Last post of the Wiimote plugin main thread:


I wonder about the poor using (at least it seems like that, using the search function) of that plugin, because having a Wiimote support in 3dGS is pretty cool (since I own a Wiimote ;))...
Why it's cool? Because it gives a great possibility to make casual games and create in a relative simple way something new.

Here's now a mix of questions:

• What about the Wiimote+ support? Will it come?
• Why does the pan angle not work (stays zero)?
• Why do the standard (->wiimote not moving) accelerator values change with its angle (values)?
  -> this means: when the Wiimote lies on the table X, Y, Z should be always zero
• Where's the ir_dot.tga file you need for WII_DEBUG ?
• Any news related to a Wii 3d Headtracking package?

Belongs less to here:

• Are there any 3dGS projects using Wiimote? Anything to see or play or experiment with?? (Only game I found is TacaTaca at AUM which planned to integrate Wiimote controle - the only place where I got information about that project...)
• Are there any tutorials and hints? Beyond the single maginal manual page and the (required) look inside ackwii.h/c?

Greetings,
Clemens

More confusing with coordinates:

• Why is the tilt angle suddenly changing for 180° when the roll angle is getting higher (/lower) 90° (/-90°)
• The same reversed: it turns 180° around at roll angle when tilt angle is higher (/lower) than 90° (/-90°)
  (you can try the example in the manual using vec_set(your.pan,wiimote.angle1): it doesn't work correctly -> "mirror like" angle switches)

I'm afraid answering those questions is beyond me, as I'm not this experienced with the Wiimote. What you get are not the raw angles and accelerations, but preprocessed values. I don't know what the preprocessor in the Wiimote precisely does, but maybe there are forums where those questions are discussed. Mostly you'll have to experiment yourself. A Wiimote+ will work I suppose.

We plan to support other console 3D entry systems as soon as they are established and libraries for them come out.

Thanks for your answer.


Sounds very good.


Ah okay, I thought it's not implemented because of Firoballs posting, quoted above. I only have a Wii Remote Plus (integrated), so can't try figure out the difference between with and without. ...furthermore I would expect, there have to be some commands to check it. There are Wii games you only can play with the MotionPlus... hmm, your answer seems to be not that plausible..;)


Okay, but at all there is a need to update the manual, especially the example. Also the default values for the commands should be termed (e.g. wiimote_ir, var wiimote_smoothfac).
Every body who's used to the wiimote programming - please share your experiences! (maybe should open a new thread...)


Yes, Firoball developed the dll, so probably he is the capable contact person - will immediately write him a pm for please take a look inside here...

But one problem should be able easily to solve: "Where's the ir_dot.tga file you need for WII_DEBUG ?" It doesn't exist, so you should update ackwii.c (by rename it to dot.tga) or create the meanted file.

Wiimote+ as far as I know needs some extra code, meaning it will not automatically deliver better sensor values. They have to be grabbed separately. I haven't further looked into it.
I don't own Wiimote+, but it should at least work for getting the accuracy of a standard Wiimote.

That pan is always 0 is how the Wiimote works. Pan can only be detected by additionally interpreting a sensor bar and the IR camera, otherwise it only delivers 2 dimensions of angles (Wiimote does not know in which direction you place your TV just by reading the sensor values, so it cannot know its pan either).
The calculating of a Pan however is to what I have read not very accurate and can lead to wrong values quite easy.
I haven't looked into how to do it in detail.

Accelerator values and angles are mixed. Actually the Wiimtoe does not deliver angles but just accelerations It is possible to extract the angles when adding all sensor values and then being close to 1. In this case it's possible to calculate tilt and roll out of the acceleration data. And that's also where the angle jumps are coming from. The acceleration data changes sign at certain positions, and this will is forwarded to the angle calculation routine.
The classic Wiimote delivers pretty inaccurate data. The feeling of good Wiimote handling is usually done by the game.
There is a reason why they released Wiimote+.


There was some headtracking stuff done by Puppeteer working just fine. I helped out with some code parts of my own old try and also wrote a version without using a shader.
Puppeteer's code is easy to understand and easy to use, so I did not recreate another headtracking script. It can be found somewhere in the Lite-C contributions.