|
|
TrueSpace, new users, and YouTube
About Truespace Archives
These pages are a copy of the official truespace forums prior to their removal somewhere around 2011.
They are retained here for archive purposes only.
TrueSpace, new users, and YouTube // Roundtable
Post by moogaloonie // Apr 17, 2008, 9:31am
|
moogaloonie
Total Posts: 64
|
It seems to me that some people seem to think TS is lucky to survive between blender which is free and 3DSmax which is the "standard'. There is concern also that TrueSpace isn't attracting enough new users. I want to point out that the company I would think has started the most people off on 3D art lately is likely Daz. While it seems they are primarily in the content business, they have nonetheless shown a serious interest in providing tools to their customers. It seems to me that while in the old days most users learned from tutorials (make the spaceship anyone?) most users today learn by modifying included or purchased content.
Caligari have a powerful foundation in TS 7.5, but it doesn't seem to have satisfied many of the power users who are doing product or architectural visualisation. I think everyone would be happier if TrueSpace were used by more people and had greater recognition. It would seem to me that the best way to bring in new users would be to catch those who are outgrowing Poser and Studio, but still are more interested in animating than modeling ultra realistic models from scratch. I know many purists look down on this type of user, but I think it's important to remember that animating and modeling are very different disciplines. There are now tools like MovieStorm and Antics that provide basic characters and scripted actions and allow to make short films. There is also iClone, which seems geared toward Poser users, and which uses hardware shading to produce short films.
This is the golden age of YouTube, and I would like to see TrueSpace 7.6 blow all of these programs away. It already has a better realtime renderer than any of those programs, but doesn't yet support the ready made content that machinima-type artists depend on to tell their stories. We need Collada import as well as export, since it's not likely we'll soon be loading brokered content (i.e. poser crap) natively. I really must stress that I know there are people who hate Poser, Studio, its content, users, etc. It's just important to remember that as 3D programs mature and add features the whole medium changes. Game studios now have dedicated UV mappers, dedicated texture artists, etc; all things those of us who have been at it for years learned gradually over time as these things came about (I remember pre-normal map days) but would be overwhelming to learn all at once now. The average hobbyist looking to produce something by his/herself often doesn't want to start from scratch. I think TrueSpace could be integral to a whole new wave of independent CG shorts, more so even than the programs that have already targeted this niche. And I don't see any reason why the current users wouldn't benefit from that, even if it means a sudden influx of n00bs of asking question in the forums. |
Post by splinters // Apr 17, 2008, 9:42am
|
splinters
Total Posts: 4148
|
S'funny moogaloonie, I was just thinking the same thing. Having just read 3D world's latest issue, it has a big feature on machinima and an interview with Daz about their integration of Mogware, to make Mogbox, into Daz studio.
Brings us naturally back to Carrara too eh? |
Post by Jack Edwards // Apr 17, 2008, 12:58pm
|
Jack Edwards
Total Posts: 4062
|
These are good points.
In my opinion, TrueSpace is currently a very strong modeler for game content and character creation. The modeling tools are really starting to mature on Workspace side and are very easy to use once you adjust to the new workflows. If you are careful it is very easy to maintain safe geometry for export and use in video game engines. The capability to see the models as they will look in game with custom shaders while you model is bonus. The only thing holding TrueSpace back here is that I'd like for it to have better capabilities for manually editing vertex normals. And of course more competitive UV tools, but UU3D solves that as well as game format translation.
For Architectural Viz work, the inference snapping tools are VERY useful and make laying out parallel geometry quick and easy. TruePlay and the customizable D3D shaders make for very nice realtime walkthroughs. The material library is a bit limited in the current version unless you know how to make your own shaders, but when Parva's Sky and Shader pack become available after 7.6 is released that will solve that problem. The lack of numerically accurate workspace side UV tools is a bit limiting though, and the TS VRay glass issue is a problem for current Arch Viz users that need to do detailed renders with caustics, transmission, and accurate reflections.
For product visualization, I think the problem is mainly the lack of spline based tools for workspace side, again the TS VRay glass bug, and the lack of layered materials for TS VRay. Other than that the modeling tools are very strong for prototyping.
