Project: Raspberries 

Objective and Summary:

I have recently come across a few digital images of food and fruit that have inspired me for a digital creation. As was the case with my last project, I would still like to work further on my texturing, materials, lighting, and rendering.

My previous project, the French Press left me quite satisfied with my glass, lighting, and final renders. It was quite simple of an object though, and was not set in a scene. As well, I did not achieve quite the level of realism I was looking for in the texture of the metal.

So for this project, I am putting a heavier focus on my materials and texture. I intend to create raspberries, blueberries, and a few side props to depict a scene showcasing a dessert-like dish. I want to embody the feeling of having just made a dessert and enjoying it freshly made.

While I have avoided using Zbrush in the past, I do know how to use it roughly, and intend to utilize it in the creation of my textures.

Look Research:

This project has required a bit more initial research than my others. As such I’ve scoured the internet in search of good photographic research. Normally, I would take the photographs myself or have the objects in front of me - but at the moment, I cannot get any, so the internet will have to do.

I’ve found that the raspberries are semi-translucent and covered in small hairs. They are built of many small nodules stacked neatly on top of each other, arranged in rings. There is a uniformity to them, but the key to getting them to look right well be having each node unique. I do not want to fall into repetition. The translucence is not much, but it is enough that light can pass through it, seemingly illuminating the flesh from within - I will have to explore subsurface scattering (SSS) when I get to materials.

The blueberries are essentially spheres, but they are wider than they are tall, and have small pits at the top where they connect to the stem as well as a larger pit at the bottom that appears similar to a blossom in shape. The blueberries are far less uniform in color than raspberries however - coming in varying shades of blue, indigo, purple, and red. The skin is quite splotchy. There does not appear to be any transparency to them however, so SSS will not be necessary here, but again, avoiding repetition will be important to sell the short.

Cloth, of course, is made of many fibers woven together. There are many different styles to choose from, but I wanted to create a quilted look to the cloth. As well, I wanted the cloth to be folded upon itself.

The bowl is a simple design, inspired by Chinese Porcelain. I expect this will be the easiest piece to create.

For my cutting board, I have decided to meld two ideas. I distinctly like the worn, wooden look that cutting boards get over their lifetime. This will require a many small criss crossing grooves from knife strokes. As well, I really like the shape that pizza boards have with the rounded edges. So I intend to create a nice mix of both in the modeling and texture phase.

Spoons, I’ve found and have known, are of many different style, shapes and designs. I have found, and intend to create spoons that fall in between ornate and simple. I want the look to be semi-fancy, the type of cutlery that is not frequently used, but neither is it reserved for special occasions. I will pay careful attention to the tarnish and light smudges and fingerprints that are inevitable on metal surfaces.

I intend to utilize chocolate in a variety of ways: whole bar, broken bar, and melted. This will provide a nice contrast of textures and shape for me to play with as well. Chocolate tends to have a very glossy look, but with no reflection. Where the chocolate breaks however, there is a distinct roughness both in texture and in gloss, something I should be able to take care of easily enough in texturing.

Modeling:

Raspberries: I created the raspberries by starting off with a polygon sphere. I then manipulated the faces into a kernel shape that did not appear to be uniform or symmetrical. Once I created the initial kernel, I then duplicated that shape and manipulated the scale and various vertices and faces to create subtle differences. I followed this approach and built each tier, with around 8 - 12 kernels per level. Once I had my basic shape, I then went into smooth poly preview and manipulated faces to close any gaps between my objects that existed. With all gaps closed, I smoothed my objects and grouped them together to create one final object. I then used the model as my base, from which I created all the other raspberries. I modified the other raspberries with scale changes and deformers.

Blueberries: The blueberries were created via box modeling. I created a a divot in the top with four flaps that extend in a little. I then manipulated the shape with soft-select to reach my desired shape. I had intended to go into Zbrush with this, but as I had been discovering on this project, I was not yet comfortable enough to utilize my results. However, I was quite happy with my blueberry results.

