Digital Art

Week 1

Rendering a Sphere in Photoshop

Today I utilised the drawing software Photoshop to render a sphere by studying values and components to rendering any basic shape, this will aid me in understanding how to shade assets and implement realism with the components interacting with light and perspective to achieve plausibility.

I used a drawing from the Youtube channel "Fine Art Academy" to demonstrate how fully rendered shapes are composed of multiple values that add tone to the object and apply plausibility - these components will be coded by letter, labelled on the photo and have their purpose explained.

Recognising Values:

A: Highlight - This is the area that reflects light the most as the plane that faces it directly.

B: Midtone - The general base value that makes up the shape and meets the rest of the values in the middle.

C: Terminator - Where the direct light cuts off and is terminated, the darkest shadow following the contour of the shape, showing a strong depiction of depth.

D: Core shadow - This is the base shadow that lies just above the darkest value and makes up the side of the shape not exposed to direct light.

E: Cast/Occlusion shadow - The darkest area of the composition, found where the object makes contact with the ground and blocks light.

F - Reflection/bounce light - Where light is reflected from surfaces and illuminates the shape, showing the interaction with its surroundings.

Loish Method:

Loish is a digital artist who gained notoriety in the 2010s for her distinct visual style consisting of vivid colours and fluid lines; she conjured up this efficient method of rendering a sphere which I will demonstrate here.

To begin, I opened a new file on Photoshop and changed the background colour to a light grey to better determine values while working as it acts as a mediator in tones. I pressed U for the Ellipse tool (also found on the leftmost toolbar) and held shift as I dragged it out, which locks the proportions to be equal as it scales up/down. This leaves me with a proportional base that I filled with a medium grey tone using the Bucket Fill tool (press G twice to prevent Gradient).

Next, I blocked in the main values such as the highlight, midtone, terminator, and core shadow. I kept in mind the contrast of these tones by seeing the grey palette on a scale from 1 - 10 with the lowest being light and the highest being dark. The highlight would be sat at a 2, while the core shadow is at an 8, and the midtone meets in between at 5.

For these shades, I kept them on a separate layer above the base shape with a clipping mask applied so my marks wouldn't spill out of the necessary areas.

I then went to the "Filter" tab atop the interface, to "Blur" and chose "Gaussian Blur" which reduces noise and leaves a smooth, blurred look at high values.

Finally, I implemented the shadow by duplicating the base layer with Ctrl C + Ctrl V. I pressed Ctrl T to transform the shape, vertically squishing it and stretching it to replicate an accurate shadow. I refined details by using another clipping mask and applying a Gradient (G tool, Foreground-Transparent setting for depth build up) with the darkest tone in the composition. To finish off, I carefully erased some edges to implement the fading effect and used the "Gaussian Blur" method again to show the shadow's slight softness.

These are my layers found in the bottom right layer tab of Photoshop's interface; I have each one labelled relevantly to its purpose. I'm demonstrating "Layer Discipline" which is a vital habit to build in digital art/producing 3D environments as organisation promotes efficient navigation and prevents messing up from struggling to find which layer you've accidentally edited.

Dev Log Evaluation:

In the areas I participated in, I believe I succeeded as I've gained knowledge in rendering with values which will apply effectively to observational studies in the traditional art module and generally drawing scenes with improved object implementation.

I carried out the task successfully and achieved an effective piece of art, efficiently made with moderate plausibility.

My critique of this piece is that the midtone isn't clear enough and the highlight is over-established, leading to the piece losing depth - to fix this, I could create a new layer and adjust the values with the same Gaussian Blur method, adding more midtone and concentrating the highlight to a smaller area to enhance its definition.

Finally, I believe I could've learned more through doing the square and cylinder demonstrations but I ended up feeling sick and leaving early which is out of my control, however I made sure to document this blog with my findings to prove the lesson's worth.

Week 2

Wood Study

This is the mood board I created to prepare myself for my wood material study; this is done through visualising the plan in my head by compiling relevant images. I gathered photos considering the texture of the bark and cross section, stumps in the wild with different lighting, and closeups of the material. I also analysed a specific image I'd use as a primary reference once I narrowed down my idea of where to take the material; this was an oak stump in warm but diffused lighting, which I annotated to understand what components to include in my drawing. With this, I was able to begin drawing my study in Clip Studio Paint.

