BPR Rendering Workflow in ZBrush 4 R2

After posting the image Angry Faun on ZBrush Central, I got a lot of questions about my render and comp workflow. This guide was written while I was using ZBrush 4 R2 and rendering with BPR Render. All compositing was done in Photoshop.

Environment Light

Press the texture button in Light > Background to import the image that you want as a base to create your lights from. Keep the HDRi in the texture slot so it can be used as environment reflection in materials (Fig.01).

Fig. 01 id_Fig

Fig. 01 id_Fig

If you are making an environment pass, make sure samples in Light > Background are about 2-3 to create a very ambient lighting. The idea is to simulate bounced light from the environments, so you don't want a very specular light (Fig.02).

Fig. 02 id_Fig

Fig. 02 id_Fig

Adjust the exposure etc., if you need to get a good intensity on your lighting. Don't forget to activate Use Material Curves to make sure that your diffuse and specular curves on the materials should be used while rendering. Otherwise the materials will share the same curves. You can find theses settings in Light > LightCap adjustment (Fig.03).

Fig. 03 id_Fig

Fig. 03 id_Fig

As you can see in Fig.04, there's a shaded pass on the bottom. The Exposure layer controls light intensity, while the Photo Filter layer gives it a slight green tint. On the top is a solid color adjustment layer with Normal as the blending mode. It is masked with an inverted version of the shadow pass from the environment light/light cap. Use Alt + click just below each of the three top layers to make them only affect the layer on the bottom of this stack and no layers further down.

Fig. 04 id_Fig

Fig. 04 id_Fig

The current pass can be seen in Fig.05, and the combined passes in Fig.06.

Fig.05

Fig.05

Fig.06

Fig.06

Key Light

This pass was rendered with a normal light. This light is supposed to be the main light. Don't forget to save out your light cap (for later access) and then delete all lights in the light cap. The reason is that the light cap light and shadow pass should not affect your other render passes (Fig.07).

Fig. 07 id_Fig

Fig. 07 id_Fig

I wanted to render this image with a very indirect lighting and indistinct shadows. Therefore I set the blur on the shadow quite high. Since the blur is quite high I could use a low resolution on the shadows (Fig.08).

Fig. 08 id_Fig

Fig. 08 id_Fig

Here I had another shaded pass on the bottom with Add as the blending mode and a Photo Filter layer to give it a warm, orange tint. On top is a solid color adjustment layer with Normal as the blending mode. It is masked with an inverted version of the shadow pass (Fig.09).

Fig. 09 id_Fig

Fig. 09 id_Fig

The current pass can be seen in Fig.10 and the combined passed in Fig.11.

Fig.10

Fig.10

Fig.11

Fig.11

Rim Light

This pass was rendered with a normal light. The function of this light is to show the shape of the character, since it comes in from the side/back (Fig.12).

Fig. 12 d_Fig

Fig. 12 d_Fig

This light has a high amount of blur, just as the key light. It is a bit unrealistic that this light hits some of the parts of the right side of the face (camera point of view), but I think it adds to the definition of the forms in the face (Fig.13).

Fig. 13 d_Fig

Fig. 13 d_Fig

Again there's a shaded pass on the bottom with Add as the blending mode. Exposure controls the light intensity and Photo Filter gives it a cold, blue tint. On top is a solid color adjustment layer with Normal as the blending mode. It is masked with an inverted version of the shadow pass (Fig.14).

Fig. 14 d_Fig

Fig. 14 d_Fig

The current pass can be seen in Fig.15 and the combined passed in Fig.16.

Fig.15

Fig.15

Fig.16

Fig.16

Ambient Occlusion

To render AO you need at least one light activated that can cast shadows. Don't forget to activate AOcclusion in the render settings (Fig.17).

Fig. 17 d_Fig

Fig. 17 d_Fig

I prefer an AO with height definition of the shape of the character. Therefore I usually lower the blur radius to 1. Gamma is nice to lower as well to get a broader AO. Think of it as longer raytrace in mental ray. In ZBrush AO is created with an array of depth maps though (Fig.18).

Fig. 18 d_Fig

Fig. 18 d_Fig

For this comp I set the AO with Multiply as the blending mode. I usually put AO further down in the layer stack and only let it affect the environment light. Letting the AO multiply the rim and key lights can make the model look dirty. In this case I just thought it looked better to put it on top of the key and rim lights (Fig.19).

Fig. 19 d_Fig

Fig. 19 d_Fig

The current pass can be seen in Fig.20 and the combined passed in Fig.21.

Fig.20

Fig.20

Fig.21

Fig.21

Cavity Map

Instead of baking the cavity map into your poly painting you can create a material for this to get greater freedom later on in the comp. Create a white texture and put it in the texture slot. Copy my values and play with them. Make sure you are using a MatCap type material (Fig.22).

Fig. 22 d_Fig

Fig. 22 d_Fig

Turn off shadows and AOcclusion in the render settings (Fig.23).

Fig. 23 d_Fig

Fig. 23 d_Fig

The cavity pass was set to with the blending mode Overlay, which means that black multiplies, white adds and 50% gray does not affect anything at all. The color balance adds a red tint to the darker areas to keep a better SSS-look on the skin. Keep the opacity low as this effect can be quite drastic. This effect has a tendency to make the sculpt look sharper and brings out the details (Fig.24).

Fig. 24 d_Fig

Fig. 24 d_Fig

The current pass can be seen in Fig.25 and the combined passed in Fig.26.

Fig.25

Fig.25

Fig.26

Fig.26

Fog and Light Correction

Depth is always rendered and can be found in Render > BPR render passes (Fig.27).

