What I want is something where the preview mode looks the same as the final render. I don't want to make 50 full quality renders to setup a scene !!!

What I want is that in preview mode the lights illuniate in the same way and intensity as in the final render including the shadows and I want the preview mode in real time. The only difference between preview and final rendering must be size, quality and detail and not the scene to look something different.

Beside this, what about something that solves the Navier-Stokes equations ?

aeilkema good post.

about the poser users. if they dont want change then this means that they are happy with ther curent poser version? right? so a new poser software will not hurt them.

i think the poser community is very specific. you have on the whole world sortware users who are complaining because they dont get new technology in their software. with poser users its different. they are complaining because they are afraid that they will get something new.

Or they would rather have improvements on the current renderer, improved animation tools, improved setup tools, improved clothing simulator, improved hair room, improved user interface or one of many other new features which would improve workflow.

For them a new "more realistic" renderer is probably much lower on the "wanted" list

wimvdb its pointless in 2011 to add imporvements to such an old renderer. things changed with CGI.

its like trying to make an old car better for todays standards.

Quote - Or they would rather have improvements on the current renderer, improved animation tools, improved setup tools, improved clothing simulator, improved hair room, improved user interface or one of many other new features which would improve workflow.

For them a new "more realistic" renderer is probably much lower on the "wanted" list

 

its not even about realism. realism comes automatic with a new renderer because of the year 2011.  

a new renderer would be easier and more simple to use. which is what poser users want.

You did not get my point. For most users it is low on the priority list. They much rather have other improvements in Poser.

Quote - wimvdb its pointless in 2011 to add imporvements to such an old renderer. things changed with CGI.

 

its like trying to make an old car better for todays standards.

You're wrong there. But I'm not allowed to tell you why.

Quote - its like trying to make an old car better for todays standards.

Which people do all the time.

@Iceboy instead of beating your head against the wall on this issue, and since your Focus seem to be getting better renders, perhaps it is time for you to just invest in one of the MANY alternatives  to rendering directly within poser.

I have been using poser since "Fractal Design's Poser2"
and I can Say that I never Did What Iwould call a final render within poser itself.
 it was first Meta C's Bryce 2 and so on and so on up until today where where I am using C4D + Vray or the base version of Vue depending on the look I want.

Vue has a low cost option
the new Pose2LuX  seems to be coming along nicely
(Congrats  in order there!!!)

I dont see the point of trying to browbeat SM into giving poser some vray+SSS, quality ray tracer.

The harsh reality of it is
that if you truly want to Achieve exactly what you want with CG
you will have to deploy MORE than ONE program to do so
or be willing to compromise  on your desired results.
and this is not just a Low end "hobbiest " Dilemma

There is a reason why MAYA owners are still buying Zbrush
or Motionbuilder
There is a reason  Why us C4D Owners have Swallowed hard and Sprung for RealFlow or Vray

and trust me there are plenty of threads at CG talk Demanding better modeling tools for C4D along with other gripes about C4D's Shortcomings yet MAXON just reported its most
profitable year ever just as MODO continues to sell quite well
to C4D users.

Cheers

Quote - You did not get my point. For most users it is low on the priority list. They much rather have other improvements in Poser.

poser users also have a big problem with new skinning for figures. weight painting is a BIG NO for them .  

looks like they dont like changes.

Quote - > Quote - wimvdb its pointless in 2011 to add imporvements to such an old renderer. things changed with CGI.

 

its like trying to make an old car better for todays standards.

You're wrong there. But I'm not allowed to tell you why.

when it comes to software you know everything million times better then me. but i will still writte what  i will now. its easier and better to create a new path raytracer for Poser.

i will not belive that its possible to make Firefly good enough for todays computer.

Quote - @Iceboy instead of beating your head against the wall on this issue, and since your Focus seem to be getting better renders, perhaps it is time for you to just invest in one of the MANY alternatives  to rendering directly within poser.

I have been using poser since "Fractal Design's Poser2"
and I can Say that I never Did What Iwould call a final render within poser itself.
 it was first Meta C's Bryce 2 and so on and so on up until today where where I am using C4D + Vray or the base version of Vue depending on the look I want.

Vue has a low cost option
the new Pose2LuX  seems to be coming along nicely
(Congrats  in order there!!!)