Note: I will use Photoshop for most of this, I just had to use CSP initially since I was in my hometown with no access to PS.

This carousel shows the process of creating a base silhouette. With this step finished, I could move onto rendering.

Above the base layer but under the sketch layer, I created a new one dedicated to the values; I applied a clipping mask too so I can exclusively work on the silhouette. This screenshot also demonstrates my layer discipline as I labelled each layer relevantly to its purpose.

I began laying down the base values, considering where the light source would be located which would affect the area of the highlights, the core and cast shadow, and how the branch interacts with the tree to cast its own shadow, strengthening the plausability. I used a round mixing brush at 100% hardness for solid lines at 75% opacity to layer the values with each stroke. I repeated this until I got this result with smoother lines and a general establishment of value.

Adding Detail to the Top:

Before and After Feedback:

My tutor assessed my work and determined that I could slim the midsection of the log to elevate its silhouette and improve its plausibility through correct perspective; he adjusted the values mainly on the largest root to harbour more shadow to imply its curve and volume. With this help, I began to improve my study (feedback was done on a separate layer) now working on Photoshop.

Fixing the Top:

To begin my independent improvements on my study following feedback, I focused on the top. At first, I recognised that the outline's curve needed to be adjusted to correct the perspective, it was important to align it with the base's curve to show that their perspectives match and are of the same object - an important lesson in ellipses. Once this was changed, I added some cracks coming from the root separations to show how their structure affects the cross section of the tree.

Following this, I fixed the values referencing my feedback so I could start adding detail.

For the next adjustment, I used a noise airbrush to add highlights lining the cracks in the wood to portray the cut off edges facing the light source. I also used this airbrush at a larger size and lower opacity to add streaks of dark and light values. This way, I was able to build up the grainy texture in the wood whilst following the curves and bends of the roots to achieve the streaky look bark often has.

To finish off the rendering before applying gradient maps, I went into Image --> Adjustments --> Brightness/Contrast and slightly increased the contrast to define the dark areas and bring out the values. After this, I lined the edges of each root with the default brush and carefully blended out the side facing the root with an impressionist blend brush for a splotchy gradient so that I could clean up the edges and give them more definition while ensuring the new lines weren't out of place.

To begin, I duplicated all of my layers and merged the new ones into a single one so I can experiment and preserve my previous work. For my first gradient map attempt, I believe I achieved a realistic wood colour scheme but it wasn't visually appealing. This gradient was simple, with a monotone scheme going from dark brown to cream which worked for a realistic stance but I wanted something dynamic with a variety of hues, too. Next time I'd refer to my reference basked in soft sunlight for a better variety of colour.

This was my next attempt applying a gradient map, which I feel was vastly improved. This time, I used dark and saturated chartreuse for the deepest value, a standard brown for the midtone, vivid peach for the highlights and then a slightly muted cream for the top section; using chartreuse for the shadowed areas allowed the hint of green to elevate the shadows with its cool tint and the inclusion of peach as a shade of orange providing warmth represents the sunlight as the light source effectively. I'm much prouder of this attempt which popped much more.

For the last part of the colouring stage, I wanted to add some variation in the overall hues as finishing touches that would take away from the monotone look I desperately want to avoid for its dull and linear nature. I applied a clipping mask on a new layer and set the blend mode to "Overlay" which would include the hue of a brush stroke exclusively, allowing you to paint over your work with new colours but maintain its appearance. I used the "texture grassy" brush from the experimental brush pack for its grainy, noise abundant texture that integrated with the current rendering seamlessly. With the brush selected, I made strokes of lightly dark and muted greens which would elevate the plausibility of this piece with the colour palette now reminiscent of woody environments filled with brown and green, and almost imitate moss with its subtle build-up on the wood. I'm satisfied with how this turns out and how the gradient map increased how convincing this study was, meaning I could finalise with cast shadows.

Final wood render. Yay!

Dev Log Evaluation:

With my first material finished, I feel very satisfied and ready to progress onto the next ones. I believe my use of colours in the gradient map and the addition of hues in the overlay layer were very effective and added to the plausibility of the study.

