicles), as it messes them up without being able to do anything about it. That's not a real problem now, as I'll just use degrees, but given I read that TV eventually wants to switch to all radians, I'm holding my breath for Newton integration with TV...

[ZaPPZion]

well, it's not that big a deal really. Degrees to radian convertion isn't that bad really. You just multiply degrees by Pi/180 that's really really easy in my opinion, so don't worry, it'll be fine

[nicolas]

I know, I can work with them. It's not me I'm worried about, it's the integration of Newton with TV3D when TV3D would switch to radians only that worries me, since ATM radians mess Newton up in TV3D.