About the axe head: Assuming you're using a PBR workflow, its basically a flat grey for the diffuse and a speckled noise like texture in the roughness with the changes in direction of the axe head being a bit smoother than the rest.
Add a mild normal map to it based on the metal roughness and you'll be mostly there.
Usually, there is always a newcomer that comes and mentions weekly things or challenges and 90% of the time it ends up in no one doing anything because it becomes a chore.
If you want to make something happening, just start making something and then ppl will see if they want in.
Making a challenge out of things usually doesn't work.
According to the sources, Diffuse and Specular, in physics, are 2 sides of the same coin:
Specular is when the light bounces from the surface linearly in the opposite direction
Diffuse is when it gets scattered around
(and, if I got it correctly, roughness and microsurface contribute to define IF and HOW the reflections are scattered)
Because of conservation of energy law, highly reflective materials tend to also have weaker diffuse, since only so much light can be bounced off in any direction (and if it's concentrated in one spot, it can't be diffused - makes sense)
Here is my dumb question:
Does the PBR theory explain white glossy plastic? (if it's white it has a strong diffuse, but it's glossy so it's also highly reflective) Where does this exist within the energy conservation law?
Please not I'm not being rhetorical or skeptical. I'm genuinely asking.
Also, don't Roughness and Microsurface pretty much do the same? (aka scatter around the light because the material has bumps and pores)
Does the PBR theory explain white glossy plastic? (if it's white it has a strong diffuse, but it's glossy so it's also highly reflective) Where does this exist within the energy conservation law?
Please not I'm not being rhetorical or skeptical. I'm genuinely asking.
Also, don't Roughness and Microsurface pretty much do the same? (aka scatter around the light because the material has bumps and pores)
This is not a dumb question, it's actually a really good question! I hope this explanation will suffice:
An object that is white is still scattering light that is not absorbed. A mirror-like surface is just scattering it in a significantly more predictable fashion. Glossiness for a white object is due to a greater amount of that scattered light being scattered in the same direction, much like a mirror, but at a significantly reduced amount compared to that of a mirror. It's important to remember that scattering is not all-or-nothing -- you can have scattering, but at the angle of reflection to the light source, you'll typically have more of that scattered light, resulting in a bright spot.
As to your second question, roughness/gloss/microsurface dictate the amount of scatter within the allowable range for a non-metal/metal. A rough, white, object will reflect and scatter all of the light off the surface. A shiny, white, object will reflect all, and scatter some of the light off the surface. And if you were to polish that object a whole bunch, it would reflect all, and scatter none of the light off the surface, resulting in a mirror.
Glossy and reflective is just a matter of degrees in how scattered the light is off and object and the short answer is: a matte or flat object scatters light a whole lot, a glossy object still scatters light a bunch, and a mirror-like object doesn't scatter it much at all.
This handy image from Marmoset's page on PBR theory gives you a good idea of how far apart dull/glossy and reflective are on the same line, and that at grazing angles, even non-reflective materials can exhibit reflective properties (called Fresnel)!
I feel like that was kind of a confusing explanation, though it does have some good points (and maybe some inaccuracies? some things I couldn't quite make sense of some stuff from the wording).
There's a few key bits of physics at play. Absorption, refraction, reflection, transmission are the most important in the context of typical materials.
Diffuse light happens because of the fact that things are made of microscopic structures. This allows photons to move through the surface and be absorbed, reflected, or transmitted. Pulling from Wikipedia, one can imagine a crystalline structure like snow where rays of light have the potential to enter the surface. At each interface a photon could be absorbed by the medium, reflected in whatever direction, or continue to be transmitted through the surface. The reason water is clear is because it lacks the structure to continually scatter light internally. No amount of polishing a surface will turn it into a mirror.
Now, refraction is the property that as light enters a medium, it changes speed. This causes a change in angle, which is the reason that you cant spear a fish while standing in clear water by aiming directly at it, you have to account for the change in angle at the interface of the surface.
Specular reflections happen when photons hit a surface and bounce off it without transmission or adsorption. The fresnel effect describes the ratio of reflected versus refracted light. This changes based off the angle that light hits the surface. As the angle of the light to the surface becomes more shallow, the greater the percentage of reflected light becomes (as it has less potential to enter the surface, it may help to think of skipping a stone). You can see this visually in Gearman's example.