For illustrators, the hair tools could be a bit better (I still thing polygon hair is a better approach though) and again the VRay glass bug is an issue.
For animators, the big problem is a lack of built in special effects. Particles, fluids, hair physics, etc. The character animation tools and scripting tools are very strong. The ability to blend keyframed tracks with BVH motion is very powerful and with morphing coming in 7.6, the only thing left is soft bodies. And as we've seen from Noko's demo vid in the Captain's blog cloth physics is already almost a "poor man's" soft-body simulator.
So I'd have to say you're probably right it's the animation features paired with the "real time render" that will make TrueSpace a very attractive package for creating animation shorts. Rendering out a 2 hour HD resolution movie in less than a day is CRAZY cool, and it brings video creation into the hands of the average person at a fairly affordable price. It's hard to believe that such an after though feature as umm.... gee... lets save the screen to file... :p would be such an innovation, but it is! :D I love the ability to use my video card to render out an animation to file in a matter of seconds.
But I still think the $600 price tag for TrueSpace prices it out of the reach of new hobbiests. One of the big things I've argued to make TrueSpace more competitive would be to make Lightworks an add on like VRay. Then Caligari wouldn't have to pay for a Lightworks license for each copy of TrueSpace and could probably lower the price. Just think how much more accessible Truespace would be if the base package was only say $299? And then the user had their choice of VRay or Lightworks for $299 each? With the free and stand alone render engines out there now, why should the user be stuck paying for a built in render engine that might not even meet their needs?
The other thing that TrueSpace needs is robust file import and export. With that and the above issues sorted, I think Truespace becomes a great well- rounded intermediate level app useful for a variety of industries and one that users can continue to grow with and useful for professional as well as hobbyist use. |
Post by Burnart // Apr 17, 2008, 1:04pm
|
Burnart
Total Posts: 839
|
As far as content and free or very cheap tools (ie DAZ) being the stepping stone into 3D goes it makes me think back to the late 70s early 80s when the home computer first got going - not only was the "serious" home user expected to build their own computer they were expected to write their own apps!
Long before I ever owned a computer I remember people complaining about how readily available graphics programs were taking the skill and fun out of it all.
Tying the software into broader marketing models is a sign of the times. Truespace's alignment with Virtual Earth is another such thing - we'll have to wait and see how it all works out. |
Post by Burnart // Apr 17, 2008, 1:18pm
|
Burnart
Total Posts: 839
|
The character animation tools and scripting tools are very strong.
The core boning/skeleton setup tools are quite nice - could do with a little bit of refining in terms of control of specific vertex assignment - the soft selection approach is a bit "smeary" compared to the select and smooth approach in Motion Studio. BVH motion attachment is easy - if you want to work with standard skeleton hierarchies. I'd still like to see some Joint Master type controls as standard on the workspace side for character animation before I would fully agree with your statement. Dragging and pulling into position is great for broad poses but sometimes you want an easy way to get in there and specify angles.
Collada import as well as export is something I definitely want to see - I can't understand how something can be an interchange format if software companies only implement export. |
Post by Jack Edwards // Apr 17, 2008, 1:49pm
|
Jack Edwards
Total Posts: 4062
|
I agree, sliders or scrubbers would be nice as well as manual entry. |
Post by moogaloonie // Apr 17, 2008, 2:18pm
|
moogaloonie
Total Posts: 64
|
I'd also like to mention that while I used Caligari Broadcast almost exclusively for nearly 6 years, I only did a handful of projects with versions 4 and 5 when I came back. I really only bought 7.5 as a vote of confidence in the re-write, DX based workspace, link editor and collaboration stuff. I played with it once or twice before doing an OS install and have just basically been waiting for 7.6 ever since.
I agree that seperating the external renderers to lower the price is a good idea, as I have zero (zilch, nada) interest in them. I think the future of independent CG will be decided by a company like Epic or Crytek selling a dedicated machinima studio, or maybe the maturation of some left-field product like Awakening. I'd like it to be TrueSpace, obviously, or I'd have just saved the money I spent on 7.5 for something else. |
Post by transient // Apr 17, 2008, 2:45pm
|
transient
Total Posts: 977
|
If Caligari wanted to join the content craze, they would be better off just starting their own content company geared around real-time content. Now that microsoft own them, this would be viable and would be win-win as the assets could presumably cross-over to xna.
