Forum Coordinators: RedPhantom
Poser - OFFICIAL F.A.Q (Last Updated: 2024 Nov 18 10:25 pm)
Creating various sooty marks and artifacts all over your render tha wee nervy there when using Pro-Pack!
BTW, there are three means:
P4 Rendering,
P5 Firefly without RayTrace,
And P5 Firefly with RayTrace
DPH
STOP PALESTINIAN CHILD ABUSE!!!! ISLAMIC HATRED OF JEWS
Some render systems like mentalray and VRay are pure raytracers, ie they do everything through rays, from casting eye rays to lighting to shading. Which is also why they are so fast in rendering the stuff mentioned above.
Firefly chugs in these things because it is a hybrid renderer. It normally renders using a non raytracing approach but has to switch to a slow raytracing engine when you use these features. Which of course is gonna cause a render hit.
**1. Raytraced shadows
** You can do shadows without raytracing
2. Occlusion
What do you mean by occlusion?
**3. Photon mapping
Is not a variant of ray tracing?
Stupidity also evolves!
Quote - **1. Raytraced shadows
** You can do shadows without raytracing
You can do Shadows without raytracing. But they are different.
Without ray tracing you need to do a whole render from each light to build a shadow map. So you need a lot of memory for big szenes to get acceptable shadows. Ray traced shadows don't need that. You don't absolutely need ray tracing for shadows, but ray traced shadows have fewer problems than their scanline counterparts.
Quote - 2. Occlusion
What do you mean by occlusion?
Ambient Occlusion. Shadows in corners, or other small places. Poser 6 have that.
Quote -
**3. Photon mapping
**Is not a variant of ray tracing?**4. Monte carlo style GI.
Monte Carlo is used by raytracing
- IBL
**I don't know what it exactly does
Image based Lighting. You use a picture to iluminate your szene. It helps a lot if you want to create a 3d-Model which will be copied in a photographed or filmed szenery. You make a picture of a mirror ball, positioned where your model will be. Render your model with the picture from the mirror ball as IBL light and it will fit in the szene.
Poser 6 uses IBL as a method for global illumination.
**
**> Quote - **6. SSS
**This one I don't know what it is
Sub- Surface Scattering. Certain materials (skin, orange juice, wax) do not reflect all light on their surface. Light enters the object and is reflected inside the object. IMHO it is possible to fake SSS without Ray Tracing.
**
**> Quote - **7. Caustics and volumetric caustics.
- Path tracing for volumetric effects like smoke, clouds etc
**You can do this without raytracing with expection of correct refraction
Correct. But without refraction it is not easy to get a realistic look.
Scanline is not alwalys faster than ray tracing. If the szenery becomes too complex, or if you have too much lights with shadows, ray tracing is faster. The main professional products (Pixars Photorealistic Renderman and Mental Images Mentalray) are AFAIK comparable in speed - although mental ray only uses ray tracing. Pixars Photorealistic Renderman uses both Scanline (REYES) and ray tracing. So, if you look at the professionals, you can do it without scanline, but not without ray tracing,
Please correct me where I may be wrong, since I am not a professional in this area.
Greetings,
Uwe
A ship in port is safe;
but that is not what ships are built for.
Sail out to sea and do new things.
-"Amazing
Grace" Hopper
Avatar image of me done by Chidori.
I think you are mostly correct...
About SSS, you can fake it but it doesnt look nearly as nice as real SSS. Mainly cause SSS is a complex phenomenon which you cant easily duplicate with fakes. A great example is something as common as milk. There is no fake method that can successfully give you the look of real milk. Not even mentalray's SSS shaders are able to do it easily.
Also you cant do proper caustics with refraction alone. Take a look at this:
http://www.motiondesign.biz/vol_start.html
There is no way to simulate those scenes with refraction alone, and they are all very common situations in real life.
Things to note:
In the shadows scene, note how the shadows properly blur with angle and distance. And how the color from the surrounding seeps into them. Both are things you can't easily do without raytraced shadows and they add greatly to the look of a scene.
The subtle color variations in the cubes come directly from the IBL. This is also very important as it makes your scene look vibrant and rich. It also gives it nice lighting.
In the AO scene, notice how 'solid' the scene looks, even though it is just a bunch of blank white cubes. That's how important contact shadows are.
In the SSS one, note how it respects the thickness of the object. and casts internal shadows (like on the base) Both are not easy to fake.