This relationship is expressed as the index of refraction. In terms of the typical rendering models you'll see, this is simplified substantially and is usually expressed as a percentage (read up on Schlick's approximation. the relevant bit is R0 = ((n1 + n2)/(n1 - n2))^2. n1 is the refractive index of air and n2 is the index of the surface. R0 is the total percentage of light that is reflected, and R Theta is the percentage based on angle, which goes to 100% at grazing angles. Read more here. http://en.wikipedia.org/wiki/Schlick%27s_approximation). Particularly for metals this is simplified as percentages of RGB color, as their reflections are related to their subatomic structure (this is why only metallic surfaces have colored reflections and no diffuse) and can not be simulated in realtime.
For the Metallic style PBR, they lock nonmetals at 4% reflection which is a refractive index of 1.5 for the material. This is good enough for most everything. The metalness map replaces this 4% with whatever is in the albedo texture. This is why people advise against painting grey values in metalness maps. For example a 50% grey would be 50-50 blending between that 4% reflection and the albedo textures color, and then removing 50% of the albedo. Note that a transition between rusted and non-rusted metal would have grey values in its metalness map in order to look decent, but this can be rationalized as thinking that at a sub-pixel level it's defining the percentage of metal versus nonmetal.
Scattering is an affect of the structure of the mediums surface. A gloss/roughness map are the same thing, microsurface normals. They describe a statistical amount of surface normal irregularities at a sub pixel level just like a normal map describes the angle of a surface normal per pixel, or the vertex normals do at a vertex normal level (in this sense it can be related to a micro bumpmap, defining the strength of the roughness/bumpiness). When light hits a surface, its reflected across the normal. By making the normals irregular, you're scattering! It's all just different degress of granularity which I feel makes understanding the principles at play much, much clearer. The final thing to note about the roughness of a surface is that the appearance of roughness is diminished as the angles become more grazing because of the fact that more micronormals are going to be aligned to reflect, which is the largest principle at play that makes the ball Gearman posted look mirrorlike around its edges.
Please let me know if I can clarify any of this I can provide code examples of all of these things and their interrelationships. Going further though gets into microfacet theory which gets a lot more complex lol.
It was probably a bit confusing, because I greatly simplified much of the physics involved. When I've had to explain PBR to other people (mainly artists) I can see their eyes glaze over as soon as I start talking about angles of incidence, angles of reflection, refractive indices, etc. I tend to not cover index of refraction at all (IOR) because it is not typically used in games. I admit that much of the confusion is most likely from the singular term for "reflection".
To put it more succinctly, an object can exhibit both a diffuse and a specular reflection, they are independent of each other. A glossy white object will reflect ~4% of the light in a specular fashion, and has a very high amount of diffuse reflectance as well. A glossy, colored, object of similar material would reflect ~4% specular, but would have a lower amount of diffuse reflectance, the amount of which would determine its color. A roughness map merely adjusts the surface roughness, which affects the sharpness of the specular highlight.
I understood everything of it, thanks to your clear explanations.
This is a very interesting subject! I have a bunch of tabs open, waiting to be read. I'm looking forward to learn more.
So I'm trying to create a tarnished brass material using a metal-rough workflow. I understand that this kind of blackening of brass is an oxidization, but I'm wondering how I could represent this using metal-rough. Would it just be considered a partial metal or would I just change the reflectance value of it in the albedo?
The darker stuff on the outside is indeed oxidation, which means its nonmetallic. However the amount of oxidation on top of the metal is of variable thickness, so its best to use nice gradients in your Metalness map so you get bits that are fully oxidized and have no metallic reflections, and parts where theres only a very thin layer of patina so you have some nice metalness leaking through.
However Ive noticed Marmoset has a real problem with gradient metalness, so its best to do you testing in UE4 or another rendered which works correctly.
How do I texture convincingly a relatively new, nearly unused and clean stainless steel object? All tutorials out there focus on adding wear & tear, scratches, rust, dust and all kinds of signs of actual use and time. But assuming the object is clean and in good condition, how and where do I add variation to make the texture interesting?
How do I texture convincingly a relatively new, nearly unused and clean stainless steel object? All tutorials out there focus on adding wear & tear, scratches, rust, dust and all kinds of signs of actual use and time. But assuming the object is clean and in good condition, how and where do I add variation to make the texture interesting?
