Forum: Bryce


Subject: Bryce's Parallel Light is equal to...?

AgentSmith opened this issue on Jan 19, 2006 ยท 38 posts


AgentSmith posted Thu, 19 January 2006 at 5:47 AM

As some of you know the Parallel Light is a lie. It just doesn't work as a "Parallel" Light should. In fact, it's actually just a Spotlight with a different lame, er, I mean name.

How can you have a real Parallel Light?

Well, I thought, place spot/parallel lights side by side, down a row. Logical idea, but it doesn't work. The spotlight inherently has some amount of falloff on its edges, no matter what you do, so placing them perfectly side by side, you can see the edges easily. Overlap them a little? Nope, doesn't work either, the overlap shows up more intensely than the previous falloff edges.

The only way to make it work on some level of decency, would be to overlap the spotlights a little, AND then blend them as perfectly as possible using the spotlights "Edge Softness" value. This is the only way I have found it to "work".

Three lights sized at 20/20/20, each placed 30 units apart, and each with a Edge Softness of 40, seems to KINDA work. Not perfectly, but its okay for a start. I can still see some slight edges, but they are faint. Yet, it still kicks the butt of the default Parallel Light.

Maybe someone can fiddle with those parameters and come up with the perfect setting, as to have it more seamless looking...

AgentSmith

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


pidjy posted Thu, 19 January 2006 at 6:12 AM

er... why would you need to place parallel light side by side, when you just gotta resize only one parallel light to get the same result?


AgentSmith posted Thu, 19 January 2006 at 6:18 AM

Nope...go ahead and try it. A Parallel light is just a long spotlight. Resize it longer and the ends will eventually fade out to nothing. AS

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


AgentSmith posted Thu, 19 January 2006 at 6:25 AM

I may be using er, slightly incorrect terminology as far as what a stereotypical parallel light is usually considered... But, we WERE short changed when it come to the parallel light in Bryce. It should give us a CONSTANT level of light. It does not though. AS

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


pidjy posted Thu, 19 January 2006 at 6:27 AM

not if you give your parallel light twice the height of it longer side. if your light is 250 long, then it should be 500 tall.


pidjy posted Thu, 19 January 2006 at 6:31 AM

Well it can't give a constant level of light, because the render engine is based on raytracing algorithm, and this algorithm just can't calculate a light source as a coherent light source (like a lazer) because it's based on natural light behaviour.


pidjy posted Thu, 19 January 2006 at 6:35 AM

and considering this, using Bryce, parallel don't mean coherent!


marcfx posted Thu, 19 January 2006 at 7:43 AM

Yep, as the 7 above 8} Marc (Lost me when you said Bryce!!!)


Smile, your dead a long time :)


pakled posted Thu, 19 January 2006 at 10:21 AM

ok..though I'm never parallel or coherent most of the time..;) what would be the use of a parallel light being parallel? just asking..;)

I wish I'd said that.. The Staircase Wit

anahl nathrak uth vas betude doth yel dyenvey..;)


Erlik posted Thu, 19 January 2006 at 10:38 AM

Er, if you increase the light and/or remove fallof, you don't get the shadows on the edges. Almost the same if you click Infinite Light. then, I created a parallel light 20x20x20. Intensity 25, Soft edge 50. Used the default gradient, Linear fallof and Infinite checked. Strung five lights on 5 BU difference. IE, the first one on 0, second on 5, third on 10... No edges inside - one rectangular light trail with fallof at the edges. Unfortunately, the light looks like there are two light bulbs at the positions of lights one and five. But the lighting difference is really not that noticeable. the main problem is that parallel light has a strong light source in the centre. BTW, try positioning parallel light above the horizon line, checking infinite and volume, and see what happens when the ray intersects the horizon. :-) Totally weird. Sorry, no pics, my Photoshop is still kaput.

