Project Alpha - Milestone 2

Our second milestone required game testing along with concept feedback during one of our sessions. After finishing the paper prototype (for our last milestone) of our 'walking simulator' game, I was able to start work on the first map.

There were various alterations to the players path and what could and couldn't be explored on the island overall. This was mainly due to things such as time constraints and modelling hours working in a small team. Above is the first draft I made, of which we could use in game as a map guide (with a refined update) and to also refer to for later plans and gameplay designing.

For several of the monuments that are scattered across the game, I planned on creating model sheets or 'turnarounds' to help Prem (our Game Designer & Environment/Modeller Artist) with construction. One of the largest monuments from throughout time, that is going into the game, is the Colossus of Rhodes statue. This statue of the Greek titan god Helios, was one of the seven wonders of the ancient world.

The first and foremost issue with having this statue in the game, is there is no reference images of it as it was destroyed during an earthquake in 226 BC. Many artists have depicted different poses over the centuries and the painting above is one of the most plausible postures that could have been made of the time. In the image above, I have tried to match a basic male model in Mudbox that I quickly posed using the joint and weight tools.

After posing the basic human form, I could then go into an orthographic view and grab shots from the front, sides and back to easily create a model sheet for our Environment/Modeller Artist. I thoroughly enjoyed working on this piece as it accumulated to a historical and innovative challenge. Furthermore, one of the central points of using this statue was to generate originality over say a stereotypical monument; like that of the Statue of Liberty (of which was believed to be inspired by the Colossus of Rhodes)!

Here is the final Colossus of Rhodes model sheet. I kept to using fine lines (like that of a blueprint) to aid with distinguishing form and size. I also added the statues base plinth on. 

As we could implement dozens and dozens of unique landmarks of interest into the game, we decided on capping the amount to sixteen or so. Having too little is better than having to many in this sense, as there is also a style to composing and merging cultural spectacles. If the Colossus of Rhodes statue was placed next to a Moai Head Statue from Easter Island, would that look aesthetically pleasing? Or would there be a visual conflict? These are some of the questions we had to consider and was really quite fun to evaluate.

Previous is the model sheet I made for the Arc De Triomphe monument (from Paris) and above is one of the smaller misc relics of time in the game, a Chevrolet Camaro (aka Bumblebee)! Like that of the Colossus of Rhodes but no where near as grand, I struggled to find references in an orthographic perspective for the 1970's car. After some of scouring online, I eventually had to combine earlier model blueprints as well as a modified one from the Transformers (2007) film; minus the racing stripes. 

Here is another large landmark we plan on including in the game, the El Castillo from Chichen Itza (in Mexico).

The distinct and strange features of the Moai monolith statues give a great sense of mystery and that was another factor we wished to include in the game. I created just two models on one sheet so that Prem (our Environment/Modeller Artist) could make the two and then simply multiply and alter rock textures on the models for variants.

One of the first signs of civilisation in the game is a small village. We decided on setting the game a few hundred years on from when the world ended and time went into disarray. The idea of putting that event/backstory in was so that we could show how the remaining humans have resettled and constructed society again; using a mixture of technology but regressed slightly to the medieval era. Here is one of the peasant huts I put together. One of the plans we developed as a team, was to be able to make assets that we could reuse and place again over other structures. This would save time modelling and give a sense of familiarity across the environment.

This is the second peasant hut model sheet. Again using elements we could reuse and that gave off a mixture of time periods. Another aspect I wanted to exhibit was that of architecture built from the environment around it. The features such as the wood could have been carved from the forest nearby and the roof, made of corrugated metal, from a destroyed garage that came out a void.

To further expand this idea of building around monuments and objects from the voids, I made this peasant hut. The house has been built directly next to a huge mine truck wheel. Incorporating in these elements to make a structure that doesn't look too irregular was a great challenge and also pushed my concept problem solving.

This is one of the upper class houses that has again been brought in by a void. I followed designs that our Game Designer was after and included influences from 18th Century European architecture. Again with features that could be reused and interchanged.

While being given basic concepts from our Game Designer, I was able to experiment and develop upon these; with a massive influence from TV and film. The idea around these portals are, 'voids' open up across time and space like a doorway, between two time periods and whatever was in the location (often the largest man-made objects) would be transported and abruptly dropped in to the present. The voids was a wonderful way to experiment fractal design and above is an early model of one. I used a simple material pass and found a chrome setting that I thought would work quite well; like a mirror breaking through dimensions.