Clean textures are generally uninteresting. A whole lot of clean textures is when things become interesting.
Someone can give some tips how to set proper worn bluing metal on spec-gloss workflow?
Here is example ("damages on edges")
So in my opinion these edges should be slightly visible on diffuse/albedo, white (well, not full white) on specular, and dark on gloss. I have tested it many times, but I would like to know what do you think about that.
I think what he means is that a single clean object or material on its own will look boring, but an entire environment or complex model of very clean materials is interesting because it then becomes a design aesthetic.
Also, look around you for examples. Even new things have signs of wear, just a lot less of it. Are there welding joints? Machined creases or divots? Anywhere where machine oil might be caught up and not cleaned off during the creation process? Are there joints where it articulates that require some form of lubrication or that come in contact with other pieces of the object when they move? These will all have some signs of wear, however minor.
Has it been handled by humans at all? Perhaps those areas will have some accumulated oils that change the surface quality.
Often something might be clean and not have anything special with the diffuse texture, but can have variations by changing the specular areas, showing the worn areas. Like rough plastic might get smooth from lot of wear, but you cant really see it except from the light reflecting from it, the shininess.
hey, i'm a beginner and i hope you dont mind me asking about this basic stuff:
i have this simple form i modelled but i wonder what's the smartest way to keep the edges when smoothing and making them nicely round:
this is my lowpoly-mesh
this is my approach i try usually, with swift loops, but i dont know, it looks messy and wrong, something wrong with MY approach here or doesnt it really fit in this case?
the easiest way seems to be to just tesselate alot
@chomp The glass or the brick? The brick would be a very simple albedo texture, with the Normal and AO really making it pop, the glass would just be a solid colour with the majority of the detail coming from the glossiness / roughness map I believe.
I wonder how to make a glass material (to use in Toolbag 2) that has some dirt (or any (diffuse) information) on it. So far I played with the transparency setting and tried loading a opacity map (to affect/show the albedo map) but had no luck getting the result I want.
Does anybody of a database or something where you can find the scientific roughness values of various materials?
http://refractiveindex.info/ is awesome but i dont think it has anything about roughness. ive only used it for reflections and there's a section specifically for 3d artists
I wonder how to make a glass material (to use in Toolbag 2) that has some dirt (or any (diffuse) information) on it. So far I played with the transparency setting and tried loading a opacity map (to affect/show the albedo map) but had no luck getting the result I want.
Some help here would be great : )
There's a preset material for glass in toolbag. If you use that and then make some textures for what you need, you should be good to go.
Does anybody of a database or something where you can find the scientific roughness values of various materials?
http://refractiveindex.info/ is awesome but i dont think it has anything about roughness. ive only used it for reflections and there's a section specifically for 3d artists
I don't think so, and it would be extremely hard to create one. The problem really boils down to the fact that roughness is a property of a surface boundary, whereas reflection is a property of a material. You can roughen or polish surfaces pretty much however you like.
That said, there are material reference databases such as the Merl 100 that you could look at. Disney has a great BRDF viewer which allows you to get an understanding of how material properties are interconnected: Understanding the 3D and Polar Plots is an awesome way to learn the science behind materials. You can see exactly how albedo, roughness, reflection, etc all work together at a glance.
Actual roughness parameters nowadays typically compute a statistical gaussian distribution of heights for each pixel. The rougher the input value, the noisier the heights. By understanding how your roughness visually changes in this manner and comparing to reference imagery you'll grow skilled at seeing the material properties through the pictures.
How would I go about making a realistic blued gun metal material? I don't have any actual guns so I can't obverse how the material behaves, but I can't tell if it's a dielectric like paint or just a very dark metal. The fresnel effect doesn't look very apparent on the barrel compared to the wood.
The 'blueing' you are seeing here is the reflection of the bright blue sky, rather than the metal material makeup. Actual blued metal is far more apparent: LINK If you are just trying to make gun metal, then that is a different question than if you are making actual blued metal. Someone can correct me if I'm wrong, but both standard gun metal and blued metal would still exists as 100% metallic in PBR.
Well it's oxidized metal, so I would think not. A good place to start from if anyones intent is to get a base material as accurate as possible: You could make a template for black oxidized metal (or MANY materials) based off its brdf in the Merl 100- http://people.csail.mit.edu/wojciech/BRDFDatabase/brdfs/
Quack, 100% metal doesn't exist, and particularly not in that case. If the weapon was made of "pure" metal it would take mere hours for it to be completely caked in rust.