Chocolate: I created the chocolate through multiple means. The solid bar was through simple box modeling which I then smoothed through subdiv-surfacing. The broken bar was modeling by using the solid bar as a base, and then cutting the model into separate pieces. I began sculpting melted chocolate with Zbrush, but due to lack of experience I was not happy with the results I was getting, so I abandoned this route and just stuck to the hard bars.

Cutting Board: The cutting board was also another simple design, I began with a cube which I then subdivided and moved faces and vertices to get the pizza paddle board kind of shape I was wanting to create.

Spoon: I attempted to model the spoon in Zbrush, but again I was not happy with my results. So instead, I fell back to box modeling and got a shape I was ok with.

Bowl: The bowl was constructed very simply through the use of cylinder. I used no special techniques here. As with all my models, I used subdiv-surfaces to smooth the model while keeping actual poly-count low.

Table Cloth: The cloth I decided to create procedurally inside Maya, as opposed to sculpting in Zbrush. I began by creating a table and placing objects and ridges for the cloth material to interact with. This did not create the folded effect I wanted however, so I played with the simulation to get my desired results. In the end, I found that angling the cloth and then dropping it onto the table creating the physics I wanted to result in the bunched up and folded cloth.

Texturing:

Raspberries: Texturing the raspberries was the first attempt at subsurface scattering that I’ve done. I began with a misss_fastskin_x and then changed the various subdermal colors to the red range. The top dermal layer I created a deep red, while the other two I kept to a lighter and softer pink color.

I applied Perlin Noise maps to the Diffuse weight, Specularity, and Bump maps. This provided a degree of variation to the material that helps the eye feel more comfortable looking at the raspberry as food, rather than plastic. 

I spent a long time tweaking the numerous sliders to get the exact look I wanted. 

I decided to forgo creating the hairs that cover raspberries because of the extreme render times I began to run into. 

Blueberries: For the blueberries, I also used a misss_fastskin_x. In my research, I came across some interesting things about blueberries. The skin is (obviously) blueish-black, however if cut open, the flesh is green. Yet, if squished, the juice is purplish-red. So in my material, I created a deep blue dermal layer, a green subdermal layer, and a red tertiary subdermal layer. I applied a color map for the diffuse node, and a Perlin Noise to the bump map to reach my final look.

Chocolate: I was careful with my chocolate material. Chocolate can have a very distinct look. I chose a Blinn material as my base to ensure I could create a shiny surface. However, while chocolate is shiny, it is not very reflective and seems to have a high amount of scatter. I dropped my eccentricity down about .42 and upped my specular roll off to .56, however I then dropped my reflectivity to 0, which provided the non-reflective sheen I wanted.

For the color, I chose a deep brown with a hint of red. This resulted in a shade that was something a little between dark chocolate and milk chocolate. To me, it looked like a higher-quality, rich chocolate. However, the uniform color was too perfect, and chocolate tends to have slight variation in it’s color due to temperature changes and cooling. As such, I applied a Perlin Noise to the specular color node which provided the right amount off variation.

For the broken texture, I added a rough stone image and multiplied it with an image of coffee bubbles. Initially, I tried to use a displacement map, but realized later on that my UVs were not set up optimally for this - which resulted in broken seams. So I instead went with a bump map, which seemed to work out fine.

Cutting Board: By the time I’d gotten around to texturing the cutting board, I’d decided to just forgo Zbrush entirely, I’d focus on it for a different project to really get the hang of it. So instead I just began with the mindset of utilizing photoshop to create my displacement maps and various texture maps.

I used three different wood textures to create my desired results. I used one for a base, and then I overlayed and multiplied the other layers with various adjustment layers to create a diffuse, specular, and displacement map, which I applied to a Blinn. 

Spoon: My intention for the spoon was for it be a tarnished silver. However, my final look seemed to look a bit more pewter-ish. I used a Blinn with a color map and normal map applied to get the final result.