This was the first render I did in a basic sun and sky pass. I was quite surprised and satisfied with the result of the mirror effect. However I felt that some particle effects were needed to bring it more to life. 

Here is another shot this time with more shards in the centre.

I thought that this one had better form with just the right amount of chaotic shards spaced out. The next step would be to add some particle effects.

To help with the work load of modelling, which is to be mostly all made by the Game Designer and Environment/Modelling artist alone, I began to model the Golden Gate Bridge. This bridge is a very iconic piece of engineering and the art deco design and international orange colour is quite unique. As the structure will be at the start of the game and feature heavily in the concept/promotional art, I wanted to do a decent job on the build. Above is the beginning document stage of blue prints, measurements and project save managing.

Instead of working in mirror mode the whole time, I decide to create half of one of the towers first; so that I could then mirror and merge later on. In the documenting stage I had to change some of the camera settings so that I could see the model from afar as the object is to scale, at two hundred and twenty seven meters high!

As the player has a third person perspective but is grounded to the floor, I could work from a high detailed base to a very low basic form at the top. This would also help with the games overall operating when it comes to the final game build. 

To create the suspension cables, I used a curve line with a bezier setting. This allowed me to create the curve line and also go back in and edit the angles etc.

Here I have used the extrude face along path function. I find this to be the best time saver when it comes to curved cable like objects. I placed the before and after next to each other to show the great result. In addition to this, I tried to use the least amount of edge loops on the curved line to once again help with the games operating. 

One of the smaller challenges was modelling the finer details. Although I have been fortunate enough to walk on the bridge itself when I was younger and have a good sense of spacial construction for it, some of the higher tower struts are inaccessible but an obvious overall detail. A website called Turbosquid with a tonne of 3D models on and several of that of the bridge, was great help for referencing things like this. Here are the two pages I used with a bundle of great renders:

http://www.turbosquid.com/3d-models/golden-gate-bridge-architecture-3d-model/621273

http://www.turbosquid.com/3d-models/golden-gate-bridge-3d-model/532375

Once the modelling of one side was finished, came the big mirroring job. I double checked the merge with parent option before I applied the mirror!

Some of the separate items had to be mirrored individually, like the strut walkways and fences etc to be aligned more accurately.

As I went along and saved each time I went into Maya to model the bridge, I made sure to keep a clean hierarchy and label things neatly so I that I knew what objects were what.

Something I discussed with a lecturer at Uni (about all the structures in the games world) was decay and deterioration. He suggested the term to think about called 'Entropy' meaning a gradual decline from an ordered state into a disordered state. There were many ways I could apply/display this in the monuments in the environment. Some examples of this are nature reclaiming a statue like ivory growing or the corrosion of metal on a building etc. One of the ways I did this destroyed atmosphere with the Golden Gate Bridge, was by drooping the suspension cables.

I played around with the sun and sky setting in my mental ray renders of the complete model. This was the best way to show off the height and shear size of the bridge tower. On the other hand, another way I could do this was low camera angles like above.

This render shot shows the hanging cables and the first strut on the tower.

I was thoroughly pleased once this render above was complete. The low sunlight and cast shadows worked fantastically.

I decided to see what the model looked like with just a primitive colour and the above render came out quite nicely. Now completed, I gave the mid/high poly model to our modeller who will then make a low version, bake the two and then texture using Quixel; before being ready to go into the game.

Project Alpha - Milestone 1

Our Year Three Final Major Project comprises of two units; Project Alpha in the first term and Project Beta in our second. We were given the choice to either work solo and build up a portfolio of a game art bible or work in a team and make a 'substantially' sized game. Several students had already teamed up over the summer, having chosen to go solo before starting the year, so I had a limited choice of who I could join (as I know working in a team was what I wanted to do from the beginning). 

After discussing ideas with other fellow students, I settled down to work with just two other students (Prem Fatek & Peter Bedding) concentrating on game mechanics/ideas already built up by Prem. We discussed our strengths and weaknesses of roles and also what we wanted in our individual portfolios, to show off to potential employees once we had finished.

I choose the role of Concept Artist with Prem as the Game Designer and Pete as the Sound Designer/Animator. Whilst these are our main roles, we will be filling in/crossing over other roles (as there is only three of us) like UI, Character Art, Environment Art etc. Above is our initial one sheet for our game 'Crimson Kingdom' (working title). Set on earth in a dystopian future, the games genre is somewhere around post-apocalyptic sci-fi fantasy with a third-person view. 

