Joshua Mollard's ePortfolio

Media Graphics Production

Photo Restoration Project

The goal of this project was to restore an old photo (left). As you can see, the photo was in pretty bad condition and was a challenge to restore. I had never worked in Photoshop before completing this assignment, which meant I was frequently referencing tutorial videos. In order to restore the photo I used the heal tool, clone tool, and patch tool. In addition to restoring the photo I also used a curves adjustment to add contrast and a warm, sepia-like color. I also added a vignette by creating a black layer, masking an ellipse in the center, and feathering the edges.

I was impressed with how quickly the heal tool was able to restore minor creases in the photo but certain areas presented a challenge. The faces of both women, hard edges like those on the lapels of the older women, and areas that had large sections missing were particularly challenging. To fill missing chunks in the background, I used the clone tool to extend parts of the tree, fill in the supports under the roof, and fill in the brick. The rest took lots of patients and time to fix. The most effective tactic I found for filling the missing areas was to first clean up the small creases surrounding the larger crease using the heal tool with a small pixel brush (3-6) and then clear the larger crease with a larger pixel brush. Cleaning the area around the larger creases gave more consistent data to the heal tool so it could better restore the larger creases. However, this would sometimes leave behind other lines or blemishes that were not in the original image which then required a retouching with the smaller pixel brush.

This project was a real challenge for me but seeing how well I was able to restore the photo inspired me to continue working on it. The photo was restored but it looked kind of cold to me so I added a curves adjustment layer to add contrast and color. I used a very shallow “s” shape on the RGB channel for the contrast. I bumped the red channel up some and pulled down the blue. This gave me a red image rather than a warm one so I pushed the green up a small amount to get my end product. I then added the vignette. I finalized the look by experimenting with different layer orders and blending modes but I ended up liking the normal blend mode the most for each of my layers. Below you will find a side by side comparison of the images.

Before

After

Image Manipulation Project

The main goal of this project was to composite two or more images together to create a scene. I had a lot of difficulty designing a scene for this project because I wanted to try to attempt a lot of the projects presented in class but I also wanted to come up with my own design. I eventually settled on recreating the night version of the Spirit of the Forest from Princess Mononoke because I thought it would allow me to use concepts from as many of the tutorials as possible. My final product is not quite what I wanted but I worked hard and think I achieved something decent for my current skill level.

The first challenge I encountered was finding an image of trees to place my spirit in. The image on the top was my original choice but I was unable to cut out the sky for replacement even when I zoomed in to the individual pixels. The result was always too dirty for my taste. I ended up using the image on the bottom.

I cut the bottom image into two layers to separate the trees in front of the dark pine cluster from the trees behind the dark pine cluster. I then removed the sky and the taller trees from the back layer. I did this so I could replace the sky and so I could fake a valley for my spirit to walk through. Below is an image with red lines roughly where I cut the images.

I created the main shape of my spirit by combining the outlines of a man walking for the body and an otter for the head. I filled the shape black and then created a double exposure effect using different star fields to give the spirit a celestial feel. I then added an outer glow effect and used the brush to paint on areas with extra glow and used a gaussian blur to create extra glow. I then painted on the dark spiral markings. To finish the effect I applied a different star field with a clipping mask and used a layer mask to accentuate different areas of the spirit with extra stars. I did a final brightness, contrast, and color correction to finalize my spirit. I then used the brush tool to paint light falling against the dark pines and the trees directly around the spirit and once again added a gaussian blur to add a smoother glow.

To create the sky I combined images of a third star field, clouds, and the moon. I used layer masks and different blending modes to layer them together behind my tree and spirit elements. I finalized the image with a curves correction and the RAW camera filter to add grain and make the image look less computer generated. The final product is below in both its original size and at a reduced scale. I am unsure why the reduced image exported with the pale blue border at the top.

Color Study

The purpose of this project was to create an image divided into four quadrants each with the same image but utilizing a different color from a complementary, triadic, analogic palette. I used a double complementary palette for this project. A friend of mine volunteered to lend me the original image for this project and as a nod of appreciation I used her favorite plant as the background and a forest scene from her homeland for the double exposure effect. I added the mandala pattern in the center because the smoke conglomeration in the center created its own mandala-esque shapes and also drew focus to the center so I thought I should put something of interest in it.

