[x3d-public] Euler rotation
Joseph D Williams
joedwil at earthlink.net
Thu Jan 10 13:45:28 PST 2019
➢ user-niceties were sometimes compromised in favor of efficiencies.
User-niceties were sometimes omitted because they don’t work reliably in the general case, especially x y z angles in interpolators.
➢ clear and nice to the user
Until it doesn’t work anymore. Real nice work, and we do need something to flag quaternion vs axis-angle, but does your hanim work?
Good Luck,
Joe
From: Leonard Daly
Sent: Monday, January 7, 2019 12:02 PM
To: x3d-public at web3d.org
Subject: Re: [x3d-public] Euler rotation
Hi Albert et al,
Another option (at least for user-friendly representations) would be to use converter symbology in the field definition. For example, right now rotations are represented by
normalized-vector angle
where normalized-vector is a 3-space vector that has been normalized that represents the axis of rotation; and angle is the angle of rotation.
A set of symbolic converter functions could be specified that allowed the user to specify the rotation is a number of different manners. These functions would accept the rotation values and convert them into a standardized internal form (or default input form, depending on the particular browser).
For example, converter functions could be defined for Axis-Angle (default), Euler, and Quaternion. The rotation formats would then be equivalent.
AXIS-ANGLE (normalized-vector angle)
EULER (angleX angleY angleZ)
QUATERNION (qw qx qy qz)
All would be legal rotation field values and would convert to the same internal format.
This would work when the X3D encoding was string based (XML or Classic VRML). The binary encoding would need to process the input (which converts do now) to a standard/default format.
This allows the user interface encoding to specify something that is clear and nice to the user; while maintaining a standardized internal format. Note that the function name is part of the value, so changing the value would require a function (or use of the default).
I implemented this in XSeen, though it hasn't documented yet. In addition to rotations, I use something similar for colors (fractional, value (decimal), byte (hex), labeled (white, red, green, etc.)). For both of these, I drew my inspiration from CSS and their specification of colors and other data values.
Essentially all of the fundamentals of X3D were taken from VRML, which was taken from OpenGL pre-1997 with a number of features added from those long-ago years. Computation was expensive then so user-niceties were sometimes compromised in favor of efficiencies.
Leonard Daly
"So the term euler would replace the term rotation and also be a different datatype than rotation"
Yes, 'euler' would be complementary, an alternative way, to set the initial orientation. To avoid Gimbals lock, one should of course use quaternions for dynamic rotations.
I think the sequence of the individual numbers for the angles and their sign should correspond to the x y z axes as defined by X3D. I don't really have an opinion on the yaw, pitch roll terms: -z seems a natural gaze direction for this.
Albert Jan Wonnink
Van: x3d-public <x3d-public-bounces at web3d.org> namens x3d-public-request at web3d.org <x3d-public-request at web3d.org>
Verzonden: maandag 7 januari 2019 14:44
Aan: x3d-public at web3d.org
Onderwerp: x3d-public Digest, Vol 118, Issue 15
Send x3d-public mailing list submissions to
x3d-public at web3d.org
To subscribe or unsubscribe via the World Wide Web, visit
http://web3d.org/mailman/listinfo/x3d-public_web3d.org
or, via email, send a message with subject or body 'help' to
x3d-public-request at web3d.org
You can reach the person managing the list at
x3d-public-owner at web3d.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of x3d-public digest..."
Today's Topics:
1. Re: Euler rotation (Joseph D Williams)
----------------------------------------------------------------------
Message: 1
Date: Mon, 7 Jan 2019 05:43:28 -0800
From: Joseph D Williams <joedwil at earthlink.net>
To: Albert Jan Wonnink <awonnink at hotmail.com>, John Carlson
<yottzumm at gmail.com>, "vmarchetti at kshell.com"
<vmarchetti at kshell.com>, X3D-Public <x3d-public at web3d.org>
Subject: Re: [x3d-public] Euler rotation
Message-ID: <E1ggVBd-0003vy-Se at elasmtp-curtail.atl.sa.earthlink.net>
Content-Type: text/plain; charset="utf-8"
From: Albert Jan Wonnink
Sent: Monday, January 7, 2019 2:03 AM
To: John Carlson; Joseph D Williams; vmarchetti at kshell.com; X3D-Public
Subject: Re: [x3d-public] Euler rotation
? Hi Joseph,
? Thank you for your reply. Just to be sure: the context I meant to discuss is about having a convenient attribute in the declarative syntax, say 'euler', to set the initial orientation of the transform, so anyone can use either 'rotation' or 'euler'.
?
? <transform euler="20 40 10" >
?
Fine, Albert,
So the term euler would replace the term rotation and also be a different datatype than rotation?
The important technical and usage detail is accepting quaternions since probably the preferred internal form.
The main thing about degrees is that they won?t always work. That is why Euler moved on. If you are making video frames, then maybe ok to author or capture frame data in degrees but when you want to have some realtime or variabletime fun, then interpolation may not operate as expected or at all.
? So I interpreted Vince's $yaw, $pitch and $roll to be just placeholders, for manual substitution.?
Thoughts about the use of pitch yaw roll names for the x y z rotation/orientation axes?
Just as a memory aid of course.
I described it as I found it in viewpoint where the default is gaze -z and hanim joint animations where the default is gaze +z.
? You are right using this with either radians or degrees would be equally simple.?
Most simple and reliable using the x3d form thinking with radians, no doubt. The x3d recognition of quaternions would improve x3d capability by being even more simple and reliable and also giving improved level of validation for rotation and orientation values.
? Regards,
? Albert Jan?
Thanks for the thoughts,
Joe
?
??