With our FMP (Final Major Project) being another step up towards looking professional and like being in a small game company, I designed a logo for our team name Project Peacock. Above is the final design I quickly made in Adobe Illustrator to get a crisp vector file that I can then resale easily for later use.

Towards getting an idea of what further concepts and designs could look like, I made four thumbnail speed-painting pieces. These were mainly to help visualise what our game may look like for our presentation/first milestone; with the games level of detail to look fairly realistic. Later on, I think this will be a good challenge, potentially making both highly detailed photo bashes of the environment and other things such as paint-overs.

Here I have made a colour palette of our games overall mood/contrasts. I liked the idea of a rich sunlit saturation. As the game is set in a world where time has stopped, this gives us free reign over a broad range of aspects and factors such as day/weather/atmosphere etc.

For our games working title, I wished to use a decent font that expressed the feel of the post-apocalyptic game and that we could then use for the final name. Out of several choices, I narrowed it down to five and then as a group, we all decided on using 'Filth Of Icarus' from the free downloadable font website http://www.dafont.com/

Games Art: 2D Pathway - Environment (M4/4)

One of the last two requirements of the Art Pack in the 2D pathway, was to create a model sheet (turnaround) and a final piece (detailed image).

Above is my final model sheet of my part vehicle/part environment scene. I decided on including a front, side, top down (birds eye view) and a back view. In addition, amongst the other graphical elements, are measurements of the heights, widths and lengths of certain objects; such as the ship and dry dock platform. I really enjoyed aligning components to scale and proportion. Moreover, this would then aid a modeller if it was to be made into 3D.

Here is my final render made in Maya's plugin, mental ray. To get the right composition was a tough choice, as I wanted to get several elements into place and direct the field of view towards the focus that is the colossal anti-gravity ship in centre. I looked through several shipping and dockyard environments on Google Images before I came across one of a dry dock in Portsmouth. As I am from the historical naval city, it seemed odd that I did not think of this setting at first (one that I am very familiar with of in the dockyard).

Above are several of the photographs I took on my travels and in my free time to use for my digital work. For example, a range of metal surfaces that I then cropped and shaped to fit certain aspects of the ships etc. After this, I then played around with the blending options in Photoshop to get the best result.

This is the step by step progress of my final piece. As it was only a simple paint over of my 3D render, there are just four shots but it is clear where I have adjusted certain areas. It was really interesting looking back on this at the different stages and the improvements I applied. I attempted to focus my time management as correctly as possible between certain aspects of the art pack requirements. However, I feel I could have spent less time on my variation compositions (regarding detail) and more time on my final piece.

Lastly, this is my final piece paint-over of my previous rendered image. I made sure to use a unique colour palette, not only because of the interesting tones but to capture the elements and mood of the futuristic state of the depleted and desperate world (in my game idea I made the content around). 

Furthermore, I found several weaknesses in my work as regards to painting proportions and large scaled 'hard surfaced' objects. To challenge and overcome this obstacle, I studied in-depth at how to use guidelines and visualise space (of the immense) through real-life examination. In other words, on my travels, in London for example, I looked at buildings etc from afar and their sizes to other bodies surrounding them (jotting mental notes and small doodles/sketches).

Overall, I believe I have refined my practices in both the digital medium as well as in my practical (sketchbook studies) while exploring my ideas revolving around present nautical and future freight transportation, through in-depth research, to the best of my ability.

Winter Submission: Triple A Asset Production - Re-submission UPDATE

For the previous Winter Submission, there were many mistakes and issues I made. This built up to me not being able to have enough time to place my gun in Quixel as well as the Unreal 4 Engine for the deadline hand-in. Fortunately, the first submission was classed as a 'formative' meaning we could re-submit our work with any updates for a re-mark.

I took this opportunity to add in the work I did not get the chance to do in the previous term. The only problem I had to work out, was my time management around my other deadlines and work.

Above and before are some TURTLE renders I created in Maya. I used a low light and contrast to capture the detail and surface variations. In the image above, I removed one of the arcing petals on the barrel to show the work I did on the inside of the gun (which would then open up like a flower for the 'firing mode', if I were to animate it).

On placing my gun and all the components that I baked in xNormals into Quixel, I realised that there were quite a few problems with the normal maps and ambient occlusions such as on the centre stock part seen above.