I didn't have a clear goal when I started this project which is why I requested to use an image from a friend. It was a lot of fun to experiment with different orientations in each of the four quadrants. I had a lot of difficulty masking the hair in the original image, a lot of the more translucent smoke was mixed behind it so I couldn't use the quick select or magic wand tool easily separate the background. The light in the original image also created a lot of contrast in the hair itself which prevented my from using different color channels to mask hair which was suggested as a method in several tutorials I watched. I ended up making a rough layer mask for the face and then spent several hours masking as many of the fine hairs as I could. The other difficulty I had with this image was figuring out what exactly I wanted to do with it. I started with the main image with the background and main image in the same arrangement and color scheme as the final product but without the trees and flowers. It seemed flat and boring to me so I added the flowers which helped but was still not quite enough. I added a circular vignette to the BG which helped add depth but the image still felt too plain. I added the double exposure of the forest which added more to the image but the center looked fairly plain and unbalanced compared to how busy the rest of the image was. I added the mandala which originally covered the entire image and played with different blending modes, transparencies, and color fills before finally landing with the image below. Looking at it now it feels too busy. I'm not sure I will ever be completely happy with this image.

Typography Poster

This project presented a unique challenge because I was restricted to using text, masks, and colors to create a poster for a fictional concert. I chose to create a poster for the band that plays in the cantina scene in Star Wars Episode IV: A New Hope. I chose this band because I am a fan of the Star Wars franchise and also because I thought a simple design would work well as a poster for that group. In addition to making my poster aesthetically pleasing, I also wanted it to be accurate to the Star Wars universe. That meant that I started this project by doing some light research on the Star Wars wiki: Wookiepedia. I looked up the name of the band, the name of the cantina, currency conversion, and time conventions so I could include details that would make sense to my audience.

For the design itself, I started with an orange layer partially covered by a yellow layer with a dissolve blend mode to divide my poster into my main space and my informational space. The dissolve blend mode also created a sort of dusty horizon. I then created two overlapping, feathered circles to serve as paths for my text. The circles and background also bring to mind the sandy planet and twin suns of the Planet Tatooine.

I then added the band's name using the circles as text paths. The font I chose for the band name is called Neo Sci-fi v.2.0 which has clean, futuristic shapes and rounded edges that give it a space-funk or space-jazz vibe which is perfect for representing the band. The entire poster is made from only two colors for simplicity and high contrast. I was able to highlight the band leader (Figrin D'an) by placing that text outside my circles in the yellow color and the rest of the band name inside the circles in the orange.

The poster is headed by the name of the cantina and uses the font FREEDOM regular. The subheading is the city the cantina is in and it uses the same font. The right side of the poster utilizes a typeface that was created for and used in the original star wars trilogy called aurebesh. The vertical portion reads "Live Music" and the smaller portion in a block is the same as the information provided in the block in the bottom left of the poster. The bottom left block provides information about the date and time of the concert and uses the font Trench thin apart from the currency symbol which is once again the aurebesh font.

I think this project went well despite the challenge of using only text, color, and masks. The design represents the band and venue well and also matches the simple style of concert posters that may have been used in the 70s when Star Wars first came out.

Magazine Cover

This project is intended to be a culmination of all skills used in the above projects and was therefore one of the most challenging. I had originally planned on doing a mashup scene of different fantasy worlds colliding but I found the canvas space was not suited for that type of image. I'm not a frequent consumer of magazines, so I did some research and found most effective magazine covers featured one, large subject image with undistinguished backgrounds. I redesigned my concept to mimic EMPIRE magazine. My main image was a portrait of myself as a conductor composited over an image of a jedi (myself) and a sith (a helpful friend). I used rounded rectangle shapes in layers with different blend modes to create the lightsabers. My goal was to portray myself as the composer for the soundtrack to a future star wars movie. I filled in the details of the magazine with other future film titles and added a gradient with a grid for depth. I am satisfied with my product but found I was still limited by my amateurish photography and inexperience in visual media. I took the photos on my iPhone 6S+ in front of a green screen but my lighting set up was poor and resulted in some green spill that required adjustment in post. I used the green screen in the hopes that it would make masking areas around the hair easier but once again my lighting was poor and the product ended up slowing down my workflow. Another problem I had with this project was choosing fonts to use. I ended up using typefaces that were used by the franchises they represented which ended up looking messy and was somewhat distracting. In the future I will be more careful with my photography and I will limit my font selection.