Wardrones: 3D illustration techniques (part 3) – lighting & rendering in KeyShot
In the final part of this tutorial, I am going to take the previously assembled and painted mesh, and move it into KeyShot for scene assembly and rendering. Using multiple different layers, or renders with different materials and textures, I will provide material for compositing in Photoshop, giving us total control of how we want to texture this robot in post-processing.
I will instruct you on the types of layers that you can add and how to successfully composite them to look not only flashy and brand new for concept and product shots, but to add dirt and wear, depicting a more realistic scene in combat. What we want to achieve is total control of the outcome, without having set any UVWs or correct topology.
ZBrush to KeyShot
The obvious first step will be to correctly export this mesh into KeyShot, while utilizing all the masks we defined in Part 2. Before we move on, you should definitely consider decimating your model to a lower polycount, unless you feel cocky about your machine today, and you’re willing to test its guts.
It’s up to you, but I’d advise not to play with the devil, because our model has up to a hundred million poly’s at this point. Now, ZBrush luckily has a built in function to transfer the ZBrush Model along with its materials into KeyShot, which is what I’ve been counting on this whole time. Just go to Render > External Renderer and make sure to check the following options:
- Auto Merge
- Groups by Materials
Now, click the render icon in the corner of your ZBrush document and you should have it load into KeyShot.
A tedious task indeed – unfortunately the second step includes selecting all of the groups and once again assigning the appropriate material inside KeyShot, as well as doing manual adjustments to some of them to fit the look we’re developing.
A good way to do this is to hide everything and then unhide and select all parts that share the same material and then assign a linked material. Then proceed with the next group and so on.
Retrospectively, I’m sure there are ways to group these differently before exporting into KeyShot, so you can experiment with merging all of the SubTools before exporting to KeyShot, something I may have overlooked.
The types of materials you are already familiarized with, but the way they are distributed in KeyShot is in groups by type: Metals, Paint, Plastics, Emissive (Light), Leather, and Fabrics and so forth.
You are going to use different types of paint and metal on different parts, and to decide whether to use rough or shiny versions of each material, matte or metallic paint, and by similar comparison. A lot of this is already decided and I’ve talked about extensively in Part 2.
Still, you’re given another layer of decision making to improve your design choices. With this tutorial I will assume you are familiar with the basic UI of KeyShot and how to do these very basic actions. The result of this step is going to give us a very clean and product showcase type rendering, but the truth is, this is only the first layer of our composite – the foundation.
Lights & environment
Now that we have a good looking foundation, we can address the lighting a bit and start setting the mood. For this particular model, I’m not looking for something too shiny, too studio like – my goal is to set some kind of dangerous doomsday scenario mood, so as far as the environment is concerned, I want something to only roughly depict some kind of soft urban environment with no harsh lights.
KeyShot has some great HDRI templates (although you can always use your own or custom craft one). In this case I’m going to use the one located in Outdoors library named “Sports Hall.” To tune the lighting properly, I am going to rotate the environment at an angle of 250-degrees to get a soft light coming from the right-side of the mesh and having the rest only lit by secondary bounces. You should also lift the model using the Move tools, as it is a drone, and we don’t want it sitting on the ground. This will be a great foundation for adding more lights to accentuate some detail and add a dramatic effect.
Once this HDRI light is setup you can discard the environment in the preview, and instead use the option to select a color for the background. Set it to black, but slightly lighter than pitch black. This is a good color to isolate the model and decide what to do with it, without visual interference. A pure black color blends badly with the shadows.
This is how I will do things in this particular case, but there are many other ways to build your scene. I’d like to focus on the model first and foremost, and base the rest of the scene and its decision making on what the model tells me by reacting to certain experimental play.
Now, I want to add some lights to accentuate the details and dramatize the picture, emphasizing the doomsday scenario and mood. Nothing looks better on sci-fi models than harsh dynamic lighting to emphasize the mechanical parts nicely, and make good use of these reflective materials.
In this case I’m going to use sort of a 3-point light setup, having one red light from the right, one regular fill light from the left, and a rim light coming from the back. I also want to experiment with the color of the rim light and see if I can make it work having some sort of a neon teal to it.
Perhaps I’d like to add more color to the moody pic, and give it a cyberpunk feel, having the red narrate sort of a nearby presence of an alert light, also depicting urgency and danger, with the teal communicating neon signs overhead in an urban cyberpunk city. I’m going to explore this direction and perhaps I find that it doesn’t work.
