As promised, I wanted to follow up on my Eames series with a few pointers about the materials created for the chairs. This is the first in a series of mini how-to tuts about materials.
The Eames chairs are a good place to start. They all use materials that are meant to show some age, a certain vintage quality and patina, yet are as simple as materials can get. They are a good way to demonstrate how materials can be made to look realistic without necessarily using complex, multi-layered setups that would take a long time to render. I will explore more complex materials later on (indeed, I’m planning a post just about multi-layered mats), but, for the sake of economy, I only use them in my work to achieve effects that cannot be obtained with single-layer mats.
For this series, I have cobbled together a material test scene. I may upload it for free to TS if there is interest, but there really is nothing to it. For illumination, it uses an HDR map and two light planes (left and above). HDRIs in material testing scenes are anathema to purists because they give the mats a non-neutral colour cast. I don’t care. I want my materials to look the way they would in a fully constructed scene and an HDRI map is the most efficient way to recreate a plausible physical environment for the ball to to reflect and refract. In any case, I would trust such a material scene vastly more than I would the mat editor’s preview, which has a way of looking very much unlike what the mat will look like in the scene.
A few points before looking at the specific mats.
1. Material creation should always be done by using real-world reference. Try to look at the actual objects whose features you are trying to re-create, or at least look at several photos taken in different light conditions. Do not create materials just from memory or by looking at other artists’ renders. Also, do not hesitate to ask colleagues, partners or your kids about what they think of your materials – a fresh pair of eyes, possibly belonging to someone who has little to do with CG, is the best way to tell you if you are on the right track.
2. Be as subtle as you can be. When fine-tuning a specular or bump map, find a setting you like, then take the effect down one or two notches. Some things should only be suggested.
3. Do not feel like you have to use variations of the same map in all your slots. Real life does not always work like that (although it sometime does). Add extra, discrete layers to your glossiness and bump maps – they will only add to the material’s realism.
4. Even when doing single-layer materials, don’t be afraid of layering your maps by using Max’s composite map or Vray’s blended textures. These multilayered maps are a good way of masking repetitions in tiled bitmaps and can help you conserve memory by deriving many different looks from a limited number of maps.
Eames plastic red
This is one of the simplest materials in the scene, used on the plastic version of the chair shells (for reference, the ball is 40x40cm). Here is how it looks like in Slate, including a few details about the diffuse slot:
I could have used a plain colour here, or a VrayColor if I needed float values. The reason I used a bitmap instead was that I wanted to be able to easily switch the colour of the chair to one of a few pre-determined hues. So I painted a small colour swatch in Photoshop showing all the available tints for these chair models and used the CroppingPlacement took in the bitmap editor in Max to isolate only one colour – in this particular case a slightly orangey red.
The next step was to add a very subtle imperfection to the surface. I didn’t want it to be visible in the diffuse, only in the specular and glossiness slots, which is enough to give a slightly aged, real-world feel to the material. Here is a low-res version of the map I used in the specular slot (note that I only used unfiltered bitmaps, regardless of the slot, even though that can lead to longer render times):
You see that I used a much lighter, low-contrast version of the same map in the glossiness map. That’s what you want to do, but you don’t have to do it that way. Most of the time, you can get along by using a colour-correct map to extract a lighter, lower-contrast version of the spec map, which you can then plug into the glossiness slot.
Lastly, I opted for a procedural noise map to replicate the rugged, high-frequency bump on the plastic shell. Here is the setup (note the very low scale):
Here are the details about the material proper. I’ve marked in red the important bits. These include Fresnel reflection on – as it should be for all materials except very reflective metals; Reflection subdivs of 64 – which can be lowered depending on your render settings; reflection cut-off value of 0,001 (instead of the default 0,01), which will ensure accurate reflections when working in linear space; specular, glossiness and bump map contributions of 40, 55 and 18 per cent respectively – essentially a way to fine-tune how heavily the map will affect the final result.
Here is what the material looks like on a real model:
And here is how it looks in black, which tells us one important thing – darker materials look more reflective:
This material (designed for the vintage version of the chair as environmentally-unfriendly fibreglass was originally used before it became technically possible to cast entire plasic shells in one go) was derived from the plastic material with the following tweaks:
1. A dirtier diffuse made by merging the colour bitmap with a dirt map using a composite map set to a “multiply” blending mode at 75 per cent;
2. A lower glossiness for sharper reflection and a slightly higher specularity;
3. A “fibreglass” bump map – actually a normal map derived from photos of real vintage chair shells (I only used the fibreglass map in the bump slot because it was too repetitive to be used in the diffuse slot, which I should ideally have done). There are many different Photoshop plugins and standalone applications to derive normal maps from photos. One cost-effective but bare-bone standalone solution is Shadermap. Another, costlier but integrated in Photoshop and endowed with more controls is Pixplant. Note that normal maps should be loaded as bitmaps with a gamma of 1.0 and plugged into a Normalbump or VrayNormal maps before being plugged into the bump slot of the material.
4. The dirt, specular and glossiness maps are the same that were used in the plastic material.
Here is how the material looks like in slate:
And in the editor:
Sharp and painted metals
This example shows how minimal changes can yield completely different looking materials. The main difference between these two is the absence of fresnel reflection (for the chrome mat) and a dedicated bump map (for the black paint mat). Otherwise, they share most of their characteristics, as you can see here:
This is the kind of material that succeeds or fails based on the quality of the textures. I used three here:
1. A specular map designed to give a very, very subtle variation in reflectivity across the surface;
2. a glossiness map, basically scratches, finger- and handprints painted with custom brushes in Photoshop;
3. a bump map to “break” the surface of the painted material (note how the bright spots on the map read like the grains of sand or dust you often see on shoddy paint jobs).
Here are the two materials in the editor. Note that I’m using a Ward this time, which works better for metals. Also, you will notice that the contribution of the glossiness map is extremely low – you really want this to be very subtle. If you cannot make it subtle enough, try loading the bitmap with a gamma of one, which will make it even fainter.
Old, tired metal
I needed an older metal for my vintage chairs. This one was totally over-the-top, but I quite liked the effect in the end so I decided to include it here, just as an example of the interesting effects you can obtain with single-layered materials and the right bitmaps.
Here are the details:
A few things to note:
1. Fresnel reflection is not active, yet the reflectivity of the metal is low, giving it a dustier look. The map only contributes 15 per cent of specularity, with the specularity colour contributing the rest. Mixing bitmaps and colours to generate the overall reflectivity will make the effect more subtle.
2. The diffuse map was desaturated via a colour-correction map before use.
3. The specular/glossiness maps are the same and, like the normal map, were derived from an inverted greyscale version of the diffuse (the white scratches had to be indented and unreflective, not prominent and reflective).
Here is the maps I used:
That’s it for today. I hope this was of some interest. I’ll be back shortly with a look at more complex materials. Of course, feel free to ask if you would like to know about a specific material or effect.