Good Renderer, Better Terrain Creation, But...

Started by rcallicotte, September 30, 2008, 09:36:40 AM

Previous topic - Next topic

rcallicotte

Oshyan (and Planetside or whoever) -

Recently, I read you (Oshyan) saying the greatest strength of TG2 is its rendering engine.  While I have no arguments about the rendering engine, I do wonder how having the basics of terrain only in our scenes will help anything we have planned in the future for TG2.  I don't want to quibble about the fact we can import a smithering of paltry objects (which do look good in our renders), but I think the strongest factor within TG2 is its terrain creation aspects.  That is difficult to find anywhere else.  Where else can we have real planets with near realistic planetary features? 

But, for most of the work I'm seeing within the matte creation community and / or movie industry and / or advertising agency, I am having a difficult time seeing what else we (the consumers of TG2) can actually do with this other than strengthen Matt's programming skills (a great skill, by the way).

Granted, a few TG2 users have used TG2 only to create some things for a few customers.  I really think it could do more, though.

Please let us know what you see, all quibbling aside.  If you don't see anything else in TG2's future that might include a complete package for matte creation, movie industry and advertising, then I'm happy to keep using this as I have been.  It's a geek fest in here.   ;D
So this is Disney World.  Can we live here?

PG

TG2 could be used for loads of stuff once more work is put into certain areas. Climate models could be graphically simulated with TG2 if an animation module was created to allow precise control of the world elements and a physics system was included too perhaps.
Figured out how to do clicky signatures

rcallicotte

@PG - The point - usually this sort of question is answered with a lot of big IFs.  This is to be expected, but I'd rather see Planetside include the elements that will make this a mega-tool for media, while not getting lost to some huge company that could well see its potential.

For example, what if someone like Autodesk saw the potential of including many of TG2's strengths in one of its many packages.  That might be good, if the end result would be an easy to handle package and we don't lost TG2's strengths.  Cost might be a little expensive, but where else could this be going?  Whether it's Planetside or someone else who comes along like this (there could be more like Planetside's TG2 in the future), it won't end here.  How could it?

Exciting.
So this is Disney World.  Can we live here?

PG

Well I can't imagine Terragen has gone unnoticed in these circles, nor the potential that this type of program has. But are you talking about someone like Autodesk buying out Planetside and using the bits from TG2 in a program like 3ds Max or them fronting the money for more large scale development of TG2?
Figured out how to do clicky signatures

rcallicotte

Both ideas (and maybe a few other scenarios) are crossing someone's mind out there, I bet.
So this is Disney World.  Can we live here?

PG

I just hope EA doesn't try to buy it out and use it as a terrain generation program. They mess up everything they touch.
Figured out how to do clicky signatures

sjefen

Autodesk is the same. They are also trying to buy the whole world :)

- Terje
ArtStation: https://www.artstation.com/royalt

AMD Ryzen Threadripper 1950X
128 GB RAM
GeForce RTX 3060 12GB

PG

yeah, and it's unbelievable how much their software costs. what is it, £3000 for 3ds Max? Most people can get the same results with blender if they used decent plugins.
Figured out how to do clicky signatures

Oshyan

Well, I'm curious what you'd *like* to be able to do with it. TG2 and related applications (Engen) has at this point been used in as many feature films as Vue, to my knowledge, and probably more individual shots considering Engen's heavy use in Stealth. So clearly these applications are being used for the purposes we've talked about being one of our focuses, i.e. effects work in feature films. But the fact remains that TG2 and applications like it are very much niche applications. That's a limitation we are not only well aware of, but that we rely on to provide us a market. We have no illusions that we could ever compete with larger, more general applications, nor would we want to.

So the reality is TG2 will never be a major part of many films simply because there is often little need for non-location-based outdoor scenery. However in cases where there is such a need, TG2 can absolutely shine and becomes almost indispensable. Stealth would have been impossible without an application like TG2. Likewise many science fiction films, like Star Trek: Nemesis, have great need for this kind of application. In the future I think more and more productions will also go the way of films like "300", with the vast majority of filming done on a set and with backgrounds added digitally. This is another prime example of value for TG2.

Yet even in those cases TG2 is not likely to ever comprise the entirety of any single shot, or even of a matte background. With the output of any application, included 3DS Max, Maya, etc., there is almost always compositing of multiple elements and other "post work" to get the final film frame. This process is vital for getting the exact results needed (as no one application ever seems to be quite right for *everything*), and for maintaining consistency in visual presentation throughout the film. So even while TG2 may be extremely valuable in many cases, you will always be working with other applications, or other artists using those applications. Even as import/export improves, there will still be elements that must be rendered in other applications, simply because that is their area of specialization (fluid dynamics is a great example).

As we have said before this is really the core of the market we are aiming for with TG2, those working in the effects industry and who will likely be working with a suite of applications to accomplish their goals. That being said TG2 is still very applicable to hobbyists, who comprised the vast majority of our users for TG 0.9, and whom we also want to maintain as satisfied customers. Most users of TG 0.9 never had any commercial success, as is true with any other application in this area (Bryce, World Builder, Vue, etc.). They used TG for artistic expression and fun. TG2 has the same value. But again if what you want is commercial viability and you're asking yourself whether the skills you're building will have any value in that market, yes there is a definite possibility. Just remember that, as it is a niche in the software market, so too is it a niche in the employment market for effects professionals. The flip side of that is that professional-level landscape artistry is rare and those people will be increasingly in demand as the film industry relies more and more on digital sets and set extension. So if you get good enough in this area, you may have career possibilities.

