FYI, long anination + smaller bucket size = sucess. If you are doing a long sequence render with a bucket size of 16 or less. Further render to image sequence, then if something blows after 2 day of cooking you can just pickup where you left off. Nerd

Fixing the displacement shadow bug would make me (almost) happy.

Attached Link: http://www.nstk.com

In the Render Options there is a setting called "Bucket Size" It effects how large an area Poser renders in one bite. The bigger the bite the faster it works, BUT the more likely it is to have memory problems (These are supposed to be fixed in SR 4, fingers crossed). So, when you are rendering a long animation, especially unattended you should set the bucket size to aroung 16. This will greatly reduce the chances of it render getting stuck. Nerd

"Like hair, breasts, belly, shoulders etc should be effected by gravity and motion. When the character is walking every time a foot hits the ground that should send a shock wave through the body, just like in the world." What your asking for is not available automatically even on systems like Max + Motionbuilder (a total cost of almost 5,000$). While it would be nice, and useful - it just plain isn't goign to happen in this price range... nor would many take the time to do the complex setup that woudl be neded for a less automatic system (liekt he ones Motionbuilder uses). Most people don't event ake the time to learn dynamic cloth - and Daz hasn't taken the time to learn to support it.

My wish list:

• Fix all extant memory leaks.
• Augment the cloth modeler's particle engine to allow all objects to have mass assigned to their vertices, spring constants to the inter-vertex edges, and similar force fields to the edge angles and possible edge torsion angles as in molecular mechanics. This would allow "body bounce" as mentioned by several posters. "Mass maps," "Softness Maps." and "Elasticity Maps" would be needed to parameterize bounciness and fleshiness.
• Evolve cloth simulator to allow much finer-grained cloth models to allow realistic draping, folding, and wrinkling, and to avoid the strange tearing of coarser cloth models.
• Evolve the walk engine into a scriptable motion and gesture engine. Perhaps using XML, users could write scripts to define complex behaviors on a scene by scene basis. Such XML scripts could reference other scripts to define simultaneous movements, facial expressions, and gestures within a single figure, as well as simultaneous movement of different figures. Time points could be set to ensure precise coincidence of motion.
• Improved physics engine for above.
• Provide a GNU/Linux version, e.g. use wxWindows or QT KDE or something for the GUI?
• Open source the basic Poser engine so that the above improvements can be made in a reasonable amount of time without Curious Labs or Poser users needing Hollywood blockbuster budgets. Curious Labs could make money selling additional content, behavior scripts, maps for bounciness and fleshiness for existing figures, training videos, and other products. If DAZ wants to invade Poser's turf, CL will have to invade DAZ's.
• Add compatibility with POV Ray, or better yet allow its use as some kind of a plug-in renderer. It is much better and more versatile than Poser's. POV Ray scene files are human-readable, and fantastic scenes can be hand-coded. It is easy to generate and manipulate elements with macros and external programs written in any programming language. If Poser scenes could be written in POV Ray format, a huge amount of resources would be made available. PoseRay can do a lot of that already.

"Open source the basic Poser engine so that the above improvements can be made in a reasonable amount of time without Curious Labs or Poser users needing Hollywood blockbuster budgets. Curious Labs could make money selling additional content, behavior scripts, maps for bounciness and fleshiness for existing figures, training videos, and other products."

Some of the other suggestions are interesting... but I don't see this one as practical or useful for CL.

Despite the protestations of much of the OpenSource community, the final result of something like this would be the complete and utter destruction of most of the value and competitive advantage that Curious Labs possesses.

The disadvantage (the complete loss of control over their intellectual property and the sum total of all the effort and investment it represents) will never be offset by any profit in selling maps and so on - Daz barely makes a living doing this kind of thing already, and the Poser market isn't nearly large enough for two such companies at the moment.

Curious Labs is in a good position at the moment to hang on to their intellectual property - they have the resources of e-frontier on board and their would be competitor is finding out that building a tool like this is much harder than they thought it would be. In other words, there isn't one good reason to give away the source code.

The theoretical idea that they would gain a large development base, let alone one that could make significant and useful contributions to the code base is not born out by most of the OpenSource projects in the world - the vast majority of them fail to garner any useful level of developer interaction and event he few that do usually get most of their core features from a development group no larger than the one they would have had commercially.