Where I struggled most was placing the values and ensuring the contrast was high and the tones readable; Once I moved my study to Photoshop, rendering values was much easier as I had texture brushes and blenders to aid me but I will still have to work hard to ensure the values are strong on my following pieces.

Overall, I'm pretty happy with this study and will do my best with the next materials.

Week 3

Leather Study

My leather mood board featuring considerations of variation in texture and how that affects lighting, patterns such as seams and embossing and finally an analysis of a leather ball with tight seams which align with the study I'll be doing with diamond embossing as a sphere primitive.

Beginning the process in Photoshop, I etched lines over a layer atop the base (created with the ellipse tool) to mark the diamond embossing pattern, mindful of how perspective would warp the edges and how the curve of the sphere would bend these lines to its contour.

Following this, I erased the edges of the sphere based on the lines to define the areas where the seams would cinch the surface to establish the silhouette.

Starting the rendering phase, I created a new layer above the silhouette with a clipping mask applied to work exclusively on the base; I made sure to place a core shadow atop the midtone along with highlights and some slight bounce light too, all blended out curved like the sphere.

Diamond Pattern Shading:

Creating the Seam:

Creating the seam lead to some of the guidelines mismatching, therefore causing the silhouette to mismatch too. Therefore, I refined that to fit the new seams and cleaned up the values as I matched those layers to the new shape; I used a scratch blend brush for a softer, but still rough gradient to set up for the texturing to come later.

Using a leather brushpack from online, I achieved the jagged hatching present in skin when texturing. I began with a layer set to the "difference" blend mode, using this brush at a darker tone to each square, followed by a normal layer with lighter tones stacked on top. This was an effective and efficient way to imply texture, proving Photoshop's technical prowess.

With all the layers sorted into a process folder and duplicated into a finalised folder, I merged the duplicates into a flat image to apply the gradient map in one step; I made a custom preset with primarily warm, brown tones with red undertones to symbolise the origin of flesh, complemented with a cool grey tone followed by a subtle cream to fill out the highlights cohesively.

Creating a Cast Shadow:

Final Result:

Making some final adjustments such as slightly decreasing the saturation of the material and brightening the seams to set the value of that and the leather ball apart, I was finally done with the leather. Yay!

Dev Log Evaluation:

As I finished my leather, I didn't feel as satisfied as I did finishing the wood. I was proud of my grasp on perspective when it came to executing the seams and making the ball appear like leather, cinched by thread. I also think I'm improving in understanding how to structure my process, beginning with marking guidelines, followed by base values, then rendering detail and texture, finally adding a gradient map at the end. Adjusting to this process will help me complete the rest with ease.

I had a hard time attaining the look of leather and general plausibility, I believe this is an issue with my blending of the shadows and gradient maps. I believe if I blended the shadows subtly with less of a smooth transition, the density of leather would be readable in this piece, along with accentuating the highlights to elevate the hatching of the leather skin. Finally, I believe adding more hues to this piece would've benefitted it, the gradient map was too monotone and having some subtle blue added in an overlay layer could add variety and intrigue to the piece.

I will take these critiques going forward, my main goals being to add variety to the hues; with this, I will continue to work hard.

Week 4

Glass Study

For my glass study, I considered multiple textures of glass ranging from clear to opaque and analysing their effects on reflections to understand how I'd render details. I also looked at shot glasses, cups and crystal balls for variation in primitive shapes (cylinders, spheres). I decided to make my study a glass ball, as I believe the complex distortion in a shot glass' reflection could cause me too focus too hard on detail, forgetting the process of rendering in different layers. A glass ball still is challenging regarding the material itself, but the circular shape itself will be easier for me.

On a new layer, I would apply the bounce light with a hard brush at a lighter value and blend it out smoothly. I changed the background colour to match the mid-tone of this material to establish the effect of transparency in the ball - this was in order to learn how to create this effect without simply erasing areas as I wouldn't be improving my ability to pick values. So I could clearly distinguish the study from the background, I applied a subtle gradient around the study underneath its layers for clarity.