EDIT : Just thought I'd add that this is XSI, not poser in case anyone was wondering. Poser as yet doesn't support some of this stuff, mores the pity. You can do AO and IBL. of course.
Quote - Sub- Surface Scattering. Certain materials (skin, orange juice, wax) do not reflect all light on their surface. Light enters the object and is reflected inside the object. IMHO it is possible to fake SSS without Ray Tracing.
Well, it can be done without ray tracing, it's only an illumination model.
The difference between ray tracing and not is that in the second the rays of propagates in straight light and ends when hit the last not transparent object.
With ray tracing the rays can be reflected and then in another direction where can be reflected again and also the rays can bend due refraction.
But the method without ray tracing can be very fast and done in real time.
I don't know which renderers beside games that use this method.
Poser 4 I don't know what it is, it looks as some limited ray tracer because the scene appears slowly line by line and not in an instant with all whole scene rendered.
Stupidity also evolves!
Quote - Sub- Surface Scattering. Certain materials (skin, orange juice, wax) do not reflect all light on their surface. Light enters the object and is reflected inside the object. IMHO it is possible to fake SSS without Ray Tracing.
Well, it can be done without ray tracing, it's only an illumination model.
SSS is no illumination modell, it's more a volumetric shading system. SSS works with all light sources. Light enters the object and is reflected in various depths. That is a material property.
Quote - The difference between ray tracing and not is that in the second the rays of propagates in straight light and ends when hit the last not transparent object.
With ray tracing the rays can be reflected and then in another direction where can be reflected again and also the rays can bend due refraction.
But the method without ray tracing can be very fast and done in real time.
I don't know which renderers beside games that use this method.
Poser 4 I don't know what it is, it looks as some limited ray tracer because the scene appears slowly line by line and not in an instant with all whole scene rendered.
Poser 4 is a scanline renderer. FireFly is a scanline based renderer which uses ray tracing when needed.
Graphic adapters are capable of rendering certain scenes in real time. They use the scanline algorythm. Up-to-date graphic processors are comparable to your main processor in complexity, to make real time graphics possible. They are limited in the number of polygons they can handle, and the texture size. Since scanline needs a collision test for every polygon in every pixel, it looses its performance when the scene gets to large.
Graphic adapters render in real time, software renderers usually do not. Real time is good for games, the best graphic adapters can do a lot. But high-res textures, ambient occlusion, real reflection and refractions can only be faked. Obviously faked.
If you don't plan to do photorealistic work this can be enough. If you go for photorealism there need to be ray tracing. And no, it is not "slow", at least not if you have complex scenery. Compare it with cars: true, a Ferrari or Porsche is faster than a truck. But if you need to move a lot of stuff, it will not help.
DAZ Studio and Poser 7 IMHO both have a preview function which is done by your graphics adapter. This preview is probably the best you can get in real time. This is scanline, with all options your graphics adapter can show. Since I haven't spent that much money for mine, I do not use that. See what you can get, see if it is enough.
A ship in port is safe;
but that is not what ships are built for.
Sail out to sea and do new things.
-"Amazing
Grace" Hopper
Avatar image of me done by Chidori.
A ship in port is safe;
but that is not what ships are built for.
Sail out to sea and do new things.
-"Amazing
Grace" Hopper
Avatar image of me done by Chidori.
"They are limited in the number of polygons they can handle, and the texture size. Since scanline needs a collision test for every polygon in every pixel, it looses its performance when the scene gets to large."
Zbuffer is something that is very very fast for culling the pixels of each polygon, the time that is used can be easily ignored in the whole time needed to render a pixel.
Stupidity also evolves!
I was doing some research and found many things.
I discovered the scanline algorithm, it's something that I don't use and is not efficient and a slow process.
The difference between ray tracing and other methods is only the way used to find the illumination sources that contribute to the illumination for each point of an object.
How the source affects the illumination depend on only the illumination model that you use and not if you use ray tracing or not. You can have surface scattering with or without ray tracing.
Ray tracing is able to find more correctly the illumination sources in some cases as with mirrors.
IBL is also a fake method, the reason is that the illuminantion in a point not only depend on the intensity, it depend also of the angle of incidence. The image used mapped on the sphere or other shape only contains the intensity information with no angle information, that would be impossible to be stored.
The equivalent of IBL in games are the light maps.
