[ODE] how to count angular velocity...

[Krystian Ligenza]

yes, it's esier a lot. I have done as Erwin said and everything work
greate. Thanks for a reply, I must start to use more quaternions instead
of matrix's.

Krystian

> I see.  If you convert to quaternions, it is easier to obtain a rotational
> difference between two arbitrary rotations, and turn this into a rate over
> time.
>
> -----Original Message-----
> From: Krystian Ligenza [mailto:s2562 at pjwstk.edu.pl]
> Sent: Wednesday, November 22, 2006 1:50 PM
> To: Tedd Streibel
> Cc: ODE at q12.org
> Subject: RE: [ODE] how to count angular velocity...
>
>
> i forgot to add, that I'm setting my symulation to initial configuration
> from motion capture data. So I need to compute by my self angular
> velocity.
>
> Krystian
>
>> You should be able to get the angular velocity from the object using
>> dBodyGetAngularVel(ID).  It returns a 3-float vector that represents an
>> axis
>> (the x, y, z from the vector) and a rate of rotation (the magnitude of
>> the
>> vector, or mag = sqrt(x*x + y*y + z*z)).  A magnitude of zero means no
>> rotation.
>>
>> -----Original Message-----
>> From: ode-bounces at q12.org [mailto:ode-bounces at q12.org]On Behalf Of
>> Krystian Ligenza
>> Sent: Wednesday, November 22, 2006 12:27 PM
>> To: ode at q12.org
>> Subject: [ODE] how to count angular velocity...
>>
>>
>> Hi,
>>
>> is't probably very easy question, but I have gave up :(
>> I have got dBody in 2 time steps. I know it's rotation matrix in both
>> time
>> step (t and t+1), and need to count body angular velocity in t+1. I have
>> tried do it in this way:
>>
>> R(t+1) = R(t)*R(transformation), so
>>
>> R(transformation) = R(t)-1 * R(t+1), after this I have converted
>> R(trasnformation) to euler rotation vector, and used it like this:
>>
>> dBodySetAngularVel(body, vEx/timeStep, vEy/timeStep, vEz/timeStep);
>>
>> Legend:
>> R(t) -> rotation matrix in first time step
>> R(t)-1 -> inverse matrix R(t)
>> R(t+1) -> rotation matrix in second time step
>> vEx/vEy/vEz -> euler rotation x/y/z
>>
>> but it don't behave good :( Can somebody help me?
>>
>> Thanks
>> Krystian
>>
>>
>> _______________________________________________
>> ODE mailing list
>> ODE at q12.org
>> http://q12.org/mailman/listinfo/ode
>>
>>
>
>
>
>