The case of Curious Labs is, to me, a great example of when NOT to OpenSource. They intellectual property has great value as it is mostly unique, the customer base is fairly large and retail distribution is in hand. They have recently acquired access to dramatically larger resources of both money and technical support (e-frontier) and the chance to mingle technology with a product that has significant features to add (Shade). Why would they give all that away now to become a company selling weight maps.

"Add compatibility with POV Ray, or better yet allow its use as some kind of a plug-in renderer. It is much better and more versatile than Poser's."

Before Firefly this might have been important, but Firefly is an extremely competent renderer - there isn't anything to gain by using POV.

Sorry, soulhunter, but POV has many features that poser is lacking, and their development staff isn't aware of. It handles radiosity much better, and there is no param of the media that you can't control. I won't go as far as to say textures are easy, but when you control every aspect of them, getting what you want may be time consuming, but not all that difficult. Making a front end that would interface the two might be the best thing CL could do. Letting poser do nothing but generate and pose characters, then leave the rest for POV, and 99.9% of my frustrations with poser would disappear. Most of my frustration comes from posers abysmally lousy light control, and a generally screwed up light system to begin with. Many programs already export to POV, and many are proud of it. It's probably the most versatile and powerful rendering engine available.

"It handles radiosity much better"

My mistake, it seems that recently POV has added a form of photon mapping and that is something that as far as I know Firefly doesn't have.

But I wouldn't trade that for the lack of micro poly displacement (POV doesn't mention it, but then there isn't a complete feature list I can find on the website). I can simulate radiosity easily enough... but displacement mapping is too cool to live without these days.

"It's probably the most versatile and powerful rendering engine available."

Surely you meant to qualify that with the words "free and open source" (even that isn't something I would grant with total ease). Clearly POV, while cool,  isn't really in the same league in either features or speed as the big engines (Renderman, Mental Ray, Brazil, VRay and so on).

I think you are confusing the engine and the interface. With the exception of Radiosity, I don't know a single feature POV supports that Firefly lacks.. and I know of at least one that apparently POV doesn't have that Firefly does. It's pretty much a toss up.

If you don't like the lighting in Poser that's certainly understandable - but don't think POV is the answer, because it isn't the rendering engine that is the problem. Slapping POV ray into it isn't going to change anything as very few Poser users will be bothered (nor should they) hand coding a scene.

While I am sure a Poser->POV converter is cool and interesting and useful, I just don't see it as something CL should spend their time on when they have both Firefly and potentially the excellent Shade renderers available to them.

Ummm, Well. Here's a quickie fer ya. to the left of the vertical column in the sorta middle,is the actual image. To the right is a reflective plane, nothing else. Mayadoll rendered "no_image" the shiny post rendered "no_reflection" the sphere endered "no_shadow" Some pretty good reasons for mixing the two. Light is one point source, the "walls" and "floor" are only infinite planes, had them textured with nice stone, but the .jpeg was like 260k, so the textures had to go. Didn't use radiosity or media, at 1500 X 1200 it might be done tomorrow if I had.

I can come up with all sorts of ways to test the odd edges of render engines ... the issue is do they matter for the use most users will put them too :) Let me put this a simply as I can. Micropolygon displacement is far more important to me and those I work with than being able to selectively show or hide items in reflections. I am also not at all sure that it isn't something that can be done in Firefly :) However, if such abilities become important then I am sure they exist in the Shade engine :)

Ummm, sorry, but I didn't hide the reflection of Mayadoll, you see that clearly, and you see her shadow. It's her reflection that you see. ====================================== However, if such abilities become important then I am sure they exist in the Shade engine :) ======================================= If they are, the price of shade is going to be too much to pay for those three keywords. Working on another render, inside four walls with a ceiling on it, one point light, not a (giggle, snort) spotlight. "Full house" render, media, radiosity, maybe by tomorrow.

"her reflection that you see."

Ah, I see. Cute. Doesn't change anything :)

"If they are, the price of shade is going to be too much to pay for those three keywords."

But it won't be too much to pay for a good rendering engine that is superior to POV and comes with an actual user interface, full on modeling and animation functions and all the rest of it. WHiel the PD front ends for POV are interesting in their own way, like Blender they simply fall far short of what the commercial folks put out in even the mid range.