I dont see the point of trying to browbeat SM into giving poser some vray+SSS, quality ray tracer.

The harsh reality of it is
that if you truly want to Achieve exactly what you want with CG
you will have to deploy MORE than ONE program to do so
or be willing to compromise  on your desired results.
and this is not just a Low end "hobbiest " Dilemma

There is a reason why MAYA owners are still buying Zbrush
or Motionbuilder
There is a reason  Why us C4D Owners have Swallowed hard and Sprung for RealFlow or Vray

and trust me there are plenty of threads at CG talk Demanding better modeling tools for C4D along with other gripes about C4D's Shortcomings yet MAXON just reported its most
profitable year ever just as MODO continues to sell quite well
to C4D users.

thanks for this post.  

i actually  was not hoping for a Vray renderer.they spend years working on that.and it needs to be production prooven.  i am hoping for a simple raytracer with simple code . since poser users use somewhere 2 textures for a figure this shouldnt be a problem IMO.

Is it just me or is there a lot of wind in this thread? Tilt! :tt2:

Quote - > Quote - You did not get my point. For most users it is low on the priority list. They much rather have other improvements in Poser.

poser users also have a big problem with new skinning for figures. weight painting is a BIG NO for them .  

looks like they dont like changes.

Most poser users do not rig, so they couldn't care less if they have weight painting or not. They only want better bending. If that is what weight painting gives them, they will gladly take it.

It has no impact on them whatsoever since they only use and NOT create items

Everyone likes changes which are an improvement. Unfortunately some changes which improve one thing make other things worse. So it depends on what they like more.

You can not generalize users like you do.

Done***..***

Cheers***

Well played, wolf359!

:thumbupboth:

The original Bryce is indeed the archetypical ray tracer showing what shiny spheres hovering over checkerboards look like.

Random thoughts...

C4D's render engine is in it's 3rd iteration. When Daz got 3Delight, it had been around for years & already used in production. If you look at VRay, Vue, etc., they are all either the product of years of development by companies that are either dedicated to building render engines, or firms that have a goodly amount of development resources to devote to the process. FireFly (IIRC) was bought - from Pixels3D, the same way the hair, cloth and face rooms were bought. While it's certainly been around a while, I'm not sure that it's a good comparison, i.e. one developer, working for different companies, with perhaps different priorities, compared to the others. Maybe it's fatally flawed or behind the times, or maybe it just hasn't had the resources given to it in order to develop as fast as it might have.

The choices as I see them are:

1. Continue to develop FireFly, building on the current base.
2. Start from scratch, tailoring a new engine to the current material system.
3. Start over and develop a new materials system and render engine.
4. Buy/license a new engine and material system.

I don't pretend to know which would be best. A new material system (not just the interface) would involve a lot of pain for users & content creators. You could maintain backward compatibility but that has its costs - probably why they dropped the P4 engine. Trying to do anything with only one developer, even a great one, is probably going to take longer than people would like.

The GPU revolution seems well underway so any choice needs to take that into account. Even folks who render 'simple' scenes in terms of materials may want to render scenes with lots of figures and GPU acceleration would be welcome. With even browsers incorporating hardware acceleration these days, I'd say it's going to be a requirement for any credible 3D application.

I think the V5 issue is somewhat misstated here. I don't think that anyone doesn't want figures that pose more naturally; some simply balked at the idea of having to buy a new version of the program to get it.

As Aeilkema said, there are a lot of (sometimes conflicting) priorities they have to look at. The release of Poser Debut seems to say that they're trying to hit all the marketing niches. Maybe that's good, maybe not. Poser has been lucky to have only one real competitor, but having that competitor being your 'frenemy' certainly complicates matters.

Whatever they come up with, I gotta see skin before I'm impressed. It's all about Vicky. I really don't care if the temple is the photo-spitting image of Mayan architecture if she doesn't look luscious - YMMV :-)

I think one of the major problems with Poser is (and has almost always been) the constant lack of communication. DAZ has changed their policy somewhat on that and is communicating much more on what they are doing and often surveying their user base how to go ahead and scanning their wishes. They put out information on what's comming (as much as they seem fit to do) and have open beta participation programmes. Although no one knows what DS4 will completely be like, a rough sketch is already known and details are filled in and showed along the way. By now DAZ has build up a good community / user base that's involved in the development of D/S, openly suggesting features and improvements and DAZ taking them to heart.