Otherwise, they would just be stuffing money in the pockets of another company (Daz). This is why you don't see autodesk or any of the other major players falling over themselves to support the content industry.
That fact is Caligari have already surrendered to the Romulans, why would they now want to assimilate with the Borg?;) |
Post by Jack Edwards // Apr 17, 2008, 4:57pm
|
Jack Edwards
Total Posts: 4062
|
LOL Transient! BTW, I thought is was having been assimilated by the Borg, why surrender to the Romulans... ;p
@Burnart,
I did some playing around and I noticed that you can set joint rotations numerically using their matrix properties. It involves scripting though.
A more artist friendly approach is to key frame some joint positions, then you can go into the F-Curve editor and manually adjust the rotation key frames to any value you want. ;)
Also see trueBlue's script object here:
http://forums1.caligari.com/truespace/showthread.php?t=2051 |
Post by transient // Apr 17, 2008, 5:29pm
|
transient
Total Posts: 977
|
LOL Transient! BTW, I thought is was having been assimilated by the Borg, why surrender to the Romulans... ;p
I guess it depends on whether having pointy ears and a bad haircut is preferable to being a nigh invincible (if a tad docile) cyborg. I've always liked the romulans, despite their fashion faux pars.
After thinking a bit more about this thread, I'd also say that Jack's idea of making truespace more modular has a lot going for it. Especially if it made it cheaper to start with. At a decent price you could probably market 7.6+ on it's promising animation tools and modeling capabilities (not to mention the physics etc.), especially with decent importers/exporters to other renderers and packages.
It would also free Caligari to focus on the core aspects of the program like design and bug-fixes.
When you look at other apps like 3dstudio, they are often used as a shell that support other superior software programs, especially renderers, but also modeling and animation tools.
The old truespace (6.6), which has surprised me so much, has this quality. The third-party add ons are really what has sold me on it, as well as the interface which I'm really liking, now that I get it. |
Post by Burnart // Apr 17, 2008, 6:43pm
|
Burnart
Total Posts: 839
|
I did some playing around and I noticed that you can set joint rotations numerically using their matrix properties. It involves scripting though.
A more artist friendly approach is to key frame some joint positions, then you can go into the F-Curve editor and manually adjust the rotation key frames to any value you want. ;)
Also see trueBlue's script object here:
http://forums1.caligari.com/truespace/showthread.php?t=2051
???!!!!
Like I said Joint Master type controls ;)
I'll check out trueBlue's script object. |
Post by Burnart // Apr 17, 2008, 7:21pm
|
Burnart
Total Posts: 839
|
re. The modular approach - I'm not against the idea of providing a core app and then buying extra relevant modules. What constitutes the core app would need careful consideration. Would all the nifty physics and anim tools be core or hived off to a module? What price the core app? free? under US$100? etc
I am against VRay being the only option for a render module because its too expensive - and I gather the current tS version is limited compared to the max version and has issues anyway. If VRay was fully functional and equivalent to the max version then surely the price would double to be on par with max VRay pricing? Ports to freeware renderers are the go or having a lightworks module at a reasonable rate. |
Post by transient // Apr 17, 2008, 7:51pm
|
transient
Total Posts: 977
|
If VRay was fully functional and equivalent to the max version then surely the price would double to be on par with max VRay pricing?
It's not too far from being close to other instances. All they have to do is enable a few more options which I would bet are already part of the license they already have.
The fact is that vray, even at full price, could be good for truespace.
There are some good free renderers out there, but nothing on a par with vray, maxwell, final render etc. imo. |
Post by Burnart // Apr 17, 2008, 8:16pm
|
Burnart
Total Posts: 839
|
I'm not disputing that VRay when fully implemented is a wonderful tool. I would be greatly pleased for the tS community if all issues with it were resolved and yet it stayed at the US$299 price. The truth is though, that even at that price I would never be able to buy - its just too expensive for me especially when updating the core program, trueSpace, probably means updating the VRay module as well. For a hobbiest like me its a bridge too far. VRay for pros and well financed enthusiasts but a budget option also please. |
Post by v3rd3 // Apr 17, 2008, 9:26pm
|
v3rd3
Total Posts: 388
|
I like the suggestion of hooking into the open source/freeware rendering engines. I still use Virtualight ... or is that I am still figuring out Virtualight.