Bowl: My references I used for the bowl were Chinese Porcelain. As such I needed to create a highly specular texture. While typically all white, I found some examples of off-white that I liked and decided to emulate. I created a cracked color layer and then used a layered material node within the hypergraph to give me more control. 

The layered material node is something I have not really experimented with before, but I have immediately found it to be beneficial. I then applied another layer with an ornate design that rests atop the underlaying porcelain texture. 

I used a Blinn to achieve the high specular, but I gave it a perlin noise map to give a slight bit of variation.

Table Cloth: I had thought at first that I would use Zbrush to create the texturing for the cloth, but after running the simulation, I decided that it would be best to just create it procedurally. It came out fine in the end, so I am happy with that. Because of the size of the table cloth and how close up the camera would be in the shot, I decided to use a 4k map instead of a more regular 2k map. Considering I intend to only take still shots, I thought this would be fine despite the increased render time.

I decided to go with a quilted look to the cloth. I used a cloth texture that I found online and then applied that to a displacement map which produced my desired results. I used a basic color pattern I created in photoshop, which I then applied to the diffuse layer.

Lighting / Rendering:

When it came to lighting, I sorta knew how I wanted to approach my shot. I began by experimenting with very basic lights, just an area light at first with no color. I know a little bit about linear workflow when it comes to lighting/render, but I will admit it’s not my strongest point - part of why I’m doing this.  So I went ahead and gave the light (and all subsequent lights) a quadratic decay.

I then proceeded to set up a three-point lighting system utilizing: area light, spot light, and point light. After checking out the initial look and assuring myself that I was happy with how the initial lighting looked, I then adjusted the colors of the lights to try to imitate life. The area light I gave a light blue light, to imitate the color a clear sky would give off; the spotlight I gave a orange-ish yellow to simulate the sun; and to the point light I gave a slightly more orange color to imitate an artificial light from behind.

I also used sub-div surfaces to achieve a smoother render without upping my actual geometry too much.

Compositing:

Once I’d finished the render, I brought my image into photoshop to do a few last tweaks to the image. This included some color adjustments, saturation adjustments, and the addition of a slight blur to create depth of focus.

Final Composition:

          

My final thoughts on my project? I’m moderately satisfied, but find significant room for improvement and am disappointed in myself in some ways. 

So where do I feel I succeeded?

I was satisfied with my modeling and texturing of the cutting board, blueberries, chocolate, bowl, and cloth. While I find room to improve on all, I am not bothered by anything and do not feel anything sticks out that detracts from them.

Where do I feel “meh”?

While I was satisfied with the raspberries in general, in retrospect, I think the modeling is a little weak, and looked a little more like blackberries than raspberries. The texturing was ok, but it was lacking the “oomph” I was looking for. As well, I could not get the hair system to work well enough and had to abandon them.

The lighting also felt so-so to me. While it worked, it didn’t quite sell the scene as fully as I’d hoped it would.

Where do I feel I failed?

Zbrush. I had intended to do a lot of work with Zbrush, but failed entirely in this realm. Now, I am ok with this as I was able to make do through other methods, and I intend to learn this better later on.

The spoon was intended to be far more elaborate with Zbrush, but became far more simple due to my limitations. As well, the textures came out looking more like pewter, instead of silver, and I couldn’t get it to look how I wanted. This really bothered me, and I intend to address this in the future.

My render times were huge. At over 10 hours to render a still frame, with only moderate success in the look, I can’t say that’s a good thing. 

In the end, I will consider the project a success, but not the success I was hoping for. I certainly learned a fair bit, and am happy with my first foray into subsurface scattering. I also realized that I again bit off a little more than I could chew, and thus had to scale back my project halfway through. The massive render times prevented me from finessing the final render lighting. I have also gotten direction in where I need to direct my studies on my next project. Primarily, I feel the need to work further on my metal materials, linear lighting workflow, and Zbrush.

What do you think? Would you eat it? I think I would.