The metal in question here is a surface-treated and oxidized alloy. The 100% idea is fairytale nonsense.
So outside of the dust and oil on top of the metal, the metal isn't 100% metal?
How would I go about making a realistic blued gun metal material? I don't have any actual guns so I can't obverse how the material behaves, but I can't tell if it's a dielectric like paint or just a very dark metal. The fresnel effect doesn't look very apparent on the barrel compared to the wood.
Assuming a UE4 shader workflow: Moderate roughness. Somewhat lower than the wood but not by much. Include scuffs or fingerprints etc in the roughness map. Low or no metalness Dark cool deep charcoal grey/black diffuse, possibly lightly mottled or speckled Normal map for actual topological changes like the screws and maker imprint.
The fresnel reflections on the rounded metal parts are quite apparent, just as much so as the wood. You can see the color of the wood and strap in the shell magazine. Other rounded parts are similarly reflecting the environment well. Most of the work for "realness" is going to be carried by your roughness map. That's where the scuffs of handling and such will show the most.
The 'blueing' you are seeing here is the reflection of the bright blue sky, rather than the metal material makeup. Actual blued metal is far more apparent: LINK If you are just trying to make gun metal, then that is a different question than if you are making actual blued metal. Someone can correct me if I'm wrong, but both standard gun metal and blued metal would still exists as 100% metallic in PBR.
Quack: Gun Blueing is the term for adding the dark blue-black oxide layer, via electrochemical processes. It is a correct term and is a separate process the temper controlled blue coloring in your link. While technically both colorations come from oxide layers, the structure of each is different enough that metalness shader settings should be applied to the tempering coloring but not gun blueing. The crystalline structure of the bluing is much more erratic than the tempered oxide layer. The color from the tempering is due to light interference within the first few nanometers of the surface of the blade. There's a physical boundary layer change going on like in other iridescent materials, or like gold, copper, or similar colored metals.
Its similar to color anodizing. The color comes from light interference rather than absorption by the material itself like in most other materials.
Tempered or annodized colored metals should have a high metalness setting in a shader, as the reflected color of the material and its environment, are changed by the material itself.
Notice how even the hotspots in that photo take on the color of the material? That is a sign you should turn up the metalness in that area.
The chemical bluing, however, should have a low or absent metalness setting since the environmental reflections are not altered by the surface.
@xSNak3rs What about it exactly? It's a dull greenish-brown. Try to find a good mid-range value just off from the highlight to get something close to a base albedo value here.
I'm trying to think how to create a circular scratches pattern with procedural texture... i think the solution is quite simple but I just cant find the right solution, help!
I feel like the original post in this thread, even if it is very old, never got good answers for Pedro's limitations.
First: the point to PBR is not the realistic rendering of materials with realistic constraints. It's not going to be used by anybody studying optics, not ever. It's as much an abstraction as any other rendering style. If mother-of-pearl exists (it does) and your shader can't render it (it probably can't), then your shader isn't very physically based, is it? The point to PBR is first-- like all other renderers-- to make pretty pictures; and second, to do so with a standard set of parameters that simplify the creation of new materials and shaders, that make it easier to reuse materials between scenes and games. When it fails at the first priority, it fails period, do something else.
With a mother-of-pearl pistol grip, you're probably looking at a first person shooter; you're probably looking at something with a small number of vertices; you're probably looking at something that's going to occupy a predictable amount of screen space, for fill-rate purposes; you're probably looking at something that's going to be on-screen a lot of the time, that the player is going to be able to inspect closely, something that has the opportunity to really wow them. This is the ideal place for a custom shader. One extra draw call isn't going to break anything.
But, maybe your shader programmers disagree-- maybe they're dogmatic, maybe they just don't want to write a new shader. You can probably do a few things to simulate the tools you'd need for mother-of-pearl, without using anything that would make fundamentalists angry.
A sharp environment map is specular. Drop your specular component and color your cube map and you have colored (blinn-phong, NdotH^n) specular. You lose dynamic lights, of course. You can regain dynamic lighting on your specular if you can rotate your cube map to the light vector. You can do this in HLSL, but I don't know if you can do it with your nodes setup. (You can create Blinn-Phong specular on any number of light sources by adding up any number of rotated cube map lookups, but these rotations are more expensive than just doing the pow().)