A fourth light is also added under the model to create some sort of enhanced dynamic in the reflections. The way I create lights in this case is just by duplicating existing meshes from within the model, assigning a Spotlight material from the Emissive material library folder, and then using the Move tool for placement, as well as adjusting some of the settings to get the best look. You can also create lights by adding pins in the environment editor.
One thing you’d have to think about when creating this sort of fast type of illustration is the environment. While it’s a totally viable option to create several quick meshes that sort of give us enough silhouette to work with in post, you can also totally utilize free stock imagery and Photoshop tricks to create the illusion of the background.
This coupled with lighting will be enough to communicate the background without actually having one. Another time saver.
Now here’s the deal. I’ve made an excessive amount of lights here, but in the following steps I might choose to remove some as the scene evolves. This is an iterative process as always, and one should not be too attached to its trials.
A little bit of foreplay with the settings is good to utilize in order to make better judgements when deciding the lighting and setup. The more finalized your preview, the more flaws you can draw out and fix. These settings include general Render settings, Camera settings, and Effects.
Basically almost whenever I start a project I do a sweep of these settings and put everything in check. Render Settings > Turn Global Illumination ON, Self Shadows ON and Ground illumination ON (you can use the Product Template, or other templates for different projects).
Image Adjustments > Pretty much everything is good to add at least slightly. For most renders, Vignetting and Bloom is a must.
Camera > Adjusting the FOV for a more dynamic shot, and playing with the Depth of Field settings. The resolution and aspect ratio is irrelevant for now.
Multi Shadow Clone Jutsu
Like at the pinnacle of the first Naruto episode, when all things seem to fall in desperation, you – the main hero of the story, are going to discover something great – something that is going to change the outcome of your project, an ability only discovered at the brink of desperation – it’s called the Multi Shadow Clone Jutsu. It’s time to utilize the ultimate cheat.
I just couldn’t resist
After I discovered my new Jutsu, I just couldn’t resist to implement one final killing blow to the opponent called mediocrity. I slay the opponent by creating a very quick city background mesh using only insert meshes on a plane. This type of scene literally takes 5-10 minutes to create, and you are about to see the power of such.
Keep in mind this mesh is going to be very obscure in the final image, but does a great deal to create more value. Another thing worth noting is that adding such large objects to the scene distorts the scaling, as well as messes with our reflections, so we will render the background separately, otherwise our previously made lights won’t work correctly. This is something I encountered while working on this project.
Render passes – Background
To keep things short – the background will have several render passes. I’m rendering at a larger resolution than will be necessary – this is always a good idea to have additional sharpness and room for non-destructive editing. Also always export at 16-bit TIFF or PSD so that you can have more color real estate.
I render based on maximum time. For my machine 20 minutes is more than enough for good results at 4000 x 2500, the chosen resolution at this point. The first layer is the regular lit and shaded layer.
The second layer is just a dust pass. A dust pass is a matte material applied to everything in the scene, and adding a bit of color noise in the diffuse color – making it look sandy / like dirt. Or you can just use a dirt texture.
The third layer is Ambient Occlusion, and we may use this as a mask for dust, as well as other uses. The fourth pass is a Depth pass, and we will simulate this by applying a Black to White gradient ramp into the diffuse slot of a Flat material, with planar mapping. This way by adjusting the gradient you can set where you want your falloff or fog to begin.
Render Passes – Drones
Now this is the main part and we want to have as many passes as possible, that we will utilize both as masks, and shader variations. To start with the basics: Main beauty pass – Regular render pass with the scene shaders and lighting. Ambient Occlusion Pass – Used for masking and additional shadow control.
Dirt Pass – Explained in previous step, applied to drones.
Texture Mask pass – By utilizing a black-and-white grunge texture and applying it to a flat material, covering the whole scene, you will create a 3D curved mask for adding additional texture.
Edge Detection Pass – Using the edge detection material, you will create another mask for edges that will be used to control texture.
Full-metal mask – Applying the basic metal you used on some of the drone parts to the whole scene will allow you to control paint damage via texture and edge detection masks.
Material ID – Flat color for each material, for further adjustments.
Once you’ve rendered all of these passes, you are ready to head out to Photoshop for the final composite.
Into the abyss…
It’s time to start layering in Photoshop. The first logical step is to arrange the beauty passes and ideally put them in separate folders since they will have a lot of edits. From here on after it’s a straightforward process.