- Ray tracing is unable to do a correct ambient or global illumination, in theory it would be able to do it, but the required computations would last forever.
The only way to do a correct global illumination is with radiosity, then you can use if you want, the radiosity generated maps with or without raytracing.
Stupidity also evolves!
Quote - I was doing some research and found many things.
- I discovered the scanline algorithm, it's something that I don't use and is not efficient and a slow process.
All graphics adapters use AFAIK a scanline algorythm for generating their pictures. So if you have an ATI or NVIDIA graphics adapter, you use a scanline algorythm.
Quote - - The difference between ray tracing and other methods is only the way used to find the illumination sources that contribute to the illumination for each point of an object.
How the source affects the illumination depend on only the illumination model that you use and not if you use ray tracing or not. You can have surface scattering with or without ray tracing.
Ray tracing is able to find more correctly the illumination sources in some cases as with mirrors.
As far as I know, you need ray tracing to get the correct angle of the light for SSS. In a way, translucy in SSS is a kind of refraction. But as I said earlier, you can fake that without ray tracing.
Quote - - IBL is also a fake method, the reason is that the illuminantion in a point not only depend on the intensity, it depend also of the angle of incidence. The image used mapped on the sphere or other shape only contains the intensity information with no angle information, that would be impossible to be stored.
The equivalent of IBL in games are the light maps.
Since the IBL picture is put on a simulated ball arround the szene, you do have angle information.
Light maps have a similar funktion, but they need to be prepared in advance, for your object, your light set and szenery. You could call it a way of saving those light informations.
Quote - - Ray tracing is unable to do a correct ambient or global illumination, in theory it would be able to do it, but the required computations would last forever.
The only way to do a correct global illumination is with radiosity, then you can use if you want, the radiosity generated maps with or without raytracing.
The earlier mentioned Monte Carlo-Method is a method to simulate radiosity with ray tracing. A true radiosity solution is a very time consuming process, so it is very rarely used. To my knowledge all renderers use simulated radiosity, which can only be done with ray tracing. Of course you could solve the radiosity for you szene and use the result in light maps. But that would not be fast either.
Quote -
- Raytracing is unable to handle particles, smoke, clouds, fire, etc must be done by other means.
Ray tracing can handle particle systems, which are used to display smoke, fire, clouds, water and other things. There are even particle systems for POV-RAY, so you could, if really intrested, even see in source code how that is handled.
If you think scanline is a slow algorythm, as is ray tracing, what kind of algorythm are you comparing ray tracing to?
A ship in port is safe;
but that is not what ships are built for.
Sail out to sea and do new things.
-"Amazing
Grace" Hopper
Avatar image of me done by Chidori.
Quote -
How the source affects the illumination depend on only the illumination model that you use and not if you use ray tracing or not. You can have surface scattering with or without ray tracing.
Then show us an example of scattering without raytracing. You can get the buddha model i used here:
http://graphics.stanford.edu/data/3Dscanrep/
Or use whichever model you feel is appropriate.
Quote -
- IBL is also a fake method, the reason is that the illuminantion in a point not only depend on the intensity, it depend also of the angle of incidence. The image used mapped on the sphere or other shape only contains the intensity information with no angle information, that would be impossible to be stored.
The equivalent of IBL in games are the light maps.
Light maps have nothing to do with IBL. They are just stored maps for any kind of lighting. You could store IBL in them or direct lighting or occlusion or whatever. IBL is calculated on the fly each frame by casting a hemisphere of rays around each surface normal and averaging the result. I thought you already knew about this as it is clearly explained on debevac's page.
Quote -
- Ray tracing is unable to do a correct ambient or global illumination, in theory it would be able to do it, but the required computations would last forever.
The only way to do a correct global illumination is with radiosity, then you can use if you want, the radiosity generated maps with or without raytracing.
You dont know what you are talking about. No one uses radiosity in its primitive form, that is the most inefficient and limited way to do GI. It is slow as hell, doesnt handle point lights properly and cant do optical effects at all. Most apps use other better approaches like monte carlo GI or photon mapping. Or they use a combination of IBL and photon mapping. About the only commercial app that ever tried the primitive approach was Lightscape for MAX, and even they folded up due to competition from superior systems in VRay and Maxwell.
Quote -
- Raytracing is unable to handle particles, smoke, clouds, fire, etc must be done by other means.