Hey, if POV does what you need then more power to ya, go forth render and multiply. It's a moderately competent render engine that while somewhat behind the current state of the art is still serviceable... especially for the price. You even have utilities that will let you use Poser files with it.

What I am unsure of is why it would be important or useful for Curious Labs to take on the support problems of having to write a converter and do customer support for it when they have access to better technology that they can control, as opposed to an open source project that would not in any way put much priority on their needs.

What, exactly, would be the business advantage of switching to POV? They lost capabilities that are are the moment in extremely high demand, they lose their ability to direct the development of the renderer and they have to then support it.

There is absolutely no upside for CL that I can see.

"Working on another render, inside four walls with a ceiling on it, one point light, not a (giggle, snort) spotlight. "Full house" render, media, radiosity, maybe by tomorrow."

I must be missing something, is this render (or the features you list) supposed to be an advocacy for POV? The features that POV has are far from unique or particularly impressive... and they fall far short of those included in the high end systems that are the next logical jump up from Shade and Firefly.

I LIKE POV. I played with it a long time ago and I still check in on it from time to time... but my urge to use it as the wonder tool for all things rendered faded when I saw what a real 3d integrated application can do for me.

A common problem of some Open Source advocates is that they fail to realize that it is rarely a good idea for a company to open all it's code and that in specialized areas with high R&D costs (like 3D) the OSS community simply isn't usually the best way to get the code done.

The one thing that stops most folks from using POV is the lack of a decent interface. Make that, an easy to use interface. Something simple, because most people like to think about what they are making, not how to do it. It is not a "better" engine than firefly -- it's just different. Each of them supports certain functions that the other does not, and, as noted, if they work, cool. I like Poseray, but I've used it less and less with what firefly offers me. I do a lot of special effects and stuff, and you know, the scripts to do what I do take forever to create, forever and a day to render, and still don't look as good as the stuff it takes me 15 minutes to setup in firefly or the old p4 renderer. The lighting, however, is much better. When someone creates an interface for POV that lets me use my 3000 pixel textures with a preview and the ability to modify them without typing in some pretty damned arcane math forumlae, I'll be thrilled (provided it can render faster -- POV is slower than firefly on the same images). And I have asked for a POV exporter. I think it would be cool. But do I want it in poser? No. Not at all, not in any way. As well, based on the licensing for it, neither do the folks developing POV, since CL is a for profit company, and really, really unlikely to opensource their hard work. Open source is not a viable business model for them at this time. If they open source the core aspects of Poser, they lose the ability to license it, since the programs primary file formats are entirely text-based. Training videos at this point would do them no good -- they can't lock the market there, and anyone willing to put in the time at present can learn about every single feature of poser 4 simply by reading whats been collected and created already. Invading DAZ's turf through open source isn't wise. DAZ is not making D|S open source. Claiming the higher ground will simply allow the competetion -- which is still developing at this point -- access to the very core elements that they are trying to mimic -- gee, does that make reverse engineering easier or what? About the only way right now CL will make money from content is if they hire the best and the brightest from the shops that currently exist. They have been trying to set up the system for it, but it seems folks really, really hate the uselessness of Content Paradise. You are right, they need to invade back, and, in true business wisdom, leverage what they do have already up so that they can gain a foothold in a market that they have not been involved with. This is pretty serious stuff, mind you -- it will require them to create stuff that is not just the same old same old and takes advantage of very specific features within Poser. This is Not easy to do when the bulk of the "names" aren't real fond of you, lol. side note: Content Paradise was the first step CL made. Was it before or after D|S? hee hee Also, I'll note that the lighting system in poser itself would not be fixed by POV incorporation -- anymore than it was by adding firefly. No, that's an issue that requires reworking internally within poser. And I agree -- the things I wish for most are lighting based. Oddly enough, firefly has a lot of lighting based aspects to it, and Shade (which, btw, has all of the features of POV and almost all of of firefly, and is better, and runs, in "pro form" about 1500 bucks) has the lighting system that I would like to see. (puts on thinking cap. Gee, licensing costs just went way down) lastly, just cause I gotta after all this, I wanna say that I really, really want a point light.