Fig. 27 d_Fig

Fig. 27 d_Fig

The top layer is a constant color (blending mode Normal) with the inverted depth as a mask. It is a very subtle effect with low opacity.

The lower layer is a constant color (blending mode Normal) where I painted a mask to darken the torso in order to bring out the head visually. The reason for not multiplying this layer is to not reach black in the image. I prefer to have my lowest values around 2-5 in a dark image (Fig.28).

Fig. 28 d_Fig

Fig. 28 d_Fig

The current pass can be seen in Fig.29 and the combined passed in Fig.30.

Fig.29

Fig.29

Fig.30

Fig.30

Lens Blur

Time to add lens blur in Photoshop. In order to do this you need to bake/combine all layers into one. We want to keep the order of the old layers so we can go back and do changes.

Combine: Create a new layer on top of the entire stack and press Ctrl + Shift + Alt + E to merge all visible layers into the active layer.

Put depth pass in a new channel: Copy the layer with the rendered depth pass. Go to Window > Channels and create a new channel. Then paste the depth information into this channel and name it "Z" or "Depth" so you know which channel is which (Fig.31).

Fig. 31 d_Fig

Fig. 31 d_Fig

Apply lens blur: Select the combined layer and choose Filter > Blur > Lens Blur. When the new window opens you should choose the channel you created as a source. Photoshop will blur depending on the luminosity in that channel. A good way to choose blur focal distance is to click on the image itself. Radius sets the amount of blur. There are a lot of other variables that you can play with to get the look you are going for. Generally you should apply a bigger amount of blur if the camera is really close to your model (like a close-up on a flower or bug) (Fig.32).

Fig. 32 d_Fig

Fig. 32 d_Fig

Fig.33 shows the model before the lens blur and Fig.34 shows it after.

Fig.33

Fig.33

Fig.34

Fig.34

Paintover

If you are making a still image, some painting in Photoshop can really contribute to a more detailed look. Don't forget to do it on separate layer(s). The idea is that you can reach it a lot faster by painting, than doing it in 3D. The downside is that it doesn't work with an animation or turntable (Fig.35).

Fig. 35 d_Fig

Fig. 35 d_Fig

Fig.36 shows the model before the paintover and Fig.37 shows it after.

Fig.36

Fig.36

Fig.37

Fig.37

Chromatic Aberration

Chromatic aberration is an artefact that can be found in real world photography. In Photoshop there is a tool to get rid of that, but since we want to make our clean CG image look more "real", we will use this tool to add chromatic aberration.

Basically what happens is that when red, blue and green light travels through a lens they will disperse in different directions and then hit the film of the camera in different places. Thanks to Robert Krupa and Gustav Tell at Milford Film and Animation who told me all about this artefact.

Combine visual layers: Add a new layer and combine all the visible layers on to it (Ctrl + Shift + Alt + E).

Add chromatic aberration: Choose Filter > Lens Correction. It's a really subtle effect, but I think it adds to the overall look in a good way (Fig.38 - 39).

Fig. 38 d_Fig

Fig. 38 d_Fig

Fig. 39 d_Fig

Fig. 39 d_Fig

Fig.40 - 41 show the before and after shots.

Fig.40

Fig.40

Fig.41

Fig.41

Unsharp Mask

In order to make the image "pop" I usually add a sharpening effect in order to bring out the details and make the image easier to read. I usually use an unsharp mask, but it is possible that other sharpening filters are better. It's all up to you.

Combine: Create a new layer on top of the entire stack and press Ctrl + Shift + Alt + E to merge all visible layers into the active layer.

Apply an unsharp mask: Choose Filter > Sharpen > Unsharp Mask. I usually set amount quite high and later control the amount with the layer opacity. Radius should be below one pixel in my opinion (Fig.42 - 43).

Fig. 42 d_Fig

Fig. 42 d_Fig

Fig. 43 d_Fig

Fig. 43 d_Fig

Fig.44 - 45 show the before and after shots.

Fig.44

Fig.44

Fig.45

Fig.45

Grading and Noise

After applying all the light passes it is always good to do a final grading. It's like the last tuning of the colors and intensity of the image. Adding noise to the image adds to the realism as well.

Add noise: Create a new layer and fill it with 50% gray. Then go to Filter > Noise > Add Noise. Set that layer's blending mode to Overlay. Play with opacity and make it subtle (Fig.46).

Fig. 46 d_Fig

Fig. 46 d_Fig

Add grading/adjustment layers: Add a Curves adjustment layer (Fig.47) and then edit the curves as follows:

• Curve 1: Blending mode = Normal. This curve generally brightens the image.

• Curve 2: Blending mode = Normal. Here I decided I wanted more brightness. It could have been done in Curve 1, but I thought I would have more control with a second curve.

• Curve 3: Blending mode = Color. This curve is where a lot of magic happens, in my opinion. I adjusted it so blue and green is added to the darker parts of the image, and blue and green is removed from the brighter parts. The effect is that you get a really dynamic color spectrum to the image and somehow it does not look so "flat" anymore (Fig.48 - 49).

Fig. 47 d_Fig

Fig. 47 d_Fig

Fig. 48 d_Fig

Fig. 48 d_Fig

Fig. 49 d_Fig

Fig. 49 d_Fig

Fig.50 - 51 show the before and after shots.

Fig.50

Fig.50

Fig.51

Fig.51

To see more by Daniel Bystedt, check out ZBrush Characters & Creatures

Fetching comments...

Post a comment