On a new multiply layer, I took a medium grey airbrush at the lowest hardness setting and slightly lowered opacity and lined the edges of the ball to establish depth.

Prominent Highlights and Shade:

Building up Shadows:

This is how the study looked with some extra dark streaks, the depth was finally becoming plausible.

For a finishing touch before applying the gradient map, I used a large airbrush to create a dark circle stemming from the centre on a new layer. I then lowered its opacity and softly erased the middle, for some extra shadow filling the shape, excluding the centre and edges.

For the gradient map, I ensured the use of a cool tone for the darker values and transitioned it into a muted cream for the lighter areas, finishing off with white as the highlight due to glass' generally light mid-tone and reflective nature.

With the cast shadow applied and some extra highlight for finishing touches, my glass material was complete. Yay!

Week 5

Metal Study:

My metal mood board consists of flat texture images showing welded and brushed metal, primitive shapes with different scales of roughness to depict reflections and finally a complex shape to understand the warping of reflections on glossy metal; I decided to do a brushed metal cube as I hadn't done a cube yet and brushed metal will have a matte tinge to it, differing from the stainless metal with clean reflections that resemble glass too heavily - I want variety.

To draw a cube, it's important that the perspective is correct and distortion is visible for a believable shape; rendering will never look truly good if the base's perspective is skewed. I started by drawing a horizon line and two vanishing points, followed by drawing a line underneath so 3 faces will be exposed. With this line added, I connected orthogonal lines from each corner to the vanishing points to establish the sides. The bottom faces were implemented once I drew two more vertical lines parallel to the original; I finished this off by connecting orthogonal lines from the furthest corners - drawing the right line's corner to the left VP and vice versa gives me the top face and a cube in two point perspective.

With the guidelines marked, I used the polygonal lasso tool to neatly select each faces and fill them in one by one with the 1-2-3 read method for distinct values on each face, showing their distance from the light source. I planned to work on each face one at a time for the initial values and then work on all at once when adding intricate details.

Here, I began the rendering with the base values using a mix of an airbrush at varying hardness levels and a noise blend brush. I was mindful of how the majority of shadow on flat, sharp faces like squares collect on the edges and built depth up around there.

To develop the texture, I used a graphite brush for the grainy, scratched effect of dull metal to blend out the values I previously laid down; this worked effectively, bringing plausibility to the piece and resembling rough metal well. However, I intended to do brushed metal with directional scratches, so I would continue adding to the piece until that goal is achieved too.

On a new layer, I used a noise airbrush at a very small size and high hardness, creating light strokes to imitate scratches. I did one layer with a dark grey and one with a light grey for variation and to ensure there was an abundance of lines.

I then went over this layer with a soft eraser to soften the lines and make them feel intertwined with the rest of the piece; to finish the rendering details, I added lines along the edges to show where light collects on sharp cut offs.

With the rendering finished, I moved onto gradient maps. Observing my references, I noticed how they were largely muted and primarily grey with brown/cream undertones, blue found in the areas away from the light source as shadows are generally cool. I applied a monotone gradient map with a muted brown to cream transition, supported with hue variation in a new overlay layer on top. I added blue hues to the face away from the light source, and chartreuse on the lightest face for cool areas to support the cool base colour and the warm light and reflections.

Adding a Cast Shadow:

This is the end result of my metal study. Yay!

Dev Log Evaluation:

I felt hesitant going into this study, considering how the brushed metal material could be considered simplistic and I feared I wasn't pushing myself, but I did find challenge in sorting the process into distinct steps; this has been a reoccurring flaw of mine where I focus too hard on details, copying references too close to actually think about the process and learn from it, simply drawing what I see. I believe this flaw comes from my greater experience rendering traditionally as that literally is just one layer, so I need to utilise layers better and be mindful of the steps of rendering (base values, core shadows, highlights, detail).

I am actually very proud of this piece and might be my favourite so far, I believe I captured the slightly matte texture of brushed metal effectively and the reflections warping into the edges worked well. My use of the overlay layer for hue variation worked well too as the gradient map on its own is relatively flat and uninteresting.