-- erlik


zakalwe posted Thu, 19 January 2006 at 11:36 AM

I tried with some negative lights. The distance from the plane is is about 1/2 tjan the positive lights. The intensity of the negative lights is 1/10 of the positive ones. This should neutralize the spotlight effect.

zakalwe posted Thu, 19 January 2006 at 11:38 AM

and this is the result: over you see only the positive lights active, under under the negative work too

zakalwe posted Thu, 19 January 2006 at 11:40 AM

but look at the shadow!

zakalwe posted Thu, 19 January 2006 at 11:48 AM

ah, of course negative lights don't cast shadows


pidjy posted Thu, 19 January 2006 at 11:53 AM

or set the negative light shadow to white ;


AgentSmith posted Thu, 19 January 2006 at 2:59 PM

Parallel light...in my book that would be a long flourescent light. (although that would also be a type of radial light, aslo.) I'll have to play with all these settings.... AS

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


marcfx posted Thu, 19 January 2006 at 3:14 PM

Its a good job 'Universe' wasnt added to this theme....i would be totally out of my class!! LOL Marc (Always playing to see what comes out!) ;)


Smile, your dead a long time :)


InfernalDarkness posted Thu, 19 January 2006 at 8:18 PM

The parallel light thing always got me down, too. I have to disagree with you Pidjy, from #6, in that even though Bryce IS only a ray-tracer, ray-tracing wasn't and hasn't been made to simulate natural light at all... Rays are "data-rays", and not "light-rays", and behave nothing like photons. Check out Maxwell or mental ray for more information on that topic. Perhaps someone should suggest to DAZ something about fixing the parallel lights for version 6? I don't think I saw it on their "to do" list last time I looked...? Parallel lighting algorithms are really, really straightforward and should have been incorporated in version 1! But that's all just techy-ness. I'm trying to think of a use for the parallel light now, and simply can't... For a tube light, you'd still have to use multiple radial lights (the more lights, the cleaner it would be) since a parallel light only emits light in 1 or 2 directions, and a tube light emits light in ALL directions EXCEPT those 1 or 2 (the endcaps). A tube light would be the opposite of a parallel light... Zakalwe, in #13 and #14 your effects were caused by negative lights emitting POSITIVE shadows. Imagine a negative flashlight (negative photons exist?!?), where the spot you shined it on would be dark, and whatever shadows that cast would be lighter! That's what's happening in that image in #13... You can test it with one neggy light and a sphere, it's pretty weird but very cool! FLASHDARKS?!?!


Vile posted Thu, 19 January 2006 at 11:56 PM

Um I am not sure what your point is or what you are trying to accomplish Smith? I mean light in the real world has fall off and does not really appear as a square even if your light source is square. give me some idea of what you are trying to do.


AgentSmith posted Fri, 20 January 2006 at 2:17 AM

Think Flourescent light bulb, but in just a strip, shooting in one direction. The Sun, in real life is a parallel light. It is so far away when the light finally hits us, the light rays are all but running parallel to one another. AS

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


AgentSmith posted Fri, 20 January 2006 at 2:17 AM

"Perhaps someone should suggest to DAZ something about fixing the parallel lights for version 6?" WAY ahead of you....;o) AS

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


Erlik posted Fri, 20 January 2006 at 3:09 AM

Sun is an omni, ie radial, light. Can't be parallel, since it's a sphere.

-- erlik


AgentSmith posted Fri, 20 January 2006 at 3:38 AM

Attached Link: http://www.peachpit.com/articles/article.asp?p=174370&seqNum=4&rl=1

Yes, yet because of its distance, its light becomes basically parallel. "Due to the fact that the sun is so far away from the earth, by the time the light reaches us, the rays are essentially parallel to one another (see Figure 7.10). That is why you will also hear the term parallel light when referring to directional lights. The farther away the light source, the more parallel the rays become. Directional lights do not fall off with distance; the intensity is constant everywhere in the scene." Link; http://www.peachpit.com/articles/article.asp?p=174370&seqNum=4&rl=1

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


AgentSmith posted Fri, 20 January 2006 at 3:40 AM

Attached Link: Etereae Studios - "Introduction to 3D"

"Parallel: This is the ideal light for simulating the sun. Radial light can be used to represent the sun since it is a star that is located at a concrete point and that emits light in all directions. However, in relation to us, the sun is located far, far away. So much so that, to position a luminous point thousands of kilometres away is not practical. Thats why parallel light is used: it is called parallel because the rays emitted are parallel, virtually like the rays of the sun when they reach the Earth. The rays remain parallel even when placed close to the scene."

