On 5/12/2017 5:31 PM, Bald Eagle wrote:
> Stephen <mca### [at] aolcom> wrote:
>> On 5/12/2017 2:31 PM, Bald Eagle wrote:
>
>>> Suggestions?
>>>
>> Only what I do myself.
>> I stick to Mpeg-2 because I had to pick one format.
>
> Oh, the _video_ encoding.   I'll have to go back and look at the still frames
> and see if they have the same problem.
>
>

Yes, that is the first step. That's if you d
see what I described.

>>> I'm going to play with parameters to see what I can get to fit on the screen and
>>> be aesthetically pleasing, but not too wildly unrealistic.
>>
>> Ghost images that fade out?
>
> Elaborate, please.

The moon leaves a trail of increasingly transparent images. It is a 
common effect but I cannot remember what it is called. but I wasn't 
being serious. A solid or  broken line would be good enough.

> I was thinking about the parameters of the radii, the distances, and the orbital
> eccentricities to get a nice but highly unrealistic animation of the
> elliptically orbiting planets- but it sounds like you have another interesting
> idea.
>


>>> Maybe even code in proximity conditions for textures vs uv-mapping.
>>
>> Why, is it not complicated enough?
>
> It is not.  At the moment it's some spinning spheres revolving around a textured
> sphere.  Earth has the very complex texture of pigment {Blue} with specular 0.6.
>
> "We need nine columns for the motions of Jupiter, nine for the motions of
> Saturn, and so on. Then when we have all initial positions and velocities we can
> calculate all the accelerations from Eq. (9.18) by first calculating all the
> distances, using Eq. (9.19). How long will it take to do it? If you do it at
> home, it will take a very long time! But in modern times we have machines which
> do arithmetic very rapidly; a very good computing machine may take 11
> microsecond, that is, a millionth of a second, to do an addition. To do a
> multiplication takes longer, say 1010 microseconds. It may be that in one cycle
> of calculation, depending on the problem, we may have 3030 multiplications, or
> something like that, so one cycle will take 300300 microseconds. That means that
> we can do 30003000 cycles of computation per second. In order to get an
> accuracy, of, say, one part in a billion, we would need 4×1054×105 cycles to
> correspond to one revolution of a planet around the sun. That corresponds to a
> computation time of 130130 seconds or about two minutes. Thus it takes only two
> minutes to follow Jupiter around the sun, with all the perturbations of all the
> planets correct to one part in a billion, by this method!"
>
> http://www.feynmanlectures.caltech.edu/I_09.html
>
> And really, all it is is a conditional statement.
> if the distance > some trigger, use a simple procedural texture.
> if it's less, use a more complex one, or a UV map.
>
>

Yes, I understand but have you experimented with the technique. When I 
did years ago I found little difference in the times taken. It is a good 
technique though.


As for as accuracy goes. I am always disappointed with PovRay's realism. 
The camera in my mind's eye is much better than RL. :-(

>
>>> Still pondering how to rotate the sun faster around the axis than the equator -
>>> how to procedurally texture that....
>
>>
>> You have started me thinking about creating a DF3 animation, of the sun.
>
> Excellent!  Inspiration strikes!
>

Maybe not. I was thinking about converting Blender's particle system 
into a point cloud or something like that but it is not obvious how it 
is done. I'll give it another decade or two to come up with an idea. ;)


> Looks like Bob Hughes has taken a stab at that already:
> http://objects.povworld.org/binaries/sun.pov
>

Always worth looking at Bob's code.

> And it looks like there are some other toys to play with there as well!  :)
> http://objects.povworld.org/cat/Space/
>

There are a lot of good ideas out there. A lot of them very old. :)


-- 

Regards
     Stephen

Post a reply to this message