"Rebuild the code base from scratch you like you said you would for Poser5" Where did they say that? And who said it? Metacreations stated that, and they went off and sold all the parts off. I've never seen CL say that. Not once. I would like to see that.

One more try, and see if renderslowcity times out again. Rhino, Amapi and some others incorporate POV export, what's the problem? Once you have the scene file, minus lights in POV, adding a light is only clicking a menu item and telling it where to put it. If you want to change the color, no big thing, falloff and intensity are very easy to control. You can even add radiosity, media, fog, and many other effects with little or no problem. Some just a click, others, a click and change the few numbers. Doesn't matter, my rendering engine is POV, and poser is only a way to pose a figure, that I use damn seldom anyhow.

"Once you have the scene file, minus lights in POV, adding a light is only clicking a menu item and telling it where to put it." That sort of type/render/fix problem? I have better things to do with my time. I want previews of my lights as I place them... and light cameras to adjust falloff and so on. It works for you, thats cool... but it isn't a genrally useful solution :)

It works for you, thats cool... but it isn't a genrally useful solution :) ============================ It's far, far, better than what poser and firefly offer. Some here have done pretty well with posers lights, but I'm willing to bet that they took more time screwing with the arcade game dials to get them than it takes me to set up a light in POV. It doesn't take long, and one soon becomes aware of the general behavior of light, and setting falloff is something I usually only have to do once for each scene. IF it didn't work right, then it was the result of the way I was thinking of the light, and where I was wrong is evident before the render is even half done. With POV, I stop the render and start over, with poser, I have to stop the render, jump through all the hoops, control-alt-delete, run scandisk, then reboot poser. Something there, in either poser or firefly, is definitely screwed, and nobody is inclined to fix it.

" I have to stop the render, jump through all the hoops, control-alt-delete, run scandisk, then reboot poser." Then something is seriously wrong with yoru installation, and I would look into it further. Me? I hit "cancel", move the lights or adjust the params and re-render. But liek I said, if you seriousl consider a text file to be the most useful way to set up a complex 3D scene then your welcome to it :)

(Me? I hit "cancel", move the lights or adjust the params and re-render.) And then you wait until poser finally stops, and wait more until it will clear the screen so you can do something else. The problem is either with poser or firefly, everything else I have runs just fine. While you're at it, move a point light inside a closed building. Assign that light to come from any given object. I often assign the light sources to come from objects, such as light bulbs, candle flames, lamps, but then, one would expect the light to come from such objects. Neither poser or POV are going to replace the other, I can't see any reason that poser can't support a POV output as many of the much more expensive and capable programs do. Were poser to offer such a thing as a point source light, I might find it useful for more than just posing figures. as it is, I do not.

"While you're at it, move a point light inside a closed building. Assign that light to come from any given object. I often assign the light sources to come from objects, such as light bulbs, candle flames, lamps, but then, one would expect the light to come from such objects."

The issue is not whether sometimes abilities are needed that Poser doesn't have - there are. The issue is this sort of odd belief that POV ray is a significant rendering engine for Curious Labs to support.

It isn't.

Much better for CL to continue working with Reiss-Studio to improve support for serious applications that their users can get some mileage out of. Long before it is important to give Poser users the ability to start mucking with text files it makes sense to let them use Max, or Carrara or Shade or Lightwave. The renderers are better than POV and the interfaces beat the heck out of a text file.

Had the suggestion simply been "help support the Poser->POV scene conversion project" I would not have objected. But it wasn't, it was this statement...

"It's probably the most versatile and powerful rendering engine available."

POV is many things, but it isn't even close to "the most versatile and powerful" rendering engine. Not even in the same league really as the big players in that technology. That's my contention here.

That and the other suggestion that somehow CL would be better off open sourcing their codebase.

Would point lights be good? Yup. IS Open Sourcing Poser and using POV as the renderer the answer? Heck no. Working to support the Shade renderer and the Shade interface into the Poser workflow is the right way to do this for CL.

Open source poser? I don't think so, just some way of allowing others to write plug-ins other than "poser Python". If some of the higher buck programs can boast that they support POV-Ray output, what is so strange about wanting the bottom of the line program to do it? Would Rhino support it if there wasn't a good reason? Or Amapi? I don't think so.