With SM no one of us has a clue what they're even up to and those who know are not allowed to tell. SM does not communicate with their customers at all and closed all the communication venues on purpose. By the time the user gets to know what the next poser will be like, all is set in stone and we can either like it or not.

I know we've got the once a year poser survey, but who knows if they even listen to it at all or it's just courtesy, making us believe we involved somewhat. If you don't publish survey results, no one knows if you've listened or not.

There's no doubt in my mind Poser would grow a lot more if SM would change their policy on non-customer interaction. They could have build up (or somewhat taken over) an active user base, but refuse to do so and we see the results of it. The DAZ community is growing, while the Poser community is decreasing and less active. Most people I know from way back have either moved on, given up on Poser (some died which is sad), but most of them are over at DAZ, abandoning Poser. Some are still around though...... Communcation has always been a problem for the owners of Poser, I don't know why, but they seem to have a real problem with it.

It's fun to discuss something amongst ourselves, but it would be really meaningful if SM joined in.

stewer is that you ?

http://forums.cgsociety.org/showthread.php?f=59&t=974578&page=4&pp=15

ice-boy,

Of course it's him. Did you not notice that under his avatar in CGSociety forum, in every post he makes, it says:

Stefan Werner

Senior Software Engineer

Smith Micro

Germany

Sometimes I wonder about you.

"Whatever they come up with, I gotta see skin before I'm impressed. It's all about Vicky. I really don't care if the temple is the photo-spitting image of Mayan architecture if she doesn't look luscious"

Interesting ,While one can Appreciate your honesty on the matter of vicky
I find it  rather sad that both programs (DS&Poser) have been reduced somewhat
to being a platform for ONE Naked Female model.

Daz has the advantage of having been able to build their entire program& Content regime around ubiquitous V-Chick and thus can introduce modern industry standard innovation going forward
While SM plays the part of the poor cousin trying to stay compatible while clinging to vestigial, cobbled together legacy code from the 1990's.

I remember one person saying how he was ready to move on to Autodesk 3D MAX
for better...well everything
but alas the on thing that held him back was losing a fully functional V-Chick.

One can only speculate on where DS&Poser might be today absent the dominance of this ONE model and here loyal worshippers

lastly I think posers node based Material System has not been the boon many thought it would be in fact I personally think that node based mat rooms in general are becoming less & less relevant in light of the new breeds  of GPU/CPUrender engines and their preset material structures.

Cheers*** ***

Quote - nteresting ,While one can Appreciate your honesty on the matter of vicky
I find it  rather sad that both programs (DS&Poser) have been reduced somewhat
to being a platform for ONE Naked Female model.

DS never had a different reputation, imo, and Poser was reduced as soon as DAZ grew bigger. It is sad, I don't blame the applications, but DAZ's marketing strategy for this. They've given Poser a certain name and once that stuck it automatically transferred to D/S once it was released. DAZ does a lot of good, but they sure don't give D/S and Poser a good reputation in the 3D world.

"Daz has the advantage of having been able to build their entire program & content regime around ubiquitous V-Chick and thus can introduce modern industry standard innovation going forward"

Except they don't introduce industry standard features and precious little innovation. The features of DS are proprietary, expensive to implement, on the back-end not front edge of industry innovation, and are not even fully compatable with the other programs that they sell. Their subdivision is anything but industry standard. Their dynamic cloth is anything but industry standard. DS is proprietary by intention, not necessity. But, they have stuck so close to Poser that SM doesn't even have to put much effort into making their own content. And, they are unlikely to introduce any new content that will exclude Poser users (hence the plausible rumors that SM and DAZ are in communication about their plans for future development, and neither companies are denying it).

Any future version of DAZ studio will be fully compatable with all of their previous content. Who really thinks that the next version of DS will only work with V5 and no other content? Does anyone believe that users of new versions of DS will be unable to open or convert their former scenes?

"While SM plays the part of the poor cousin trying to stay compatible while clinging to vestigial, cobbled together legacy code from the 1990's."