The technical term is ray marching, it is a modified (more complex) form of raytracing:
http://graphics.ucsd.edu/courses/rendering/2004/gmromer/index.html
Quote - "They are limited in the number of polygons they can handle, and the texture size. Since scanline needs a collision test for every polygon in every pixel, it looses its performance when the scene gets to large."
Zbuffer is something that is very very fast for culling the pixels of each polygon, the time that is used can be easily ignored in the whole time needed to render a pixel.
Again as far as I know you use the scanline algorytm to fill a Z- or W-Buffer.
A ship in port is safe;
but that is not what ships are built for.
Sail out to sea and do new things.
-"Amazing
Grace" Hopper
Avatar image of me done by Chidori.
Quote - As far as I know, you need ray tracing to get the correct angle of the light for SSS. > Quote -
No unless the ray reaching the surface is a reflected or refracted ray.
Quote - Since the IBL picture is put on a simulated ball arround the szene, you do have angle information.
No, you have not. First at all the intensity depend on the angle, as a point in the IBL image can illuminate many objects and points of an object, each direction has other angle and this is not stored in the image.
You must remember that any pixel of the IBL image can iluminate the same point of the object, so the illumination on this point is the integral over all the IBL image that is able to reach this point. This is something that ray tracing don't do. It can work only for perfect mirrors because only one direction is illuminating the point and so only one point of the IBL is taken into account, but fail with surfaces that are not mirrors where all pixels of the images makes the contibution.Quote - Ray tracing can handle particle systems, which are used to display smoke, fire, clouds, water and other things.
To simulate properly you need a great amount of particles, if you are going to ray trace each particle then the render will never end. Ray tracing can work with systems that use few particles, but for large amount of particles there are other methods.
Quote - If you think scanline is a slow algorythm, as is ray tracing, what kind of algorythm are you comparing ray tracing to?
Scaning polygon by polygon using Z-buffer, no scanlines, in the rendering buffer the polygons appear in their respective position as are rendered.
Quote - Again as far as I know you use the scanline algorytm to fill a Z- or W-Buffer.
Scanline can or not use z-buffer depending on the method used, but if used the z-buffer is for one line scanline and not the whole image..
Stupidity also evolves!
My sources seem to differ from your point of view. But then, I may very well got something wrong. The whole subject is complicated enough and I surely do not plan to dig deep enough to write my own renderer.
I, for my part, need ray tracing for accurate shadows and translucency, reflections, refractions and ambient occlusion. If there is another way to archive that without ray tracing, then fine, use it. I can only use the tools that I have, and I still have no good solution for translucency available. I'm far from being professional in this business.
Greetings,
Uwe
A ship in port is safe;
but that is not what ships are built for.
Sail out to sea and do new things.
-"Amazing
Grace" Hopper
Avatar image of me done by Chidori.
I did a portrait. First I rendered using Deep shadow mapping time 5mins. the result was very good, then I used Raytraced shadows the result was Amazing!! (Render time was 4hours though! 3 hours of that was her hair! )
In conclusion, Raytraced shadows for my image incerased the quality enormously, but it sure took a long time :)
If it ain't free, I can't afford it.
The simple answer is that everything can be done without raytracing but using raytracing is easier to implement and will in the end give better results and often in a shorter time (given the same accuracy).
If your research leads you to another conclusion you are of course free to build a renderer according to your research. If you build the perfect renderer (accurate and fast) I am sure you will find a market for it.
In the opening message you asked for our opinion, we have provided this. If you are not pleased with the answers seek answers elsewhere. This discussion is not leading anywhere if you don't want to hear about the implementations using raytracing. Once you know what has been implemented using this technique and you have studied those implementations, you can draw your own conclusions and implement whatever you like best.
This site uses cookies to deliver the best experience. Our own cookies make user accounts and other features possible. Third-party cookies are used to display relevant ads and to analyze how Renderosity is used. By using our site, you acknowledge that you have read and understood our Terms of Service, including our Cookie Policy and our Privacy Policy.
I am making an research and I would like to know your opinions.
There are two ways for making a rendering, maybe are more, one is the classic way and the other one is using ray tracing.
There are no differences with both methods in the quality of the result, the the difference resides that are some things that only can be done with ray tracing. On the other side the conventional way can be done in real time but no real time for ray tracing.
As I know the only two things that can be done only with ray tracing are:
1- Real mirrors (not fake or simulations).
2- Refraction.
There's something to be added to the list?
Stupidity also evolves!