Specular masking should be replaceable with multiple meshes with complementary alpha channels. Might be a few problems if you don't use add blending for the layers (see below), but probably not noticeable.
Multiple directions of anisotropy can be managed in HLSL with multiple meshes or multiple UV coordinates, if you have them. Pretty much the same thing anyways.
In the case of your source image, I would want to use additive blending. Can you do that? You have a nearly transparent surface on which you're applying white specular / environment map, and the difference in depth between that and the mother-of-pearl fracture layers is part of what gives it its look. If you can do add blending, you can have multiple, simple fracture planes placed inside of the handle, each receiving different specular from a differently colored environment map, and probably contributing only specular, followed by the highlights of the surface. Would be best to have a hard black backing for your add blends, not present in the source image but I think the concept would permit it.
Normal maps are appropriate for the fracture planes, and probably necessary to get good color from a simple mesh and these specular work-arounds.
All of that is probably more expensive computationally than just creating a custom shader that actually makes good mother-of-pearl, though. It's just a bunch of dumb workarounds for a dogmatic workplace.
Its similar to color anodizing. The color comes from light interference rather than absorption by the material itself like in most other materials.
Tempered or annodized colored metals should have a high metalness setting in a shader, as the reflected color of the material and its environment, are changed by the material itself.
Notice how even the hotspots in that photo take on the color of the material? That is a sign you should turn up the metalness in that area.
The chemical bluing, however, should have a low or absent metalness setting since the environmental reflections are not altered by the surface.
Hope that helps.
Ive been looking for something like this image for ages Thanks!
Anyone know how to get the grass to look like this?
Looks a lot like up close meshes and alpha planes for the plant tops rapidly LODing to clusters of alpha test planes with altered normals that match the underlying terrain. Also looks like there's some hue shift based either on the underlying terrain texture, or some other global position sampled map.
I'm trying to think how to create a circular scratches pattern with procedural texture... i think the solution is quite simple but I just cant find the right solution, help!
attaching a sample picture:
Hey! Super old question, but it's a very interesting one.
These "circular" scratches are in reality a shit ton of crosshatching scratches sitting on a spherical surface. Since they're created when something is scratched against the surface with a direction, the direction of material is changed on a micro level. It's almost like there's fibers being created, all of them pointing in the direction of how the scratch was first created. Soooo, the light response is bent in that direction. It's essentially an anisotropic light response.
Replies
eyyyyy
Add a mild normal map to it based on the metal roughness and you'll be mostly there.
Usually, there is always a newcomer that comes and mentions weekly things or challenges and 90% of the time it ends up in no one doing anything because it becomes a chore.
If you want to make something happening, just start making something and then ppl will see if they want in.
Making a challenge out of things usually doesn't work.
According to the sources, Diffuse and Specular, in physics, are 2 sides of the same coin:
Specular is when the light bounces from the surface linearly in the opposite direction
Diffuse is when it gets scattered around
(and, if I got it correctly, roughness and microsurface contribute to define IF and HOW the reflections are scattered)
Because of conservation of energy law, highly reflective materials tend to also have weaker diffuse, since only so much light can be bounced off in any direction (and if it's concentrated in one spot, it can't be diffused - makes sense)
Here is my dumb question:
Does the PBR theory explain white glossy plastic? (if it's white it has a strong diffuse, but it's glossy so it's also highly reflective) Where does this exist within the energy conservation law?
Please not I'm not being rhetorical or skeptical. I'm genuinely asking.
Also, don't Roughness and Microsurface pretty much do the same? (aka scatter around the light because the material has bumps and pores)
This is not a dumb question, it's actually a really good question! I hope this explanation will suffice:
An object that is white is still scattering light that is not absorbed. A mirror-like surface is just scattering it in a significantly more predictable fashion. Glossiness for a white object is due to a greater amount of that scattered light being scattered in the same direction, much like a mirror, but at a significantly reduced amount compared to that of a mirror. It's important to remember that scattering is not all-or-nothing -- you can have scattering, but at the angle of reflection to the light source, you'll typically have more of that scattered light, resulting in a bright spot.