Absolutely not so. The core code of Poser 8 has been completely rewritten using "modern" programming standards. Hence the ability of PoserPro 2010 to run in 64 bit mode. The interface consistant with previous versions is cosmetic not systemic. Compatability with previous features (Which DAZ will certainly maintain also) does not mean that the new versions of Poser are not possitioned to advance with technology, and implement features that users want. In fact Poser has implemented many features in their product line that DS has not yet added.

The "rivalry" between the two companies benefits both, and I welcome it. It is far more likely that the development that WE want to see in our prefered programs and content will occure if there is competion between DAZ and SM. If one program ever lags seriously behind the other, than the other will have little reason to invest in improvements.

SM sells a wide variety of popular software applications and content, and they sell thru multiple popular outlets. They do not buy software that does not have a future, and they have the capital to invest in the development of their product lines. SM actually has the resources to buy DAZ if they want it. They chose Poser.

The material node system is powerful, and does not need to be replaced, but improved. Poser would rival anything on the market if the program had nodes that could be programmed with its Python script, if the Python script was allowed to call external libraries and programs, if the features and capabilites of Poser were fully exposed to its Python script, and if there was the capability to link directly to other rendering engines from inside Poser. With these tools innovative programmers could add feature after feature to the program.

I don't use it myself, but how about a version of either program for Linux? That would open up some serious market growth.

DAZ Studio is a great application, and can be highly recommended to people looking for a program with its features. But, it is not enhanced in the slightest by underestimating or belittling Poser. If Poser did not exist, there would be another program on the market in its slot, because no application will satisfy everyone. Far better if DS and Poser concentrate on improving what they do well, and implementing what their customers want. Let people migrate to and from these platforms as they will, there are still over 6 billion people not using either program so I do not think the potential market is completely saturated.

lmk

"Let people migrate to and from these platforms as they will, there are still over 6 billion people not using either program so I do not think the potential market is completely saturated."

All of the human population can not realisticly be assumed to be a potential market for any commercial product that is not Eaten,lived in, or worn on the body for protection& Warmth so yes market saturation for a software program is quite possible considering the limited use of such a product.

Cheers

"All of the human population can not realisticly be assumed to be a potential market for any commercial product that is not Eaten,lived in, or worn on the body for protection& Warmth."

LOL! Of course not, hyperbole is rhetorical not logical.

"so yes market saturation for a software program is quite possible considering the limited use of such a product."

That might be for software used to calculate the number of coils of wire to wrap on a transformer, or to develop a pattern for a Singer embroidery machine. Certainly not for applications like word processors. When speaking in the billions that is not likely for DS and Poser products. Even in the USA, the average population is growing daily more sophisticated in terms of their use of technology (computers).

Among the six billion people made our neighbors by the Internet, the potential market of folks that could be interested in DS or Poser dwarfs those who already know about these products. The increasingly worldwide nature of this market is well demonstrated in this forum and marketplace (where our international friends charitably put up with using English, almost exclusively). Both DAZ and SM are well aware of local and global trends, and show every indication of actively positioning themselves to attract (a lot) more users, not just satisfying those of us who already use their applications.

But, whatever the future does bring, I hope the up-coming releases from DAZ and SM will contain a lot of interesting features. They will be sure to stimulate some lively conversations after the fact. Seeing that actual firm information about those features is so closely guarded, for now, speculative discussions like this are about all we have to bid the time (except rendering).

Cheers in return,

lmk

Quote - LOL! Of course not, hyperbole is rhetorical not logical.

Quote of the day!  :thumbupboth:  :thumbupboth:

"I don't use it myself, but how about a version of either program for Linux? That would open up some serious market growth."

The only thing that keeps this ol' girl in Windows is Poser. If there were a Linux version, I'd willingly go back to the GIMP (from PS3) for most of my 2D stuff and be done with Windows nonsense. "Networking with credentials is only available for Win7 Pro..." pshaw! What utter rubbish! We had it in WinXP! Now I have to log in to my desktop from my laptop every morning simply because the "Remember network Credentials" tickbox: doesn't.

"must look into VMWare - I really must"

Forgot you were in the house Wolf or I wouldn't have mentioned your bête noire. Actually I would have, but the point was that  Poser is about people IMO, not architecture or cubist neo-realism. Those things are all well and good and I want them to look good, but at the end of the day, I want Vicky, Aiko, Stephanie, Posette etc. etc., ad epiderminum to look as real as possible. That is a technical challenge unto itself which products such as SkinVue for Vue, vReel Skin for C4D & MentalRay's Fast Skin strive to achieve. Creating some new 'raytracer' without taking that into account is, again IMO, foolish.