This type of facility would be terrific for those wishing to use them. The only problem is that open source denies potential revenue to the mother ship.
I think a lower priced highly functional TS with integration potential to the open source/freeware render engines coupled with more capable import/export features would be the right product at the right time to garner a boatload of TS enthusiasts.
I know the participants of this forum would certainly welcome a large influx of new people.
It could be the big shot to get TS mentioned in the same breath as ... ahem... tools with larger customer bases. |
Post by transient // Apr 17, 2008, 10:11pm
|
transient
Total Posts: 977
|
Yes, this is definitely a potential win-win situation. There's a growing number of users that are using 7.5 exclusively for real-time work.
With no renderer in the core product, it may promote competition among third-party developers. I would happily pay for a decent indigo exporter. |
Post by 3dfrog // Apr 18, 2008, 5:52am
|
3dfrog
Total Posts: 1225
|
Here is a little video displaying numerical joint rotation. You see on the small simple skeleton there are no limits on the joints and you can numerically set them as you wish. On tank girl there are limits imposed and it is difficult to set rotation. It would be good if you could set them with local coordinates numerically I think. Currently if you try to rotate on one axis but it is limited on another axis it will stop it because the numerical rotation is not local. If I understand it correctly. |
Post by Jack Edwards // Apr 18, 2008, 5:56am
|
Jack Edwards
Total Posts: 4062
|
Thanks 3dFrog :D
So the feature *is* there, but looks like it can be improved. |
Post by moogaloonie // Apr 18, 2008, 8:32am
|
moogaloonie
Total Posts: 64
|
If Caligari wanted to join the content craze, they would be better off just starting their own content company geared around real-time content. Now that microsoft own them, this would be viable and would be win-win as the assets could presumably cross-over to xna.
Otherwise, they would just be stuffing money in the pockets of another company (Daz). This is why you don't see autodesk or any of the other major players falling over themselves to support the content industry.
That fact is Caligari have already surrendered to the Romulans, why would they now want to assimilate with the Borg?;)
Hmmm... That's kind of where I was headed with my thoughts. I brought up Daz attracting new artists because someone in another thread expressed worries that things like WOW and modding were catching the interest of people who'd otherwise be getting into 3D animation and possibly joining the TS community.
I don't think Caligari should get into the content business, but rather streamline some aspects of TS so that people looking at getting into or out of Poser/Studio have an attractive option in TS. Then maybe someone else could make money off of TS content. Can't hate Daz though, they are bringing people into CG, although mainly hobbyists of course. Advanced tools do not need to be intimidating to people who have simpler goals. Independent CG films just can't reasonably be done by one person these days, at the level of quality people expect. The content market doesn't just exist because of talentless amateurs, but also because of talented artists whose visions exceed what can be accomplished by one person working alone in a reasonable amount of time. (I saw this coming years ago while playing Resident Evil for Playstation and realising the car between me and the zombies was on a Lightrom disc of models I'd bought a few years before. If big companies outsource and buy brokered content, why are individuals still expected by so many to do everything from scratch?) |
Post by jamesmc // Apr 18, 2008, 8:53am
|
jamesmc
Total Posts: 2566
|
There are no impediments from bringing Poserites and Daz Dolls into trueSpace that I'm aware of. :D |
Post by moogaloonie // Apr 19, 2008, 10:34am
|
moogaloonie
Total Posts: 64
|
There are no impediments from bringing Poserites and Daz Dolls into trueSpace that I'm aware of. :D
So I've been waiting on 7.6 to add support for morph targets for nothing?
Still, that was only a small part of my original point... |
Post by jamesmc // Apr 19, 2008, 11:03am
|
jamesmc
Total Posts: 2566
|
So I've been waiting on 7.6 to add support for morph targets for nothing?
Still, that was only a small part of my original point...