Finally, in regards to the areas we need to improve to best serve our markets, I think it's clear that import/export will be very important, and likewise animation. These are areas we plan to focus on once we finish the initial release of TG2. Fortunately the skills you can develop now in scene creation will absolutely be applicable in the future once more powerful tools become available for working with other applications.

I hope that helps answer some of your questions.

- Oshyan

mr-miley

I can see Terragen being used for some pretty dooozy architectural visualisations (once we loose the 16 shader limit on imported objects. I see quite a few with what I do for a job, and most of them suck. They are either dodgy 3d renders of new buildings superimposed onto even dodgier photo backdrops or very obvious "artists interpretations" which are fine, but like concept cars, never look like th finished thing. I think its an area that Planetside should investigate getting a "toe in the door" of.

I would use TG2 in my job for things like borehole location plans, 3D ground cross sections etc if it weren't for the fact that UK DEM data is crap and very expensive. I would say that there are a lot of uses that people could find for TG2 that are even more niche than the film/games work, its just that a lot of the people out there dont know TG2 exists (nor should they, unless they've gone looking)

Anyway, thats my 2p worth.... so there  ;D

Miles
I love the smell of caffine in the morning

cyphyr

The current Import Export limitations do need to be addressed but I dont think theyre that bad really, most objects can be imported without any real issue, the 16 shader limit is a hassel, not a full stop on getting a project done. As regards integration wityh other application I think, as Osyian has intimated, that it is not so much the direct integration with the app (re: Vue Infinite) way of going (very buggy in Vue's case) but rather the integration of the output ie the actual images, bitmaps that TG creates. In this light I would propose that Terragens development lent towards better compositing tools as in customisable alpha masking of differant scene emements (individual Cloud Layers, terrain depth etc). This will make Terragens output much easier to integrate with the output of other packages in proper compositing packages such as AfterFX etc. Just my 2pence worth too
Richard
www.richardfraservfx.com
https://www.facebook.com/RichardFraserVFX/
/|\

Ryzen 9 5950X OC@4Ghz, 64Gb (TG4 benchmark 4:13)

Hetzen

#11
My brief experience with TG2 has highlighted a few limitations I'd certainly like to see addressed on a final release.

1/ Camera import position scaling and curvature correction based on altitude. I tried to get around the problem of not being able to rotate a planet in TG by exporting a camera move from max, linked to a null rotating around a sphere. I found two problems. First, when I scaled up my scene in Max to work in metres, my objects expanded out of the limits of Max's viewports, so had to re-scale down to make corrections, then re-scale back up for export. Secondly, due to the way TG curves land into a sphere as altitude increases, it's pretty impossible to compensate for that in an external package.

2/ Z-Depth output, would be a serious boost to being able to mask out areas in compositing. (There maybe a node work around, but not found it)

3/ Layered output. Being able to isolate shader groups to separate outputs. (Again, there maybe a node workaround, but would be good to create a node which directs a node path to a layer output) Vue has a PSD output which is very powerful, it would be good to have something very similar in TG, with the added ability to define through nodes exactly what you want to isolate.

4/ Curves graph on TGs camera.

5/ A geometry cut tool, so you can isolate a piece of terrain to be exported into an external 3D package, so we can get precise interaction with specific areas in the landscape (Think how good it would be to get a particle system to work against fake stones for example)

6/ A mask shader which you can apply to an imported object and output to a new layer from TG, so we could comp stuff flying through a cloud for example, or waves on a boat. (Again, I'm sure you could set this up through a clever use of nodes and RGB colouring)

These would be seriously powerful additions to an already powerful app.

Another twopence in the pot.

Jon

Oshyan

Jon, we do have plans to work on many of the areas you mention, including better camera (and other data) import/export, z-buffer and other output layers, and curves for animation (not just cameras).

The export of specific areas of terrain is already possible (square areas at any rate). Only heightfields can be exported in this way, so no overhangs are possible, but still it can be very useful. Use the Heightfield to LWO node. You define the area exported in meters in the size settings of your heightfield generator, and the detail with the resolution parameters (currently unlabeled) in the same area. A more complete explanation is here:
http://forums.planetside.co.uk/index.php?topic=4727.msg49460#msg49460

- Oshyan

Hetzen

Thanks Oshyan for the reply. I jumped feet first into this program, and have a habit of hitting things hard on touch down. I think you have a seriously great app here, and have no reservations on my spend on it at this stage in it's development, especially as you say you will be addressing things which will make this more industry friendly.

As for the export issue, I was thinking more in line with a planet side (lol pun sort of intended) pro-boolean which allowed geometry output after displacement shader activity. For example, using FumeFX against geometry, or Thinking Particles to interact in collision fragmentation.

Oshyan

Are you aware of the "LWO Microexporter" functionality? It sounds like it might do at least some of what you are looking for in geometry export...

- Oshyan