As to your second question, roughness/gloss/microsurface dictate the amount of scatter within the allowable range for a non-metal/metal. A rough, white, object will reflect and scatter all of the light off the surface. A shiny, white, object will reflect all, and scatter some of the light off the surface. And if you were to polish that object a whole bunch, it would reflect all, and scatter none of the light off the surface, resulting in a mirror.
Glossy and reflective is just a matter of degrees in how scattered the light is off and object and the short answer is: a matte or flat object scatters light a whole lot, a glossy object still scatters light a bunch, and a mirror-like object doesn't scatter it much at all.
This handy image from Marmoset's page on PBR theory gives you a good idea of how far apart dull/glossy and reflective are on the same line, and that at grazing angles, even non-reflective materials can exhibit reflective properties (called Fresnel)!
There's a few key bits of physics at play. Absorption, refraction, reflection, transmission are the most important in the context of typical materials.
Diffuse light happens because of the fact that things are made of microscopic structures. This allows photons to move through the surface and be absorbed, reflected, or transmitted. Pulling from Wikipedia, one can imagine a crystalline structure like snow where rays of light have the potential to enter the surface. At each interface a photon could be absorbed by the medium, reflected in whatever direction, or continue to be transmitted through the surface. The reason water is clear is because it lacks the structure to continually scatter light internally. No amount of polishing a surface will turn it into a mirror.
Now, refraction is the property that as light enters a medium, it changes speed. This causes a change in angle, which is the reason that you cant spear a fish while standing in clear water by aiming directly at it, you have to account for the change in angle at the interface of the surface.
Specular reflections happen when photons hit a surface and bounce off it without transmission or adsorption. The fresnel effect describes the ratio of reflected versus refracted light. This changes based off the angle that light hits the surface. As the angle of the light to the surface becomes more shallow, the greater the percentage of reflected light becomes (as it has less potential to enter the surface, it may help to think of skipping a stone). You can see this visually in Gearman's example.
This relationship is expressed as the index of refraction. In terms of the typical rendering models you'll see, this is simplified substantially and is usually expressed as a percentage (read up on Schlick's approximation. the relevant bit is R0 = ((n1 + n2)/(n1 - n2))^2. n1 is the refractive index of air and n2 is the index of the surface. R0 is the total percentage of light that is reflected, and R Theta is the percentage based on angle, which goes to 100% at grazing angles. Read more here. http://en.wikipedia.org/wiki/Schlick%27s_approximation). Particularly for metals this is simplified as percentages of RGB color, as their reflections are related to their subatomic structure (this is why only metallic surfaces have colored reflections and no diffuse) and can not be simulated in realtime.
For the Metallic style PBR, they lock nonmetals at 4% reflection which is a refractive index of 1.5 for the material. This is good enough for most everything. The metalness map replaces this 4% with whatever is in the albedo texture. This is why people advise against painting grey values in metalness maps. For example a 50% grey would be 50-50 blending between that 4% reflection and the albedo textures color, and then removing 50% of the albedo. Note that a transition between rusted and non-rusted metal would have grey values in its metalness map in order to look decent, but this can be rationalized as thinking that at a sub-pixel level it's defining the percentage of metal versus nonmetal.
Scattering is an affect of the structure of the mediums surface. A gloss/roughness map are the same thing, microsurface normals. They describe a statistical amount of surface normal irregularities at a sub pixel level just like a normal map describes the angle of a surface normal per pixel, or the vertex normals do at a vertex normal level (in this sense it can be related to a micro bumpmap, defining the strength of the roughness/bumpiness). When light hits a surface, its reflected across the normal. By making the normals irregular, you're scattering! It's all just different degress of granularity which I feel makes understanding the principles at play much, much clearer. The final thing to note about the roughness of a surface is that the appearance of roughness is diminished as the angles become more grazing because of the fact that more micronormals are going to be aligned to reflect, which is the largest principle at play that makes the ball Gearman posted look mirrorlike around its edges.
Please let me know if I can clarify any of this
To put it more succinctly, an object can exhibit both a diffuse and a specular reflection, they are independent of each other. A glossy white object will reflect ~4% of the light in a specular fashion, and has a very high amount of diffuse reflectance as well. A glossy, colored, object of similar material would reflect ~4% specular, but would have a lower amount of diffuse reflectance, the amount of which would determine its color. A roughness map merely adjusts the surface roughness, which affects the sharpness of the specular highlight.