Scrabble points for working in vestigial again but you'll never top your own 'vestigial versions' :-)

Quote - stewer is that you ?

http://forums.cgsociety.org/showthread.php?f=59&t=974578&page=4&pp=15

It's me, although the photo is a few years old - I never changed it since signing up at cgsociety. As a reminder, whatever I write there or here at Renderosity is 100% my opinion and I do not speak for Smith Micro. So please do not try to interpret anything about the future direction of Poser from this post here, it's all just my very own personal view.

My thoughts on a few things mentioned in this thread:

Software modernization: Software is not like a car and not like a building. I can rip out the very foundation of a program and replace it with something else and still have a lot of the upper level code intact. Also, code doesn't age, rust or deteriorate - perfectly working code written 30 years ago is still perfectly working code today. Best example: Unix. A system that was designed decades ago for room-sized machines with text terminals is now playing videos and 3d games on my cell phone. There are many dramatic changes that a good programmer can do to existing software without having to throw out the baby with the bathwater.

Ra tracing: There's nothing revolutionary or new with ray tracing or path tracing, both have been around for decades. The main reason why they're getting more attention lately is faster hardware that reduced path tracing render times from weeks to hours. Poser uses a ray tracer for things like reflections or shadows since verison 5 and expanded its use to AO and IDL, in 6 and 8 while making numerous improvements to the tray tracer itself on the way.

Unbiased: Doesn't mean what most people think or what marketing departments want you to think. It doesn't mean physically correct, it doesn't mean path tracing, it doesn't mean artifact free and it doesn't mean converging to the correct result. I could put a physically based path tracer in front of you and pretty much the only way to tell whether or not its biased would be studying its source code. In fact, most renderers that advertise being unbiased are biased anyway. (Does their max bounce option go to infinity - real infinity, not just 2^32?)

GPU acceleration: cool stuff, I love how much computing power we can get these days for so little money. However, there are severe limitations: CUDA and OpenCL don't do more than 128 textures, and textures have to fit into main memory in addition to geometry and code. Before Poser 7, textures were what made many scenes run out of memory when rendering, but the current version of FireFly can handle pretty much any number of textures in any size using less than 100MB RAM for them. I'm sure some day we'll be able to use texture and geometry caching for GPU ray tracing, but at this point, I haven't seen any practical solution for this problem.

There were a number of things thrown in together in this thread that are separate - physical shaders, ray tracing, progressive rendering and path tracing. Shading models like the one used in Cycles don't depend on ray tracing, if you look at the other sessions in that SIGGRAPH 2010 course, they are being used in games too. Also, changing a renderer from REYES to ray tracing will not add global illumination or change the way its shading works (you can try it out with 3Delight if you like, the Hider option: the render will look the same as before).

Don't get me wrong, I'm not saying that I'm against any of those things and I'm impressed as well by all the cool things that keep coming out. It's just that you don't go from a simple demo to a robust production engine in just a few months - for example, the Arnold renderer mentioned in the SIGGRAPH presentation that ice-boy linked has been in the works since 1997.

http://www.graphicall.org/106

change the shading mode to ''rendered''. and Cycles will change the rendering on the fly.

maybe in the future a connection with Cycles and poser ?

interesting thread on the Poser renderer, again.

Between the lines, some people were reporting memory issues. Poser 8 is a 32bit program which can address up to 2Gb user memory under normal conditions. You can boost this to 3Gb, I did a tutorial (in the Rendo Tutorial area) to tell you how. P8 does not do so automatically in a 64-bit environment.

Rendering as a seperate process saves you from the Poser-overhead, giving a bit more room as well. Don't forget to make the FFRender.exe 3Gb-aware as well, then !

happy rendering.

BTW, I agree on Steves analysis on GPU benefits. Luxrender uses GPU for just a few routines in the process, while MAX as well as Octane and 3DCoat use GPU mainly for speeding up the interactive work in the viewports. In the meantime, one can do raytracing with FireFly (just set the materials accordingly, this was exactly what the REYES engine is about) and it is faster than Bryce. So, what's up?

