>Thomas de Groot  on date 22/02/2013 8.59 wrote:
> On 21-2-2013 19:21, clipka wrote:
>> Am 21.02.2013 15:27, schrieb Paolo Gibellini:
>>>  >Thomas de Groot  on date 21/02/2013 15.22 wrote:
>>>> On 21-2-2013 15:15, Paolo Gibellini wrote:
>>>>> Or a Gancaloon cityscape generator ;-)
>>>>
>>>> Lol! That means digitizing my brain ;-)
>>> Mmm. How much disk space is estimated for an off-line version of such a
>>> data?
>>
>> Well, it can't be /that/ much...
>>
>> (ducks & runs)
>>
>
> Right. I shall keep the floppy disk to myself...
>
> Thomas
>
Ok. It _should_ be included in next release!
Provided with the comments obviously.
Paolo

Post a reply to this message

On 25-2-2013 9:58, Paolo Gibellini wrote:
> Ok. It _should_ be included in next release!
> Provided with the comments obviously.
> Paolo

See image in p.b.i.

Thomas

Post a reply to this message

Here's something else I thought to mention (but sorry for the delay) concerning
the camera statements in the various demo scenes.

The cameras are all constructed differently--some radically so. It isn't so much
that they *are* different (since they all 'work') but rather, it's difficult to
get a sense of what is the 'correct' way to state a camera block. (And this
affects how the basic <x,y,z> scene axes are interpreted.)  This could really
confuse a beginner, for various reasons. Granted, the POV-Ray documentation
explains the various camera parameters (and the pitfalls of using
a 'right-hand' coordinate system vs. a left-hand one); but the demo scenes are
probably where new users go first, to get that all-important first taste of the
program, by seeing an actual render and playing around with the code.

Some cameras use old methods, like 'direction' for zooming instead of 'angle.'
And the "benchmark" scene, as well as some others, uses a truly non-standard
set-up-- up z and direction y.

Here's a different example, from DIFFRACT.POV (posted verbatim)...

camera {  //  Camera StdCam
  angle 90
  location  <3.50, -15.00, 3.00>
  direction <0.0,     0.0,  1.6542>
  sky       <0.0,     0.0,  1.0>  // Use right handed-system!
  up        <0.0,     0.0,  1.0>  // Where Z is up
  right     <1.3333,  0.0,  0.0>
  look_at   <0.000, 0.000, -2.7500>
}

In the docs about CAMERA (at "Placing the Camera"), there's a nice diagram which
shows that 'direction' and 'up' are at right-angles to each other--as they
normally should be. Here, they are not.  As the docs about "Up and Right
Vectors" say:

"Note: that the up, right, and direction vectors should always remain
perpendicular to each other or the image will be distorted."

Also, 'up' (and 'sky') are in the z-direction, as well as 'direction.' So that's
not a right-handed coordinate system *according to POV-Ray*. In the docs about
"Handedness" it says:

"To use a right-handed coordinate system...Your thumb still points up in the
+y-direction...."

All of this might sound like nit-picking, but here's why I think it matters:
When I first started learning the program, I picked a demo scene at random to
render and work with...knowing nothing at all about POV-Ray or it's coordinate
axes. The scene happened to be FISH13. Lucky for me, it had an 'understandable'
camera statement, more or less--where the scene's <x,y,z> directions/axes meant
<right,up,into the scene>. Easily grasped, if only from a simple mathematical
standpoint. So if I wanted to change/experiment with something--like translating
one of the objects in +y--then the render actually showed it moving in +y. But
with a few of these demo scenes, that's not at all what will happen in the
render. Trying to learn even the basics of POV-Ray with a non-standard camera
could turn off a new user--there are just too many fundamental details to
comprehend, all at once.

Of course, to fix this situation would mean re-coding some of the scenes
themselves, if they are still to render as originally intended  (a lot of work!)
But an easier solution might be a simple statement, something like "This scene
uses a non-standard camera set-up. (See CAMERA in the included documentation for
details.) If you are new to POV-Ray, you might want to try a different demo
scene."

Post a reply to this message

On 28-2-2013 4:22, Kenneth wrote:
> Here's something else I thought to mention (but sorry for the delay) concerning
> the camera statements in the various demo scenes.

I fully agree with you!

> camera {  //  Camera StdCam
>    angle 90
>    location  <3.50, -15.00, 3.00>
>    direction <0.0,     0.0,  1.6542>
>    sky       <0.0,     0.0,  1.0>  // Use right handed-system!
>    up        <0.0,     0.0,  1.0>  // Where Z is up
>    right     <1.3333,  0.0,  0.0>
>    look_at   <0.000, 0.000, -2.7500>
> }

I think that this set up was originally made in Moray, which used a 
'different' right-handedness.

Thomas

Post a reply to this message

Am 28.02.2013 04:22, schrieb Kenneth:
> Here's something else I thought to mention (but sorry for the delay) concerning
> the camera statements in the various demo scenes.
I had this feeling too!

> ...    In the docs about
> "Handedness" it says:
>
> "To use a right-handed coordinate system...Your thumb still points up in the
> +y-direction...."
'right-handed' or 'left-handed' does not necessarily mean y=up or z=up
  (or may be also x = up ! - Why not ?? ).
It only means, that to detect the used sense of rotation, we can help
us by looking at our right or left hand!
(This has specially repercussions also on the way, how the rotation and
  the cross product works!)

The docs are here a bit incomplete!