The wikipedia entry for diffuse reflection covers this really well, and will mostly clear up any remaining confusion: http://en.wikipedia.org/wiki/Diffuse_reflection
This is a very interesting subject! I have a bunch of tabs open, waiting to be read. I'm looking forward to learn more.
The darker stuff on the outside is indeed oxidation, which means its nonmetallic. However the amount of oxidation on top of the metal is of variable thickness, so its best to use nice gradients in your Metalness map so you get bits that are fully oxidized and have no metallic reflections, and parts where theres only a very thin layer of patina so you have some nice metalness leaking through.
However Ive noticed Marmoset has a real problem with gradient metalness, so its best to do you testing in UE4 or another rendered which works correctly.
Clean textures are generally uninteresting. A whole lot of clean textures is when things become interesting.
Here is example ("damages on edges")
So in my opinion these edges should be slightly visible on diffuse/albedo, white (well, not full white) on specular, and dark on gloss. I have tested it many times, but I would like to know what do you think about that.
Also, look around you for examples. Even new things have signs of wear, just a lot less of it. Are there welding joints? Machined creases or divots? Anywhere where machine oil might be caught up and not cleaned off during the creation process? Are there joints where it articulates that require some form of lubrication or that come in contact with other pieces of the object when they move? These will all have some signs of wear, however minor.
Has it been handled by humans at all? Perhaps those areas will have some accumulated oils that change the surface quality.
i have this simple form i modelled but i wonder what's the smartest way to keep the edges when smoothing and making them nicely round:
this is my lowpoly-mesh
this is my approach i try usually, with swift loops, but i dont know, it looks messy and wrong, something wrong with MY approach here or doesnt it really fit in this case?
the easiest way seems to be to just tesselate alot
I wonder how to make a glass material (to use in Toolbag 2) that has some dirt (or any (diffuse) information) on it.
So far I played with the transparency setting and tried loading a opacity map (to affect/show the albedo map) but had no luck getting the result I want.
Some help here would be great : )
That said, there are material reference databases such as the Merl 100 that you could look at. Disney has a great BRDF viewer which allows you to get an understanding of how material properties are interconnected:
Understanding the 3D and Polar Plots is an awesome way to learn the science behind materials. You can see exactly how albedo, roughness, reflection, etc all work together at a glance.
Actual roughness parameters nowadays typically compute a statistical gaussian distribution of heights for each pixel. The rougher the input value, the noisier the heights. By understanding how your roughness visually changes in this manner and comparing to reference imagery you'll grow skilled at seeing the material properties through the pictures.
https://cdn2.artstation.com/p/assets/images/images/001/833/742/large/victor-kam-c-harvester-01.jpg?1453449287
Just paint up a self illumination map like usual. Can be single channel or RGB depending on your target engine.
I don't have any actual guns so I can't obverse how the material behaves, but I can't tell if it's a dielectric like paint or just a very dark metal.
The fresnel effect doesn't look very apparent on the barrel compared to the wood.
Actual blued metal is far more apparent: LINK
If you are just trying to make gun metal, then that is a different question than if you are making actual blued metal.
Someone can correct me if I'm wrong, but both standard gun metal and blued metal would still exists as 100% metallic in PBR.
You could make a template for black oxidized metal (or MANY materials) based off its brdf in the Merl 100- http://people.csail.mit.edu/wojciech/BRDFDatabase/brdfs/
Disney's viewer is here:
http://www.disneyanimation.com/technology/brdf.html
TY!
Assuming a UE4 shader workflow:
Moderate roughness. Somewhat lower than the wood but not by much. Include scuffs or fingerprints etc in the roughness map.
Low or no metalness
Dark cool deep charcoal grey/black diffuse, possibly lightly mottled or speckled
Normal map for actual topological changes like the screws and maker imprint.
The fresnel reflections on the rounded metal parts are quite apparent, just as much so as the wood. You can see the color of the wood and strap in the shell magazine. Other rounded parts are similarly reflecting the environment well.
Quack: Gun Blueing is the term for adding the dark blue-black oxide layer, via electrochemical processes. It is a correct term and is a separate process the temper controlled blue coloring in your link.Most of the work for "realness" is going to be carried by your roughness map. That's where the scuffs of handling and such will show the most.
While technically both colorations come from oxide layers, the structure of each is different enough that metalness shader settings should be applied to the tempering coloring but not gun blueing.