From: Joseph D Williams
Sent: Sunday, January 6, 2019 7:20 PM
To: Albert Jan Wonnink; vmarchetti at kshell.com; X3D-Public
Subject: Re: [x3d-public] Euler rotation
?
Hi Albert and All,
When you use euler (or Tait-Bryan) angles for a while you will find that you get problems.
Actually there is plenty of writing about the details maybe even in the x3d or hanim archives and widely on the web.
?
In addition, when you create or edit the things much then you will find it is actually a lot easier to use quats or axis-angle radians. In general, for degrees, maybe your editing system may offer as a learning tool if radian is really an elementary block to learning, but really, never in the user code due to misleading simplicity and what ?cha finally gonna tell ?em when the gimbals lock?
?
? <Transform DEF="yaw-rotation" rotation="0 0 1 $yaw">
? <Transform DEF="pitch-rotation" rotation="0 1 0 $pitch">
? <Transform DEF="roll-rotation" rotation="1 0 0 $roll">
?
Besides the fact that a variable is not allowed there
(maybe use a prototype if you need a variable?)
I don't think your naming of yaw and pitch and roll is reasonable for x3d.
?
For default transform rotation 0 0 1 0, sit on it and face +z and find +y is up and +x is left.
then fly the thing using pitch and yaw and roll.
The x3d axis-angle parameters are most likely pitch yaw roll then the scale,
and I think the same order for quaternions, except where the 'scale' is placed and the special limits for values.
?
Why not try the following:
?
DEF Pitch OrientationInterpolator {
key [ 0, 0.25 0.5, 0.75 1 ]
keyValue [ 0 0 1 0, 1 0 0 1.5, 0 0 1 0, -1 0 0 1.5, 0 0 1 0 ] }
?
DEF Yaw OrientationInterpolator { key [ 0, 0.25 0.5, 0.75 1 ]
keyValue [ 0 0 1 0, 0 1 0 1.5, 0 0 1 0, 0 -1 0 1.5, 0 0 1 0 ] }
?
DEF Roll OrientationInterpolator {
key [ 0, 0.25 0.5, 0.75 1 ]
keyValue [ 0 0 1 0, 0 0 1 1.5, 0 0 1 0, 0 0 1 1.5, 0 0 1 0 ] }
?
Thanks and Best,
Joe
?
From: Albert Jan Wonnink
Sent: Sunday, January 6, 2019 2:25 PM
To: vmarchetti at kshell.com; X3D-Public
Subject: Re: [x3d-public] Euler rotation
?
Thanks Vince, for your reply.?
?
As long as the angles and X3D code is? generated by code one can of course easily calculate the individual parameters for a quaternion. But during the testing of our experimental Unity based browser, and preparing scenes for that, I found it very easy to extend the transform specification with a degree-based Euler attribute (Tait-Bryan would also have worked).
?
The current system still makes it relatively difficult to manually apply an angle of the most used fractions of PI, doesn't it?
?
I guess this issue must have been discussed in one of the early days of the X3D specification. But it seems to me a lost chance to make it more convenient.
?
Regards
?
?
Albert Jan Wonnink
?
?
Van: vmarchetti at kshell.com <vmarchetti at kshell.com>
Verzonden: zondag 6 januari 2019 22:49
Aan: Albert Jan Wonnink; X3D-Public
Onderwerp: Re: [x3d-public] Euler rotation
?
There is no single node Euler angle alternative representation of a rotation transform defined in the current (v 3.3) X3D standard.
?
You can get an easier to edit representation of Euler angle rotation (specifically the Tait-Bryan angles,?https://en.wikipedia.org/wiki/Euler_angles#Tait%E2%80%93Bryan_angles, commonly used for vehicle oriented
with respect to the horizon, by nesting 3 Transforms:
?
<Transform DEF="yaw-rotation" rotation="0 0 1 $yaw">
<Transform DEF="pitch-rotation" rotation="0 1 0 $pitch">
<Transform DEF="roll-rotation" rotation="1 0 0 $roll">
<!-- put vehicle model here -->
<Shape>...</Shape>
</Transform>
</Transform>
</Transform>
?
where for $yaw, $pitch, $roll you need to substitute the numerical value of the rotation angle in radians.
?
?
Vince Marchetti
?
On Jan 6, 2019, at 12:53 PM, Albert Jan Wonnink <awonnink at hotmail.com> wrote:
?
Is there, or should there be, an Euler alternative for the transform 'rotation' property? This makes direct x3d code editing much easier.
?
Regards,
?
?
Albert Jan Wonnink?
_______________________________________________
x3d-public mailing list
x3d-public at web3d.org
http://web3d.org/mailman/listinfo/x3d-public_web3d.org
?
?
?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20190107/b500d676/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 9B68D1D68F1247D1A07F4C368D0D7215.png
Type: image/png
Size: 156 bytes
Desc: not available
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20190107/b500d676/attachment.png>
------------------------------
Subject: Digest Footer
_______________________________________________
x3d-public mailing list
x3d-public at web3d.org
http://web3d.org/mailman/listinfo/x3d-public_web3d.org
------------------------------
End of x3d-public Digest, Vol 118, Issue 15
*******************************************
_______________________________________________
x3d-public mailing list
x3d-public at web3d.org
http://web3d.org/mailman/listinfo/x3d-public_web3d.org
--
Leonard Daly
3D Systems & Cloud Consultant
LA ACM SIGGRAPH Past Chair
President, Daly Realism - Creating the Future
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20190110/511f840a/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 9EE5CF0D69ED4AAC857936DFB68618CA.png
Type: image/png
Size: 164 bytes
Desc: not available
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20190110/511f840a/attachment-0001.png>
More information about the x3d-public
mailing list