I don't think you are confusing it, but the term "morph target" in modeling from what you are used to has a different primary function than "morph target" in the proposed tS7.6 vertex animation.
Vertex animation is what is being implemented in tS7.6 as I understand it.
In Poser and all 3D modeling applications, one can make Morph Targets so they can be adjusted later (aka transposition)
This can be done in Poser by dials and sliders. Some programs have "transposition" programs that will import Poser morph targets made by poser, complete with the dials and sliders.
If the latter is what you are asking, there is nothing currently or I suppose immediately in the future that would transpose morph targets from Poser or Daz characters.
Some modeling programs use Morph Targets as an area that assigns certain functions/properties so it can be adjusted later in animation or poses. trueSpace can do this(via scripting), but the amount of character animators here is very small.
Of course, if you are handy with scripting, you could write your own custom morph target transposer routines.
Note: Other software that creates morph targets can be used in trueSpace, but it's basically up to the user how to manipulate it; i.e. no programmed help. |
Post by moogaloonie // Apr 20, 2008, 8:39am
|
moogaloonie
Total Posts: 64
|
Oh, Ok. So vertex animation like is used in MD2 (I believe) models is what we are getting. So there still won't be a way to do morph target animation then, or is there already? This isn't about getting poser figures into TS, but rather whether I can continue to use the techniques I had to learn to use Poser. I always thought facial animation using bones looked a lot harder than using morphs and was surprised that that didn't seem easily doable. |
Post by jamesmc // Apr 20, 2008, 9:17am
|
jamesmc
Total Posts: 2566
|
Oh, Ok. So vertex animation like is used in MD2 (I believe) models is what we are getting. So there still won't be a way to do morph target animation then, or is there already? This isn't about getting poser figures into TS, but rather whether I can continue to use the techniques I had to learn to use Poser. I always thought facial animation using bones looked a lot harder than using morphs and was surprised that that didn't seem easily doable.
Just based on my "best guess" from what Roman showed.
One will probably be able to create their own morph target areas when modeling and later use vertex animation for expressions and movement functions.
It's always good to zone your object.
Edit: I don't know the capability of what is being released, but I'm guessing it's similar to morph animation, but on a specifically created object.
Whether or not, one can use it on modeled targets, I suppose it could with enough work. |
Post by Délé // Apr 20, 2008, 11:20am
|
Délé
Total Posts: 1374
|
You might want to watch the morph video that Prodigy made if you haven't already. Roman posted it in his blog a while back. It shows morphs being used with a character. You should be able to pull out quite a lot about how the morphs work from that video. ;)
http://forums1.caligari.com/truespace/showpost.php?p=62604&postcount=40 |
Post by jayr // Apr 20, 2008, 12:06pm
|
jayr
Total Posts: 1074
|
You might want to watch the morph video that Prodigy made if you haven't already. Roman posted it in his blog a while back. It shows morphs being used with a character. You should be able to pull out quite a lot about how the morphs work from that video. ;)
http://forums1.caligari.com/truespace/showpost.php?p=62604&postcount=40
what i liked in that one was the 'breathing' morph that just loops the character breathing in and out, it'll be great to have something like that available |
Post by jamesmc // Apr 20, 2008, 12:31pm
|
jamesmc
Total Posts: 2566
|
what i liked in that one was the 'breathing' morph that just loops the character breathing in and out, it'll be great to have something like that available
You can do that now, don't really need to morph anything using assigned targeting.
The ability to set targets, name them and provide controls for them was the basic context I was referring.
In some programs like Poser/Daz and content dolls, they have named morph targets that use scripting conventions that can be translated to other programs.
(i.e. the area around the mouth is modeled in a very specific pattern and assigned to a morph target or targets)
Because the attributes of the target is already known because of conventions, then transposition to other programs that accept those conventions can be transposed by sliders, wheels and entry input.
trueSpace doesn't have a design in any of its features to do that, although it could be done, but not without a lot of work.
The vertex animation of the morphing targets that I saw from Roman's post work pretty much like I have in other programs (I'm guessing they would)
In that, you create an object, give it sufficient number of symmetrical vertices, then you "paint" assign those vertices by selecting them, give them a name and then stretch or compress them to perform an action.
