Rendering : 3rd Party Rendering Engines

[holosynthetic]

maybe it goes to show bigger companies prefer max? :)

who knows?

[halo]

its probably down to a combination of a large market share/users, a lack of competitive standard renderer and the fact that a plugin has only to be developed for one OS unlike maya, xsi, lw, c4d etc

it would be interesting to see how the sales stack up between them all though

[BrandonD]

MAX was one of the first 3d apps with a very accessible and free SDK (ships with every copy). Others have done the same later on down the road, but still, no other app has such a big market in add on features, so rendering is just part of that. Personally I think the MAX plugin market has died down significantly since R4 shipped. Long gone are the days of "a new plugin every week" back around 1998.

[3D BIG]

they know very well that 3d max is weak in rendering so they want to make profit.. they find demand

but i dont know how they make profit
cracks are everywhere

[mouj]

I'm sure a lot of people around here would tell you that Max's Scanline is not that "weak" in the first place; still it does lack some features that third party renderers have.

mouj

[3D BIG]

defualt renderer does not support some lighting effects..for example if u want to make a glass ball..and u put a light.
there should be bright point on the shadow shows the light..
3ds max defualt renderer does not show that effect

i think the best thing in 3ds max renderer is advanced lighting. .. there are some features that u cannot find in other renderes



thanx

[mouj]

I believe what you are referring to is called caustics, and indeed Max's scanline cannot render caustics, as it cannot use photons, and all those nice and somewhat fancy features that one can find in Brazil, Vray and the likes.

Now, that does not mean it is no good, actually i'm pretty sure that the scanline has been used lots of time in production environements, Max has been used in films and games production long before any renderer could produce caustics or use photons or whatever.
Just to make the point that by no means Max's scanline is "weak" or "bad" or anything.

As a side note, i'd also have to disagree on the fact that Advanced Lighting has features that are nowhere else to be found, actually Light Tracer is 'just' a GI tool like Brazil's Skylight for instance, and the Radiosity engine i believe comes from Lightscape (derived ?), and in fact is not one of a kind - it does hold some intersting features like photometric lights and such, but once again these features exist in other renderers.

mouj

[Alex Morris]

The main reason for this is usually speed not quality. The scanline is fine for most things, but the minute that you start using raytracing or GI it becomes really slow. All 4 renderers that you mention basically speed up GI and raytracing by at least a factor of 2 and add more functionality.

[CapnPanic]

while i'm a brazil user now, i WOULD like to point out that the scanline renderer, in the hands of a skilled artist, can be used to create some very impressive work. For example, the cinematics that made Blizzard so famous over the years, i believe are straight-up scanline rendered, no fancy gi, just alot of skilled artists :)

[gandhics]

Max has the biggest user base, so there is more posibillity to sell.


Maya's user base is growing, but Most of Maya house use Renderman. Some small Maya shop wants other cheap solution, but their numbers are too small as a market.

P.S. Max's renderer is fast and production proven.

[holosynthetic]

while on 3rd party renderer, does anyone know how much mental ray gets for being intigrated into maya and XSI? do they make more money from intigrating it with the software or selling separate licenses to people?

[ToddD]

Mouj"Just to make the point that by no means Max's scanline is "weak" or "bad" or anything."

Capn_"while i'm a brazil user now, i WOULD like to point out that the scanline renderer, in the hands of a skilled artist, can be used to create some very impressive work. For example, the cinematics that made Blizzard so famous over the years, i believe are straight-up scanline rendered, no fancy gi, just alot of skilled artists "

I agree 100%. I am always impressed by work when it is lit traditionally, too many people overlook the basics. Depending on GI and the like has produced an attitude among newer users that lighting isn't something to be learned. Scanline isn't the lame duck it is made out to be. Took me a year to realise that. I love these new renderers as well, but to be honest, I think 75% of what you see can be produced or "faked" with scanline.

Of course in a production environment I might feel differently. :thumbsup:

[CapnPanic]

yeah there are lots of things that CAN be faked with scanline, but you just don't want to have to do it that way, becuase the rendertime hit is so bad. for example lighting a large outdoor scene with a skydome-type light and a single bright sunlight. while you can do this in scanline, in order to get decent shadow quality you have to really crank up the shadow map size or use raytraced shadows (both will really hurt your render time) OR you can use a 3rd party renderer which has a skylight (ambient occlusion) built in and is way faster with way better quality :)

i think its way better to learn for a while on the scanline, because it teaches you all the tricks how to keep render times low etc and THEN move to a 3rd party renderer if you need to, as your render times will only get faster :)

[3DDave]

You only get ONE license that will render on your machine ONLY. To render on a network with a farm you need to buy network rendering licenses.

Thanks for the responses about the volume of renders for Max.

Originally posted by holosynthetic
while on 3rd party renderer, does anyone know how much mental ray gets for being intigrated into maya and XSI? do they make more money from intigrating it with the software or selling separate licenses to people?