> Of course, to fix this situation would mean re-coding some of the scenes
> themselves, if they are still to render as originally intended  (a lot of work!)
> But an easier solution might be a simple statement, something like "This scene
> uses a non-standard camera set-up. (See CAMERA in the included documentation for
> details.) If you are new to POV-Ray, you might want to try a different demo
> scene."
I'll try to integrate something like this in as many files as possible.
Re-coding all is not possible and also not the best way!
It's also a good thing, that the bundle of demo scene files shows
the variety of the ways, how to use the 'camera'.

  'right-handed' is just a convention, often used math teachers,
to make it easier to talk about how axes are orientated and how
angles turn. (later they tell us, that electrons are left-handed :-))
POV-Ray is able to do also very inconventional things.
That's very good!

Friedrich

Post a reply to this message

Friedrich Lohmueller <Fri### [at] t-onlinede> wrote:

> 'right-handed' or 'left-handed' does not necessarily mean y=up or z=up
>   (or may be also x = up ! - Why not ?? ).
> It only means, that to detect the used sense of rotation, we can help
> us by looking at our right or left hand!

Yes, that's true of course. But using an 'unconventional' handedness system
would cause trouble later IMO, because *everything* in POV-Ray is set up for the
left-handed way (or maybe its *own* left-handed way.) It seems to me that even
an experienced user would have to change over to the 'POV-Ray way' of doing
things at some point, if the coding of scenes is going to make any sense. I
can't even imagine the constant mental re-orientation that would be required
otherwise! ;-)  Perhaps some people CAN do that; but it sounds like a lot of
extra work, conceptually.

> I'll try to integrate something like this in as many files as possible.
> Re-coding all is not possible and also not the best way!
> It's also a good thing, that the bundle of demo scene files shows
> the variety of the ways, how to use the 'camera'...
> POV-Ray is able to do also very inconventional things.
> That's very good!

Yes, I do agree. When I was thinking about all of this, it occurred to me that
the 'unconventional' cameras do show that almost anything is possible in
POV-Ray!

Everything should be OK, with just a little 'warning' to new users.

Post a reply to this message

>   'right-handed' is just a convention, often used math teachers,
> to make it easier to talk about how axes are orientated and how
> angles turn. (later they tell us, that electrons are left-handed :-))
> POV-Ray is able to do also very inconventional things.
> That's very good!
>
> Friedrich
>
>
>
>

In my math classes when looking at geometry, we only used the 
left-handed coordinate and rotation system. Also, Y was always UP and Z 
forward, NEVER relative to the paper's surface.

It was the same in my physics courses.

The right-handed system is mostly used by architecs, and, as most early 
modeling applications where made for architecs, it stuck. It's also why 
we have the infamous Z for the up direction. Architecs use the X and Y 
axis along the ground and on paper where +X is right and +Y is the top 
of the paper.

Post a reply to this message

Am 01.03.2013 04:14, schrieb Alain:

>
> In my math classes when looking at geometry, we only used the
> left-handed coordinate and rotation system. Also, Y was always UP and Z
> forward, NEVER relative to the paper's surface.
>
> It was the same in my physics courses.

What? For us it was /always/ right-handed, for both maths and physics. I 
can't believe it's different across the world in such disciplines.

... unless of course you use a different scheme to assign axes to the 
fingers. We used thumb=X, index=Y, and middle=Z.

> The right-handed system is mostly used by architecs, and, as most early
> modeling applications where made for architecs, it stuck. It's also why
> we have the infamous Z for the up direction. Architecs use the X and Y
> axis along the ground and on paper where +X is right and +Y is the top
> of the paper.

If you put a proper math-style 2D coordinate system (X axis right, Y 
axis "top") onto paper, and then add +Z as height above ground, that's a 
left-handed system to me.

Post a reply to this message

clipka <ano### [at] anonymousorg> wrote:

>
> If you put a proper math-style 2D coordinate system (X axis right, Y
> axis "top") onto paper, and then add +Z as height above ground, that's a
> left-handed system to me.

Egads, then +Z becomes -Z in POV-Ray!  :-O

But you're right, of course.

My own concept of the axes (putting aside what I learned in school) is kind of
an odd one--or maybe not so odd?  I've been sketching things all my life
(mechanical stuff mostly, in semi-perspective views), and 'right' is +x
naturally, but z is *mostly* INTO the page, and y is *mostly* up, but kind of
sticking OUT of the page a little. (Similar to an isometric view when drafting,
but not as extreme.) Maybe this is why POV's left-handed system was such an easy
learn for me.

Post a reply to this message

Am 01.03.2013 07:08, schrieb Kenneth:
> My own concept of the axes (putting aside what I learned in school) is kind of
> an odd one--or maybe not so odd?  I've been sketching things all my life
> (mechanical stuff mostly, in semi-perspective views), and 'right' is +x
> naturally, but z is *mostly* INTO the page, and y is *mostly* up, but kind of
> sticking OUT of the page a little. (Similar to an isometric view when drafting,
> but not as extreme.) Maybe this is why POV's left-handed system was such an easy
> learn for me.
Well, that depends on how we start a plan and how we continue.
Starting with the front view:
   x right, y up. The house behind the front goes to +z = left-handed.
Starting with the top view on ground plan:
  x right front line, y = depth, so z = up seems okay! = right-handed.

Interesting the different positions in the last 4 contributions here!
So all positions are justified - except those, who want to
declare a type of system as 'natural' and 'absolutely correct'!

Her in Germany, some regions, by 'orders from above', have decided that
their teachers at school must use a right-handed system.
And teachers, who use POV-Ray at school, are often confronted with
big resistances, because 'it is using a wrong system of coordinates'.
The students themselves do not have so much trouble with it.
  :-)

Friedrich

Post a reply to this message