What's up is that soft/blurred reflections are missing in Poser. That's the look characteristic of the unbiased renderers that people notice is lacking Poser. The other aspect is how easily the many unbiased renderers correctly produce the Fresnel effect, whereas in Poser you have to at least build the Schlick approximation yourself and that's 4-5 nodes more you have to work into a shader. Most users can't/won't do it. When they render with, for example, LuxRender, they marvel at the realism of the surfaces. It's just those two things.

The biggest thing that bugs me about Poser is the materials (sorry BB). There is little to no documentation on any of the nodes, what they do or how to use them. Most of us go into the material room long enough to cry and leave again. The preview is minimally useful at best - in order to really see what that great material you just made looks like, you have to render it. 99 times out of a hundred it looks nothing like the preview or even close to what you were hoping for. Glass is a joke - can't even make it transparent, no caustics. Reflections are slightly better, but not great either. And there's no SSS. For folks who render primarly people and skin, this is important ;).

Lighting is another thing. Not that it hasn't improved. It really has. And things like BB's envirosphere and the addition of IDL were very welcome, don't get me wrong. VSS has also been a godsend for those of us who are material room deficient. Even taking all that, there isn't much that I would miss about the material room or the Firefly renderer if it were to go bye-bye ;).

edit: I would really miss dynamic hair and cloth tho...lolol. I wish they would improve Poser in those areas.

Laurie

No need to apologize to me when noting that nodes don't help people. I don't need nodes, I don't want them, and I don't use them. I write shaders as what they are - software. My matmatic converts them into the language of nodes, not because that is desirable, but because it's all we're offered. I have to do it that way. I'd much prefer to just write the shader as a C++ component and give it to you as a fait accompli.

The premise of nodes is you have the freedom to make it do anything, with training wheels so you can't fall down. You may not get what you want, but it will never crash. That's not true of free-form software components. And the sad thing is the training wheels limit you. You can't actually build any sort of material with Poser nodes, because some things are just impossible to say. The language of the nodes is too restrictive. There are no light loops, I can't directly get at certain key things like launching my own rays, so a good number of physical phenomena simply can't be implemented as is.

The reality is that a physically correct shader is a lot of math, and the math should be left to developers, not artists. (Unless you're Da Vinci - then you do both.)

This is the belief system shown to work for LuxRender et al. A handful of brainiacs produce a new built-in shader, and the rest of the people have no idea how it works, but they can plug in the color map, dial in the "shine", and it looks great.

I would not like to see Poser become "closed" like that - some dynamic component assembly is great. What I would like is for me, a software developer, to produce a new "Poser Surface" - or maybe 100 new ones - and skip all the nodes just to create the fundamental nature of the material - glass, metal, plastic, etc. The nodes are very cool for patterns and such, but not cool at all for the actual interaction with light.

Quote - stewer and bagginbsill.  try out a windows build of Cycles.

http://www.graphicall.org/106

Thanks, but I already compiled my own in Ubuntu.

The other problem with nodes is familiar to everybody. 

1. Spaghetti

2. No composites

In my previous job, I built a "node" based system for doing complex analytics against disparate data sources. It was very powerful but complexity overwhelmed people quickly. More than 50 nodes, and nobody understood how the analysis worked. And any common themes had to be built over again by hand in each solution.

In version 2, I introduced composite nodes. These were like integrated circuits. Users saw and used one node. Inside, there could be hundreds. They didn't need to see those. They just saw a few plugs on the outside. With this refinement, the node system took off. Today, there are hundreds of people using that system, and many of them are business people with no understanding of software development. Yet they "write" programs comprising 5000 or more steps in them in minutes.

Poser needs the same - libraries of composite nodes. Users who want glass would load the glass node - which is built from Shlick's approximation for Fresnel, and a handful of other nodes to mix and blend the reflections and refractions. All of this stuff inside, and you don't have to think about it or even look at it.

Quote - " I don't need nodes, I don't want them, and I don't use them. "

'

Thank You Sir!!
I continue to be Amazed at the insistance of some C4D users that Maxon implement
the mighty node based shader building system while at the same time drooling over the prebuilt shader libraries we have for Maxwell

Some are even demanding that Adobe make photoshop "node based"
while I under stand the true power of node based shaders I have never been personally interested  in  doing all that "wiring" myself .

