http://www.deviantart.com/deviation/52030680/
i recently finished a render and noticed that in the middle/top/left part theres something that suggests a rainbow
i never knew tg2 could do that
bug?
new feature?
o.O
Hmm, I think I see what you mean, although it's extremely subtle. TG2 can't technically do rainbows just yet, but there are ways to fake it (use a large planet with a big atmosphere on the horizon of your main planet with appropriate lighting).
- Oshyan
i see, then how was this lil one created?
mm, not to be rude or anything, but i don't see any rainbow. would you like to outline it or something?
I think I see what you are talking about. Looks to me like some kind of a lucky mix of rays and colors.
Nice image BTW.
maybe this will help....I once had a go at rainbows...I didnt quite succeed but I came pretty close....
http://forums.planetside.co.uk/index.php?topic=789.0
scroll down till the end of the page...
Quote from: Nameless-Designer on March 30, 2007, 05:21:04 PM
i see, then how was this lil one created?
I'd say it's just a coincidence, possibly due to red decay or other atmospheric effects. These are related to how rainbows are created but rainbows are the result of optical refraction in small drops of water (rain, fog, etc.) rather than that due to overall atmospheric density. TG2 only simulates the latter. Most likely any rainbow effect TG2 would incorporate would not be based on simulation anyway as it would be quite demanding for a more or less identical result to simply "faking" it.
- Oshyan
Would you have to simulate it, at the moment we have some ray tracing capability within TG2 for shadows and water would it not be possible to extent the Ray tracing abilities of TG2 to phenomena such as rainbows and other atmospheric phenomena?
Regard to you.
Cyber-Angel
Sure you *could* simulate rainbows, but doing so properly would be extremely time-consuming to render and the results would be about the same as doing a simple gradient overlay. Your fascination with true simulation is a bit ahead of its time I think. Given the choice between a 200 hour render with a simulated rainbow and a 2 hour render with essentially the same visual result, which would you choose? Why is simulation "better" if it looks the same in the end? In some cases you really do need to simulate at least the most significant aspects of a phenomena in order to accurately portray it visually, but in others - as I think is the case with rainbows - it's largely a waste of time. More complex atmospheric refraction phenomena such as halos and the like might be more worthwhile to simulate, but they're rare enough to not be worth significant development time at this point. They would be better handled as a plugin, for example a direct connection with Halosim: http://www.atoptics.co.uk/halo/halfeat.htm
- Oshyan
I didn't say you would have to necessarily simulate any thing just extend the ray tracing capabilities that TG2 has right now to halos and related phenomena. In any case if one day Terragen theoretically could do water falls then rainbow type phenomena should be present around the splash zone when seen from certain POV's but thats for the future.
To answer your question I would take the 2 hour render if it had the feature set and optimisation to stop the Vue range of products stone cold dead, but thats just my opinion on what I would like for the future unless that is not consistent with the market segment that Terragen is aiming for.
Regards to you.
Cyber-Angel
Hey man, nice pic :)
Looks a lot like mine though...:P And I mean, a LOT like it :p Especially the squiggly clouds in the distance :)
I can't seem to find a rainbow though..but there is some colour diffusion near the middle/top places that may have been termed as a rainbow.
Quote from: Cyber-Angel on April 01, 2007, 11:14:02 PM
I didn't say you would have to necessarily simulate any thing just extend the ray tracing capabilities that TG2 has right now to halos and related phenomena. In any case if one day Terragen theoretically could do water falls then rainbow type phenomena should be present around the splash zone when seen from certain POV's but thats for the future.
Cyber-Angel
That's essentially simulating at least the light refraction, which is what I was referring to. Simulating rain itself is another issue, but again one that is probably better addressed with a simple non-simulation based approach. Raytracing of water particles to achieve a rainbow would provide a pretty noticeable render time hit and doesn't seem particularly worthwhile.
- Oshyan
Light refraction...mm...
Just wondering, will 'caustics' be implemented in the future?
Crystal water with caustics in an underwater shot would just be magnificant.
By the way, if you actually ARE able to make rainbows using raytracing of water particles, we would really not care about the render time. To be able to achieve such a feat, as to create a rainbow from certain POVs (or easier, from all POVs), is so good that I wouldn't care at all about the render time (unless it took..a LONG time (250 hrs + for 640x480 pic)
It would certainly topple the competition ;)
I would say that the implementation of caustics within the final commercial release of TG2 even if its only the High-end TG2 Deep Edition would be a logical edition to the feature set, since the TG2 FAQ dose say that TG2 will be capable of underwater images.
Given that TG2 has GI as part of its rendering and the fact that caustics are a part of GI then I don't see why not as long as TG2 supports Shadowed and Volumetric caustics I don't mind.
They would have to consider and there is no way round this both reflection and refraction of light via ray-tracing; there is not way round this.
Since you dealing with underwater caustics you will need to have the render consider light transport through water, important factors to consider fresh vs. salt water, water clarity (Parts per-billion of sedimentation content of the water body which would affect the way light is transmitted), Salinity level of the water (Sea water due to its salt content transmits light differently form fresh), the geometry of the receiving surface, and support both reflective and refractive caustics among other things.
Ideally the caustics should change with changes in the water surface as should the shadows they cast as any one who's ever dived (Not Me) in shallow water will tell you about the caustics changes seen on the ocean floor.
It would be good if caustics could be cast on to other objects such as rocks in the water and so on. On the whole then these are a few things to think about for the future and as a subject area caustics are been intensely studied in the field of CGI and are not some thing to be rushed into without due care and consideration.
VUE form what I understand has the ability to do caustics already so this is an area where Terragen is lagging behind right now, but may gain this capability some time in future if not for the first commercial release but in a latter update.
Regards to you.
Cyber-Angel
;D