Making Of 'Old Man' by Oskar Swierad and Kamil Makowski

In this tutorial we will explain how we have created the artwork "Old man - Portrait". At the beginning - modelling
and unwrapping was done in Blender, detailing and rendering in Zbrush and some post production with Photoshop. You must be a bit familiar with these tools, the more advanced artists can do first and thelast part of this tutorial in different application. Ready? Let's start.

First add a plane and delete three verticles. Using the "Extrude" option create basic eyelids shape.

Now select all verticles, and extrude them. Then move single verticles to get a proper shape, repeat this action a few times to receive somethink like this.

At this point select the external edges and start modeling the rest of the face using "Extrude".

You have half of a head at the moment. Create the ear and join it to the head.

Select all verticles, duplicate them, symetrize by "Global X", join the two halves (you can also use Mirror Modifier in modifiers stack).

When the main model is ready, you should do UV Mapping. In Blender it can be done verysimply and fast. Just mark thee seams like shown below, go to UV Face Select mode, press [U] in 3D View and choose "Unwrap".

Export the model to (Wavefront-) OBJ file (if you have the Linux version, download the script from Blender.org -> Python
Scripts). Import it to ZBrush, check the mapping ("Uv Check", then "Fix Seams").

Swordmaster_eBook

Swordmaster_eBook

Now it's time to give the head its basic shape. Multiply the polygon amount by pressing "Tool -> Geometry -> Divide" and use the "Inflat" tool ("Transform -> Inflate"; in "Edit" mode).

Each time you divide the mesh, add bolder and precise details. Select "Pinch" tool and make wrinkles on the forehead, at the lips and near the chin. With "Inflat" fill spaces below them. You can "Smooth" the sharper places. Your 3D View should look like this:

Save the head for safety. ("Tool -> Export", DON'T save it as a Document!) Let's paint the texture. Turn off "Edit", press "Undo" (Ctrl Z) to erase the model, load the reference image, then press "Crop and Fill" and again load the head model. Press "Edit". Now use the "Move" and "Scale" to match it to the reference. Create new texture in high resolution, load "Image Plane" script (you can download it and get help from zbrushcentral.com). Go to "Master Projection", keep the "Color" option turned on. Move the texture till you have the proper effect, then press again "Master Projection" with "Color" marked.

Part of the head is already textured. Go again to "Master Projection" to correct the image. Select "Simple Brush" tool, choose alphamap named "Brush 01", select "DragRect" stroke and "Import" some skin photos. Set "RGB" to [100], turn off "ZAdd", put textures on model and repeat the process, until you have the whole head painted.

Now ZBrush will generate the details. Go to "Tools -> Masking", press Int, turn off texture displaying for a while. Set "Deformation -> Inflate" to about [30]-[40], "... -> Smooth" to [10] and uncheck "Masking -> View Mask"

It's done, time for render! :) I've used shader and light "Sokar" (zbrushcentral.com) with some modifications. Go to the "Render" panel, turn on the "Shadows" option and press "Best".

In the same panel, press "Bake", create a new layer. Put two white spheres as eyes and render again.

Our render looks like below:

Time for some postprocessing. For educational purposes we will do the bigger part in ZBrush ;] Before we begin, adjust the colour values in Photoshop, it's much easier than with ZBrush's curves. Check the resolution and try to remember, then save image as PSD file.

Open ZBrush again, "Resize" the canvas, then "Document -> Import" Photoshop-corrected render. If canvas' and file's resolution aren't the same, the last one will be scaled (good algorithm, but blurry effect). In "Render" panel, select "Flat" type.

Take the tablet pen (with mouse this part will be MUCH harder), select "Highlighter Brush" tool, turn off "ZAdd" (and do this each time we will select any tool), set RGB Intensity to [20], Draw Size to [20] or lower and highlight the eyes where it's necessary - they should look 'more three-dimensional'
;)

Then choose "Simple Brush", set Draw Size to very small, eg. [3]. Use harder alpha then default ("Focal Shift" to soften too hard alpha). Set RGB Intensity to [70], select very dark green shade, and paint basis of iris.

Next, Draw Size to [1], RGB Intensity to [10]-[40], stroke "Freehand" (lowest possible spacing), and green shade about center of palette. Draw the lighter parts of iris, and pupil. Highlight the lower part of iris, then add a reflection with RGB Intensity [100] hard "Simple Brush".

Colorize a bit left and right side of eye with "Colorize Brush", choose a tint close to orange. With "Simple Brush"/"Drag Rect" and proper alphamap add some details.

Now you should correct the shadows on the skin (fake SubSurface Scattering, SSS). The best tools for such operations are "Highlighter Brush II" and "Shading Enhancer". Use default alpha, with Focal Shift set to [50]. Save your document and experiment.

Done? OK. Draw details, if you have forgotten about some parts (eg. the top of head or ears) and we can continue. Select "Simple Brush"/"Freehand" (lowest spacing), draw size ~[35]. RGB Intensity [8]. Alpha should be set to "Brush 32", Focal Shift to [75].

With dark brown color, make strong shadows on brows and lighter above ears. Change alpha to "Brush 31", RGB Int. to [3], Draw Size to [10]-[20]. tint to reddish gray.

Paint the "darker" fibers, then highlight hair and brows with light gray. If some parts are too bright, touch them up with shade from dark part of head ("Picker", then drag field near "Once Col" to the 3D View). It's all about drawing and correcting :) You can also load a texture, and use with with "Freehand" or "Drag Rect".

Finished! :D We've lowered red saturation a bit, and used Unsharp Mask filter.

Fetching comments...

Post a comment