[3D BIG]

advance lighting supports real radiosity not only glubal illuminatoin

Radiosity in advance lighting is very close to the real Radiosity..
as we know Radiosity is rendering technology that realistically simulates the way in which light interacts in an environment.

brazil and all other renderers works under global illuminatoin rule...which shows fake shadows and effects.
when u use radiosity in 3d max u should be carefull about segments in the scene the light interacts independently... no matter where u put ur camera. this feature is not exist in other renderes.

u can get more informatoin (if u know arabic) :) and i am sure u dont know..through this link from maxforums
http://maxforums.net/showthread.php?s=&threadid=24496&perpage=30&pagenumber=1

sorry for my bad english
;)

[ZeBoxx]

O_O

GI, although highly generalized in just those two letters, generally does not create 'fake' shadows.

The shadows created through this method are an absence of light much in the way that a finite element radiosity approach creates them.

I would highly recommend looking up at least two of the common algorithms in many GI engines, Quasi-Monte Carlo sampling and Photon Mapping, and reading through the outset of how they work and why they work.
You'll then see that the shadows these methods create are every bit as 'real' as that of FEM radiosity.

[3D BIG]

there is a big deference between global illuminatoin and radiosity

can anyone explain this deference ( i want to know not to challenge ur knoledge)..

i will be glad if someone explain it



thanx

[ZeBoxx]

Heya,

For that, we'd first have to make the general notions specific.

Let's state that when we refer to GI, we're actually referring to QMC sampling.

And for 'radiosity', let's state that we're referring to a finite element method.

Then very much generalized, because these things literally fill dozens of research papers, we can state that...

GI calculates the lighting in a scene by sampling said scene using a series of semi-random probes of the scene surrounding the point in the scene you're calculating the lighting for.
This semi-random nature can often be easily combined with information from the surface you're rendering to get both a (lambertian) diffuse and a specular (view-dependent) component
The results of these calculations are, in general, not stored in any way. They are calculated once and then thrown away.

radiosity, calculates the lighting in a scene by dividing meshes up into (usually) triangles (or 'patches', not to be confused with 3ds max's patch primitives), and for each triangle calculating how much the other triangles' values would affect the triangle. If the difference between two neighboring triangles varies too much, the triangles are divided up further.
This process continues until a desired level of uniformity and completeness is reached.
The results of these calculations are, by very definition, stored in some way. In LightScape with 3ds max, for example, this is stored in a mesh form with vertex colors being used to indicate the lighting. This makes the technique very useful for architects who may only wish to do a fly-through of a building. The calculation only has to be done once. It is even possible to use this information for realtime representations (VRML, games, etc.)
By contrast, however, these techniques traditionally ignore view-dependent effects and thus only represent the (lambertian) diffuse component of the lighting.

Both methods have their merits and downfalls, and of course there are dozens of variations on the themes.
(e.g. GI with photon mapping, or FEM using hierarchical 'direct-lighting first' adaptive meshing)

If you really want to get to the heart of these things, reading research papers would be your best bet.

[Dave Black]

I'd also like to interject a few things.

First off, anyone how has used max in production knows that high-poly counts can make max become unstable. I've heard(and this could be completely baseless)that max has problems with memory allocation. Final Render, Brazil, etc. all have better memory allocation than scanline. I've seem lots of animation where brazil was used, but there are no GI, Radiosity, Caustics, etc in the animation. Usually, it's explained that brazil is simply capable of handling high-volumes of geometric data without barfing.

One must also look at some of the newer tech that is being implemented in these systems. FinalRender Stage 1 seems to have it's own "micro triangle" displacement system. I don't know if it's akin to renderman's displacement utilizing normal mapping, but if it is, we could soon see undercut displacement coming out of these renders. Not to mention, allot of the engineering and toonshading rendering capabilities. So besides caustics, SSS, GI, Radiosity, and the like, these new rendering engines allow for more features that scanline currently delivers.

I'd also like to point out, that in production, my trusty scanline renderer has enough muscle to do 95% of everything I need. In fact, I feel like these plugin renderers(through no fault of their own) are doing new max users a disservice.

Learn scanline. If you have exhausted all approaches using that, then find the plugin renderer that fits your needs.

Just some thoughts. I could be wrong on every one of my points.

-3DZ

:D

[xynaria]

I would concur with Alex Morris that the scanlines biggest drawback is primarily speed. Ironically for simple scenes and materials, Max's scanline is possibly one of the fastest around but as soon as your material trees grow so does your rendering and in the case of raytracing .. and how. The main trouble there is that unlike some other programmes which lead one to believe that their native renderers are under relatively continuous optimisation and development, Max's doesn't appear to be. That to me would have been more of a priority than the 'solutions' included in R5 which I still maintain where inappropiate for a lot of Max users. Mind you if the what the world and his dog are saying on the rumour farm happens to have any truth, then it would appear that I'm not the only person to think this. :)

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.