The crystalline structure of the bluing is much more erratic than the tempered oxide layer. The color from the tempering is due to light interference within the first few nanometers of the surface of the blade. There's a physical boundary layer change going on like in other iridescent materials, or like gold, copper, or similar colored metals.
Its similar to color anodizing. The color comes from light interference rather than absorption by the material itself like in most other materials.
Tempered or annodized colored metals should have a high metalness setting in a shader, as the reflected color of the material and its environment, are changed by the material itself.
Notice how even the hotspots in that photo take on the color of the material?
That is a sign you should turn up the metalness in that area.
The chemical bluing, however, should have a low or absent metalness setting since the environmental reflections are not altered by the surface.
Hope that helps.
guessing you make a brush in photoshop and then tweak colors and go for it
lots of youtube videos on how to make your own brush
I'm trying to think how to create a circular scratches pattern with procedural texture... i think the solution is quite simple but I just cant find the right solution, help!
attaching a sample picture:
Update:
Found a nice small node setup the can fit, check it out:
https://blenderartists.org/forum/showthread.php?216113-Brecht-s-easter-egg-surprise-Modernizing-shading-and-rendering/page478
First: the point to PBR is not the realistic rendering of materials with realistic constraints. It's not going to be used by anybody studying optics, not ever. It's as much an abstraction as any other rendering style. If mother-of-pearl exists (it does) and your shader can't render it (it probably can't), then your shader isn't very physically based, is it? The point to PBR is first-- like all other renderers-- to make pretty pictures; and second, to do so with a standard set of parameters that simplify the creation of new materials and shaders, that make it easier to reuse materials between scenes and games. When it fails at the first priority, it fails period, do something else.
With a mother-of-pearl pistol grip, you're probably looking at a first person shooter; you're probably looking at something with a small number of vertices; you're probably looking at something that's going to occupy a predictable amount of screen space, for fill-rate purposes; you're probably looking at something that's going to be on-screen a lot of the time, that the player is going to be able to inspect closely, something that has the opportunity to really wow them. This is the ideal place for a custom shader. One extra draw call isn't going to break anything.
But, maybe your shader programmers disagree-- maybe they're dogmatic, maybe they just don't want to write a new shader. You can probably do a few things to simulate the tools you'd need for mother-of-pearl, without using anything that would make fundamentalists angry.
A sharp environment map is specular. Drop your specular component and color your cube map and you have colored (blinn-phong, NdotH^n) specular. You lose dynamic lights, of course. You can regain dynamic lighting on your specular if you can rotate your cube map to the light vector. You can do this in HLSL, but I don't know if you can do it with your nodes setup. (You can create Blinn-Phong specular on any number of light sources by adding up any number of rotated cube map lookups, but these rotations are more expensive than just doing the pow().)
Specular masking should be replaceable with multiple meshes with complementary alpha channels. Might be a few problems if you don't use add blending for the layers (see below), but probably not noticeable.
Multiple directions of anisotropy can be managed in HLSL with multiple meshes or multiple UV coordinates, if you have them. Pretty much the same thing anyways.
In the case of your source image, I would want to use additive blending. Can you do that? You have a nearly transparent surface on which you're applying white specular / environment map, and the difference in depth between that and the mother-of-pearl fracture layers is part of what gives it its look. If you can do add blending, you can have multiple, simple fracture planes placed inside of the handle, each receiving different specular from a differently colored environment map, and probably contributing only specular, followed by the highlights of the surface. Would be best to have a hard black backing for your add blends, not present in the source image but I think the concept would permit it.
Normal maps are appropriate for the fracture planes, and probably necessary to get good color from a simple mesh and these specular work-arounds.
All of that is probably more expensive computationally than just creating a custom shader that actually makes good mother-of-pearl, though. It's just a bunch of dumb workarounds for a dogmatic workplace.
Thanks!
Also looks like there's some hue shift based either on the underlying terrain texture, or some other global position sampled map.
These "circular" scratches are in reality a shit ton of crosshatching scratches sitting on a spherical surface. Since they're created when something is scratched against the surface with a direction, the direction of material is changed on a micro level. It's almost like there's fibers being created, all of them pointing in the direction of how the scratch was first created. Soooo, the light response is bent in that direction. It's essentially an anisotropic light response.