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


Erlik posted Fri, 20 January 2006 at 9:49 AM

Astronomy... Get a Straight Answer "Are the Sun's Rays Parallel? The Sun covers about half a degree of the sky. So rays that come from opposite edges of the Sun have directions which differ by half a degree and are not parallel. Rays which reach your left and right eye from a distant star, on the other hand, are very close to parallel--they may meet somewhere at the star, at the same point (and then they converge) or separated points (and then they probably diverge), but the eye and even the the telescope cannot resolve such details. Rays from the same point on the Sun are pretty much like those from a star. But again, "same point" is hard to pin down--even sunspots may be thousands of kilometers wide." And besides, that's why we have shadows with soft edges, 'cause rays come from different positions in space.

-- erlik


pakled posted Fri, 20 January 2006 at 11:25 AM

does it matter?..;)the sun is a radial light, of the portion that strikes the earth consists of parallel rays, not withstanding the Einsteinian curvature of light rays due to gravitation...yada yada..;)

I wish I'd said that.. The Staircase Wit

anahl nathrak uth vas betude doth yel dyenvey..;)


AgentSmith posted Fri, 20 January 2006 at 11:34 AM

Zactly, lol. Good counter, though. ;oD Your Forum-Fu is strong... AS

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


PJF posted Fri, 20 January 2006 at 8:34 PM

For anyone still in doubt, I hope the above image illustrates the limitation of the Bryce "parallel" light. In this and the following images the shadow colour has been set to bright red for illustrative purposes. The falloff is set to "none" with the intensity set by using a dark grey for the light colour.

In this scene there are five vertical cylinders with a parallel light placed 500 units directly above the centre one. If the light were truly parallel, the shadows of the four outlying cylinders would be identical to that of the centre - i.e. hardly visible.

In the real world, if you spaced four poles twenty feet away from a centre pole in this arrangement, an overhead sun would provide five identical (non) shadows. Sunlight is indeed, to all intents and purposes, parallel. I suspect the Bryce sun is a radial light set at a massive distance.


PJF posted Fri, 20 January 2006 at 8:35 PM

In this image, the parallel light has been changed to a round spotlight with otherwise identical characteristics. The shadows are the same.

PJF posted Fri, 20 January 2006 at 8:36 PM

In this the parallel light is changed to a cylindrical light. Same shadows.

PJF posted Fri, 20 January 2006 at 8:37 PM

In this the light has become a radial. The shadows are still the same. I think it's fair to say that all Bryce lights are just the radial light with various vignetting effects applied.

PJF posted Fri, 20 January 2006 at 8:37 PM

A work-around for outdoors scenes is to place the light 10000 units (the max) above the scene. This is the parallel light.

PJF posted Fri, 20 January 2006 at 8:38 PM

But again, the shadows from a radial light are identical from the same height. You'd be better using the Bryce sun if "parallel" light is your prime target.

Incarnadine posted Fri, 20 January 2006 at 9:02 PM

I always used the set it far away to simulate idea that PJF mentions once I discovered the un-parallel lights function.

Pass no temptation lightly by, for one never knows when it may pass again!


AgentSmith posted Sun, 22 January 2006 at 3:39 AM

Nice conclusion PJF, thank you! AS

Contact Me | Gallery | Freestuff | IMDB Credits | Personal Site
"I want to be what I was when I wanted to be what I am now"


InfernalDarkness posted Sun, 22 January 2006 at 2:59 PM

Still, as a I mentioned, almost every other program has a true parallel light.. It would be very easy to code for such a thing. Falloff coding is much more sophisticated... Hope that Daz takes this small bug into account on the next run of Bryce!


Gog posted Mon, 23 January 2006 at 10:29 AM

Nice work pjf, sometimes I avoid the sun as a parallel source as it doesn't feel as controllable as lights, but it tends to be on close up scenes rather then distance!

----------

Toolset: Blender, GIMP, Indigo Render, LuxRender, TopMod, Knotplot, Ivy Gen, Plant Studio.


Aldaron posted Thu, 26 January 2006 at 2:40 PM

Attached Link: http://www.renderosity.com/messages.ez?Form.ShowMessage=1918892

Sorry so late in posting. The link is to a post on some lighting experiments.