Cheers

Quote - Users who want glass would load the glass node - which is built from Shlick's approximation for Fresnel, and a handful of other nodes to mix and blend the reflections and refractions. All of this stuff inside, and you don't have to think about it or even look at it.

It's similar to what we're trying to achieve with Pose2Lux actually. Prebuilt shaders that still use the texture maps of the Poser model or totally new ones that one can apply in the exporter. Obviously it's not perfected yet, but some of us are working on it behind the scenes. However, some adjustment by the end user with preview is planned for a future release.

I find it ironic too, that writing a shader for Luxrender by hand is easier for me to understand than using nodes in Poser. Go figure ;). Even LuxBlend makes more sense...lol.

edit: and yes, Snarlygribbly is a bit of a brainiac. That's a compliment...lol.

Laurie

Quote - It's similar to what we're trying to achieve with Pose2Lux actually. Prebuilt shaders that still use the texture maps of the Poser model or totally new ones that one can apply in the exporter. Obviously it's not perfected yet, but some of us are working on it behind the scenes. However, some adjustment by the end user with preview is planned for a future release.

I find it ironic too, that writing a shader for Luxrender by hand is easier for me to understand than using nodes in Poser. Go figure ;). Even LuxBlend makes more sense...lol.

edit: and yes, Snarlygribbly is a bit of a brainiac. That's a compliment...lol.

Laurie

We're lucky to have Snarly that is for darned sure.  Otherwise I would probably still be staring crosseyed at the .xml.  I can write a material by hand for P2L far faster than I could ever put one together for Poser and I don't need to worry about whether or not my math nodes are in the right order, just whether or not the .xml is formatted correctly, which with a good text editor is easy.

Which reminds me, my sanity break from skin is over, time to get back to work on it.

Quote - What's up is that soft/blurred reflections are missing in Poser. That's the look characteristic of the unbiased renderers that people notice is lacking Poser. The other aspect is how easily the many unbiased renderers correctly produce the Fresnel effect, whereas in Poser you have to at least build the Schlick approximation yourself and that's 4-5 nodes more you have to work into a shader. Most users can't/won't do it. When they render with, for example, LuxRender, they marvel at the realism of the surfaces. It's just those two things.

and from what i understand its hard to have fast soft/blurred reflectiosn in firefly because it is firefly.

Who said anything about fast? The unbiased renders are not fast. If Poser could do it in an hour or two I'd be fine with it. Trouble is it won't do it at all. And it isn't because it's Firefly. The same argument could have been applied to why we had no IDL, which we now have.

I would say more about this topic, but it would require that I discuss what is going into the next Poser and I'm not allowed. But the point you should realize is that the reason Firefly is limited is time and money - i.e. because Poser is cheap. It has nothing to do with Firefly technology.

i should have wrotte ''faster'' and not fast in general.

Quote - Who said anything about fast? The unbiased renders are not fast. If Poser could do it in an hour or two I'd be fine with it. Trouble is it won't do it at all. And it isn't because it's Firefly. The same argument could have been applied to why we had no IDL, which we now have.

I would say more about this topic, but it would require that I discuss what is going into the next Poser and I'm not allowed. But the point you should realize is that the reason Firefly is limited is time and money - i.e. because Poser is cheap. It has nothing to do with Firefly technology.

unbiased renders are not fast. but biased raytracers are super fast.

I would like to see a container node added to Poser, with a variable number of programmable input plugs, and even a variable number of programmable output plugs. (In Poser all of the nodes except the various Root Nodes have only one output plug. The Root nodes have no output plugs). Inside the (opened) container node, one could make as complex a node set-up as needed. Outside the (closed) container node it could look like the Root Node, with lines where values, or colors could be changed, and plugs for attaching spagetti connections. I would like the various channels/lines to be namable by the "programmer" for clarity. If one is going to go to the trouble of making something like this, go ahead and give it tabs or pages. (If I am allowed to really dream provide for internal connections between the tabs/pages).

I would like to see a node that can be programmed with the Python Script language built into Poser (and some lines/channels on the Root Nodes that can be programmed by the users using Python). I do not know enough about the possibilities to be able to dream of what some of the people in these forums could do with Python programmable nodes.