I will layer these textures and masks in a logical way starting from paint wear and tear, then layering dust and onward. Utilizing all the masks at your disposal you can pretty much live-texture these drones.
For wear and tear combining the grunge mask with a harsh threshold, then filtering through the edge detection mask you should get something very close to wear and tear on the edges.
Feel free to customize the mask by hand to remove any unwanted effects. For the Dust map, you will use AO as a mask as one layer, and then another very thin (transparent) layer by using the grunge mask to add some surface noise. Then you can play with AO, even use AO as a color adjustment mask to create other effects.
You can overlay any type of grunge texture, rusted metal and so forth, and use the grunge mask to distribute it properly and create a 3D illusion. This is where the beauty of this method starts. Using these masks you have absolute control to texture your model live in Photoshop and experiment with more power than you can do with usual UVWs and texturing rules, since you are venturing into the 2D realm.
A lot of these choices are up to the artist, and experimentation is a good way to get the result you’re looking for. The first step in our post processing is creating a desired final render – effect free. You are essentially creating the raw render that you would otherwise get by months of work.
Do the background as a separate raw render, and try to get the most out of it, before we turn to actual level adjustments that will fade this background into the, well, background
Before you move into the final steps of the post-processing stage, you need to gather some assets. Your most utilized assets in this type of scenery will be Clouds / Fog transparent elements that you can either find online or create, or both, as well as fire and smoke elements, explosions and perhaps some mountain and sky images. Pixabay.com is a good resource for free stock photography that I often use.
Before moving on, gather the resources and get ready to layer them. First, you must create a foundation, and this means setting up the proper levels. Starting from the furthest point in the background, you will adjust the levels to create a proper sense of depth and scale. This is a crucial point that must be done correctly or the background will end up looking laughable. I suggest adding a sky and some mountains in the far back just to set the maximum depth level, and then as you adjust levels and scale it from there. The background should be very limited in its dynamic range, meaning you should squash the min and max levels, but create some contrast for the shadows to push through. This is an intuitive thing and you must do it in a comparative manner until you get it to look just right.
Other than the background, there are drones located at different depth levels, so using the depth mask we’ve created to mask a level adjustment layer you can do the same – alter the levels as they push into the background.
After you’ve set the levels, it’s time to add some texture to the shot, mainly using these smoky assets you’ve made and layering them using the Screen modifier. You can also play around with creating fires and explosions in the buildings and adding smoke areas.
What you don’t want to forget is that this is only the background so you don’t want it to overshadow the main focal point, but you still need it to be present to create depth. This is sometimes a tricky maneuver if your picture has unusual lighting.
This last step is where you can truly utilize your creative freedom to shape this image and drastically affect its mood. I am going to add a final layer of effects such as volumetric lighting, artificial glow, lens flares, and do the necessary color adjustments to push this image to its final state.
Keep in mind, you can play around with different layering effects and create more than the standard round of adjustments. I really love playing with shadow colors and gradient passes, trying to achieve some kind of unusual look, but it isn’t necessarily going to work well on everything. Let the image speak for itself – what does it want to become? What is the greatest thing it can become? That’s where you should take it.
Some of the usual adjustments include levels, brightness / contrast, color balance, but I also like playing with the color lookup presets and layering them at lower opacity. Once you’re done with the adjustments you can flatten the image, rescale it to a desired resolution, and add a final percentage of film grain to smoothen things out.
When it comes to grain it can be quite useful when you’re going too far with effects like color dodging with saturated and intense colors – putting a percentage of grain on every layer will dissipate the color artifacts – these are less visible in a 16-bit image, but it doesn’t hurt to have more polish.
With this three-part series, you’ve accomplished creating a full blown 3D illustration with minimal effort and investment in the model production pipeline, yet creating great results as if to serve as a movie still. I’ve hopefully taught you how to get from A-Z in the most minimalistic way possible, with the highest output quality, and gave you food for thought along the way, as in how to approach the design.
Honestly, there is more that can be done with the model and the result could’ve gone a totally different route, but it is not bad for what it’s meant to be – a guide to show you how to utilize certain tools and techniques to create a polished looking 3D Illustration.
With evolving technology it’s getting easier and easier to save time, while producing great artwork will only get so much more in the future, but for now you can utilize the tools at your disposal to bring your ideas to life without working for months at a time. This is the conclusion to the series and I will allow you to dwell on the complexity of the final piece, while you remember – it only contains a single ZBrush sculpt.