Once that is complete, then the named morph target can be assigned a name and if wished, a slider, wheel or entry input.
The slider or whatever mechanism is then used to perform the "at rest" position and then the farthest or closest most definition of the vertices stretching or compressing.
Like a smile from a closed mouth and back to a close mouth for one slider named "smile."
Or the forming of the lips into the letter "O" for another slider. |
Post by moogaloonie // Apr 23, 2008, 9:24am
|
moogaloonie
Total Posts: 64
|
Here are a few things I think are important to machinima and the content markets, and a couple ideas with regard to TS.
Machimima is about story telling, the use of game engines allowed the medium to develop as they became advanced enough to not distract from the story telling. In a sense they traded the things people now expect to see from ultra-real CG (soft bodies, hair, fluids, ultra real material and light modeling) for the freedom to quickly tell a story. The use of Halo content in Red vs. Blue isn't much different than an artist's choice to use Daz content in a web comic (or whatever that stuff gets used for aside from NVIATs :rolleyes:) as both reveal the need to move beyond creating the content and toward putting it to use.
Using a game engine forces limitations to some extent, but also offers certain freedoms, not the least of which is recording actions in real time as a player using the pre-defined motions and physical properties simulated by the game logic. With a game as cinematic as Halo, or even some obscure game like Super Runabout for the Dreamcast, the cinematic cameras and replay features almost beg to have a narrative overlaid on them.
I've heard the term Daz dolls a few times, and while it seems a little dismissive it does call to mind an idea I've always had about making content. That is to design and package it like a collectible figure. A good collectible figure will often have alternate accessories as a high quality digital figure may have alternate textures and a range of available accessories (though these are generally provided so that the figure can be altered to create a number of different characters). I like the way that the morphs, poses, materials etc. are part of the Daz figures and they can all (in theory) be loaded into any scene intact yet with a minimum of files. I think it would be nice if this were taken perhaps to the point of having everything compressed into one file, perhaps even with animations/actions that many programs could read, but this isn't likely to happen.
The realtime rendering TrueSpace offers is a start. Being able to set up a figure so that it can be dragged into any project is important. I think I'd like to see more modifiers and effects primitives. I'd someday like to be able to drop a barrel into a scene, drop an "exploding" modifier on it and then run a physics sim that causes the barrel to explode when contacted complete with audible boom, physical shockwave, fragments etc. It's probably doable already, but I don't think it would be as drag-and-drop as it would be in Poser or even the recent Carrara 6. I want figures to share actions that allow characters with different body proportions to adjust to unlevel, uneven or shifting ground. I think ideally, once the hard work has been done, any Joe Videogamer should be able to make a movie with the content as it is provided. There needs to be some kind of insulation between setting it up and making it do something (but not exactly in the way that Poser is toy-like to use but yet horribly hard to learn to make content for).
It's not that I don't think I could do most of these things in TrueSpace but rather I'm not sure if I can package them for sale and expect a buyer to be able to drop them in and use them with the ease they would expect coming from Daz/Poser. I could probably explain this better if I weren't trying to keep the specifics of my project under wraps. |
Post by Burnart // Apr 23, 2008, 4:06pm
|
Burnart
Total Posts: 839
|
Moogaloonie I think what you are suggesting is absolutely do-able according to my understanding of the way scripting objects can be used to control mesh objects. The only real problem at this stage is you have to do it all yourself. If some of the scripting gurus got together with ideas people such as you they could start creating "action paks" with ready mades and do-it-yourself kits. (I thought this kind of thing was what the Rube-Goldberg machine thread from a while ago was going to start creating.)
As an LE imbecile with a scripting understanding of zero I'm not the one to help - as much as I like the way you think. |
Post by moogaloonie // May 5, 2008, 10:53am
|
moogaloonie
Total Posts: 64
|
I've just watched the new videos and I am very, very pleased with the recent developments. The facial morph control (with Puppeteer like pose dots) is exactly the type of thing I had been wanting to see. So, it's not like I will just be dropping my existing models in, but it does mean that I will still be able to animate in a way that is natural to me. I am now very anxious to get my hands on 7.6. |
|