I believe I could've improved on the colours slightly however, perhaps by adding a blue tint over the whole piece to resemble the cool colours often associated with metal but this muted brown I chose does resemble rust and copper. I also believe the scratches could've received some extra attention, adding lighter scratches around highlighted areas to match their reflectivity to the base surface. Other than that, I'm very happy with this study.

Oil Painting Analysis Task:

In this week's session, we worked in groups of three to analyse an oil painting and considered how elements of the composition makes the scene effective. Components such as colour distribution and transitions, utilisation of cool/warm hues, rule of thirds and the golden ratio all contribute to allure in artwork.

We analysed the 1860s piece "Oranges, Nuts and Figs" by Lilly Martin Spencer, each contributing points listed below.

The orange creates a focal point using core shadows - the audience's eyes are instantly drawn to it before observing the rest of the piece.
Different values suggest whether a fruit is fresh or dried, showing the artist's understanding and skill in communicating texture.
The artist is very skilled at shifting hues; the highlights and general light source is warmer, however there are cool tones lining the edges of the piece, especially in the shadows, communicating temperature change too.
The falling raisin conveys a story as it casts a shadow onto the tablecloth extending to the vase, enforcing leading lines within the piece that guide your eyes throughout it.
In certain objects such as the bowl or the orange, the light bounces off of it showing the object's opaque and reflective surface. In examples like the raisins, due to its element of transparency, light passes through the objects and internally heats it, bringing warm light to it; this is a demonstration of "subsurface scattering".
A mixture of man-made and organic materials is present; this is evident through observing textures, like how the fruits are rough, dotted with imperfections and the manmade bowl is smoothly crafted.

Dev Log Evaluation:

This task was a good way to improve my understanding of composition, but I was mostly grateful for the observation of texture and techniques such as subsurface scattering, which I can apply to my renders from now on, especially in natural pieces which will occur often as we carry out still life studies and environmental observations.

Week 6

Pearl Study

This is my pearl mood board, featuring considerations of texture with smoothness and reflections, and colour iterations. I decided to draw a peach toned pearl because, while I desired drawing a unique one with chrome or deep tones, I'd have to change the values a lot and felt like drawing a traditional pearl will help me understand the fundamentals of their structure before tackling these ones.

I started with a circle drawn with the ellipse tool, holding shift to ensure it was proportionally equal height and width-wise; pasting a light colour onto it, I had the foundation for this study, as my peach toned reference has a light mid-tone. With this, I used a large graphite brush to put down the cast shadow on a clipping mask above. Since the ground's bounce light and the highlight would negate the majority of the cast shadow, I kept it exclusive to the top half to save time rendering.

Implementing the Core Shadow:

Added some bounce light with a value lighter than the mid-tone, darker than the highlights will be to bring emphasis to their light.

Placing the highlight.

Brightening it with the study's lightest value.

Blended out all the layer outlines slightly for a smoother, glossier look.

The result. This was the final stage of the rendering, meaning I could apply a gradient map now.

Observing my references, I achieved the base colour of the pearl with my monochrome gradient map. I used a warm brown to muted cream transition to resemble the desaturated traditional peach pearls.

With the gradient map applied, I added a clipping mask layer with the "Overlay" blend mode to shift the hues of certain areas. I added chartreuse and muted green tones to the centre, pale indigo to the core shadow and extra coral warmth on the edges to accentuate the bounce light.

Adding a Cast Shadow:

Final Render:

Dev Log Evaluation:

I'm feeling happy with this piece and much more confident about the brief than I was starting; my strong points this time were my efficiency in the rendering process as this whole material took just under an hour to do, much quicker than when I started my studies. I've greatly improved in breaking down the rendering process into steps of the fundamentals (core shadow, highlight, etc) and my gradient maps and colour application feel dynamic and improved too. While this study was rather simplistic, I'm still happy with it. However, I plan to switch things up and challenge myself with something detailed and organic for the final study to test my new skill.

Week 7

Banana Study

This is my mood board for the banana study; I compiled images mostly of the cross section at different angles and lighting and analysed the difference in reflections based on the difference in moisture between the sides and cross section. I planned to draw a peeled slice of banana as this aligns with the goal of using primitives, this one a cylinder.