I would like to see most of the capabilities, functions, procedures, and variables exposed to Python, or available in a (reasonablly priced or free) SDK. I would like for Poser Python to be able to link to compiled libraries, and call external executable files.

I would like to see a nodes library. In the Advanced Material Room, right click in the nodes window, and have a library choice there under the New Nodes menu. I would like to save nodes there that I have set up. Even simple things would be a boon to me like a math node that makes an image negative. I just can't remember a lot of those things, and have to look them up time and time again. Why not name and save the node, and just add it already set up.

With additions like this, Poser could move from low-end to high-end very quickly, as people who can do the math make nodes that the rest of us can use. Well, as all can see, I have a very active fantasy life.  :)

lmk

lk, that's exactly what I built in my last job. Multiple inputs and outputs. The names were whatever the user wanted.

I also built things called "bundles" - these were composite wires. You could combine various wires in a bundle, and then you could connect all the wires in the bundle to other nodes by just dragging one fat wire. Inside, you unbundled to get the individual wires again.

There were also user-defined parameters that could be assigned to a composite. These parameter values were then available to configure the interior nodes. For example, you could set the scale on one top-level parameter, and inside it would alter the scale of all the internal nodes that require synchronized scaling.

There could be composites within composites, of course. And once a useful composite was built, it was a single click to publish it as a new library component.

The most cool thing about it was this: if a library component was updated (to fix a bug, or make it faster or whatever) then all the deployed copies in specific solutions were automatically updated as well. It was not a copy - but a reference.

I have several patent applications filed in the US and Europe on all this. Still waiting for the lawyers to sort it out.

Quote - ...The most cool thing about it was this: if a library component was updated (to fix a bug, or make it faster or whatever) then all the deployed copies in specific solutions were automatically updated as well. It was not a copy - but a reference...

I.e, a class? The more I work them, the more impressed I am with the concept!

It would be geniunely brilliant if, in one of the next sorties of Poser, matmatic could somehow be integrated and one could write "composites", so no nodes as we traditionally know them would be used or could be seen at all: they'd be like part of a specific PoserSurface, maybe.

And no, I'm not hinting you comment on the likelihood of something like this anytime soon, BB - I do respect non-disclosure agreements.

Quote - The other aspect is how easily the many unbiased renderers correctly produce the Fresnel effect, whereas in Poser you have to at least build the Schlick approximation yourself and that's 4-5 nodes more you have to work into a shader.

It has nothing to do with biased or unbiased renderers, if you put Fresnel equations into any renderer, biased, unbiased, raytracing, scanline, z buffer and so, you will have Fresnel effect in the rendering.

It only depends on the math you use for illumination, you can use Maxwell that is complete, Fresnel or even the Schlick approxiation and it will do almost the same for Fresnel effect. Approximations for math functons words pretty well, the eyes don't have the precission of a digital measurement instruments and so it will see no difference.

The involved math can be very complicated, but from the point of view of the artist, you only need to define the characteristic impendance of the two mediums and all is done by the math. If you want to ignore polarization of light you can use real numbers for the characteristic impendance instead of complex numbers and if you want to sacrify even more you can use the refraction index instead of characteristic impendance.

Lot's of interesting stuff.

bb:

I could use containers or composites in Poser, you could REALLY use them. I would say imagine how structures like this could be used with VSS or Mathmatic, but apparently you have already imagined this.

Nesting these structures inside of other containers/composites is better than pages or tabs. I like that. The "bundles" are a great concept. One really only needs one input conduit seeing that inside the structure the connections "un-bundle." Fascinating.

lmk

Quote - ...
I would like to see a nodes library. In the Advanced Material Room, right click in the nodes window, and have a library choice there under the New Nodes menu. I would like to save nodes there that I have set up. Even simple things would be a boon to me like a math node that makes an image negative. I just can't remember a lot of those things, and have to look them up time and time again. Why not name and save the node, and just add it already set up.

...

lmk

For the time being, why don't you do it, like I do...

I save all those small snippets of nodes as single materials to my library
and apply them to a hidden dummy sphere in my scene, then copy to the
material needed.
I am sure, provided one knows a little about python, one could create Wakros to handle this more efficiently.