To create the base, I drew an oval and duplicated it, connecting them with straight lines to form a cylinder; I duplicated the ovals to ensure my ellipses were correct as the base and top ones must be proportional to each other for the perspective to be correct. Finally, I filled these in with two values to separate the top face and the curving side.

I repeated a process of rendering to achieve the matte, clumpy texture of banana skin across the planes of the side. I'll walk through the steps of this process.

To begin, I used a noise airbrush to lay down the base values of the skin; this brush in particular was to achieve the matte surface through the grainy texture of noise.

Next, I used a "grassy texture" brush with curved strokes following the contour of the skin, implying direction and flow within the shape. I built these strokes up with dark and light values for variation.

I went over with a dark value to implement creases in the skin.

I then decreased the brush and increased the darkness to add depth to the creases, followed by going over this with a light value instead to create slight highlights; the highlights were only subtle since the skin is so dull.

With the skin complete, I could move onto the cross section; I began by drawing some guidelines for the seeds and pattern in the core, lowering the opacity of this new layer to see my brush strokes clearly.

This was how it looked after implementing the darker values; I followed the guidelines and blended out certain areas to achieve the soft, glossy texture of this face. I built up the depth of the darker areas like the seeds by using small brush strokes to stack the value. I used the grassy texture brush to blend the value so it still has a hint of grain and isn't 100% smooth.

With the same small grassy texture brush at the lightest value of the piece, I added highlights lining the curves of the dark areas to imply their depth.

For the final step of the rendering stage preceding gradient maps, I went to Image --> Adjustments --> Brightness/Contrast and increased the contrast so that the values were readable and the faces were distinct.

For the gradient map, I applied a cream tone with a muted gold for the darkest tone.

In a new clipping mask layer with the overlay blend mode applied, I added some extra warmth and depth to the skin and some extra highlights to the top face so the values between that and the skin are distinct and the glossiness is evident.

Finally, I added a cast shadow by duplicating the base, filling it in with a deep, warm orange colour. I erased the edges with a soft eraser to show how it fades with distance, and applied some gaussian blur to show its general softness.

Dev Log Evaluation:

With the banana complete, that marked the end of my first project. Tackling these materials was a daunting task, especially since I didn't delve into realism digitally often. A struggle I find with digital art is choosing what colours to use; traditionally, you have limited choice of tones and build them up to achieve an effective render but digitally, every colour on the wheel is handed to you. The overwhelming choice stumps me, which is why I'm glad we got to use gradient maps that I'm experienced in.

I find myself impressed by the result of these material studies and believe I pushed myself hard to produce good results and meet my own expectations. I believe I'm good at rendering traditionally because I've done it for years, but I always took a cel-shaded and cartoonish approach to digital art. Attempting realism in a digital format was an interesting challenge; I found it difficult initially when I was producing the base values on Clip Studio Paint at home. Once I translated these to Photoshop and began building texture with the large selection of brushes, my confidence increased and I finally began to see the appeal of Photoshop. Using the textured blend brushes elevated my work as they placed and built values efficiently, along with textures simultaneously. I'm very proud of how I captured certain textures, especially the brushed metal and wood. I also believe my confidence has improved in regards to recognising the core components of a rendered object such as the highlight, mid-tone, terminator, core shadow and bounce light; with every material complete, I found it easier to follow the process of individually implementing these and realised I could place them independently without relying on a reference. My independence and executive decision-making skills increased too, as none of these materials are a one-for-one photo recreation.

While I'm proud that I utilised overlay layers for hue variation, I still believe I need to improve my understanding of when to use certain colours; I'm aware that we'll be covering colour theory in the upcoming so I aim to practice and study it then. I also believe some of these materials, such as the glass and pearl, are relatively simplistic when I could've challenged myself further by increasing the shape's complexity slightly (shot glass for example). It's vital that I always push myself and challenge my ability, even if I fail, to keep learning from my mistakes and improve. Regarding failure, I must keep in mind that it's okay to fail; it simply means that I have the opportunity to try again. I wasn't entirely happy with the result of my banana study, but in reflection, I can see that I didn't convey the moisture collection on the top face which resulted in the obscuring of texture distinction. Next time, I will be sure to emphasise highlights on glossy surfaces such as this.