Quixel allows the user to preview all of the textures and applied materials in the 3DO viewer through the keyboard shortcut (of pressing the numbered keys 1 to 5). Here I have just compiled all of the modes together to compare my practice test. I found this out with some other tips through reading the Quixel online manual at:

http://quixel.se/usermanual/quixelsuite/doku.php?id=3do_hotkeys

Here are the cables at the rear base of my gun model in selected view in Quixel.

I experimented with the size and other variables in the inspector tab to get the best looking textures possible (in this case, the size of the leather effect on the handle of the gun).

I was really pleased with the test outcome above and how powerful Quixel really is (plug-in with Photoshop). For this one, I added on the 'Optimus Prime' material which gave an interesting scratched metal effect on the petal ends.

From adding in my model and its four parts into Quixel (top handle, centre stock, rear cables and the barrel petal ends) I could judge how much time I needed to fix my mistakes and then progress up to placing it in a scene in Unreal.

Likewise with Quixel, I placed my model into Unreal and experimented with placing the materials onto the gun, to decide and also learn how to use/harness certain aspects of the software's; in time for the re-submission hand-in. 

Here I placed my different maps into their corresponding material nodes. I enjoyed placing these in and looking into the different channels and their basic uses such as for emissive lighting.

I made sure to organise my separate maps into their representative asset folders with that part of the gun (four folders).

Here I have displayed clearly my worst map of the stock part that needed correcting. I believe that I baked, not knowing I had changed the original obj UV map layout, after combining several parts together in Maya. This resulted in UV faces not aligning where they should be.

From working with Maya more closely, overtime, I have learned and understood more elements from independent learning. That being said, I have also had a fair amount of help from some tutorials such as that on the main Quixel website at:

http://quixel.se/hub/tutorial/uv-mapping-for-the-suite/

Understanding the UV layout editor in Maya better, has definitely helped with my work flow and method of producing the best possible results. Above is the wires in the centre stock part of my gun (that I focused on mainly as a thing to fix). It is quite clear to see how the surface of the UV's are wrong, but the unwrapping process to make them right would be something that I had not yet done (a cylindrical unwrap or automatic unwrap would not suffice).

I was able to find an excellent tutorial on YouTube (see link below) on how to unwrap a curved pipe manually. By forming a seam along the length of the tube and then cutting the UV's. I was able to then unfold vertically and to give a decent even layout, I used the unfold UV tool in the UV Editor.

https://www.youtube.com/watch?v=dahgVetfUZg

After around seven hours of resizing all the faces and correcting the scale in the Editor, I was able to complete the new UV's for the centre stock which was one of the main issues.

Here I have a before and after of the difference in the unwrap of the stock on the gun. As I was pushed for time last term, this was one of the reasons the UV's were messy. However, after working on other projects and models in Maya, I have more experience in the way they are laid out then I did before and so was able to work to the best of my ability and successfully overall.

As I had re-done the low poly models UV's in Maya, this meant I had to re-bake my guns normals and ambient occlusion maps for the stock part in xNormals.

Since last using Quixel, many updates have been applied improving performance and optimisation of things like the 3DO viewer. I selected in a white box above, the way Quixel orders my separate four components neatly. This allowed for a simple method of selection when it came to adding on the material texture.

Here is the final material called 'Galvanized Metal' that I used for the stock from the material library. I was really pleased with the clean result and how the UV's turned out with a texture on it as a whole.

One of the final stages for the re-submission was to place my gun model in a scene in Unreal. Here I attached the petal end maps to the right nodes; as like in the previously mentioned.

For some reason, the texture and ambient occlusion maps appeared very dark when I placed them on the model parts and pressed test play. This resulted in the stock on the model in the scene view to look black; rather than a shiny metallic finish. To counter this, I simply bumped up the brightness to 5 on both the albedo and ambient occlusion maps on the adjustments panel in the Editor which surprisingly worked.

The Unreal build required us to set up a sort of diorama scene showcasing the model. I decided to go on a simple sci-fi vibe (following the background of the gun) and set it up like it were being examined with side lights projecting onto it.

I placed a few tall pillars around the outside of the gun (in centre) and made a thin tall rectangle to represent fluorescent tube lamps which I then duplicated and placed on the pillars. Although, I had a try independently at emissive placement on an objects material (with no success); I was able to find a fantastic, quick and easy to understand tutorial, of how to apply a Multiply expression to the emissive colour (which is different to the more familiar Unity engine) to make an emissive glow.

https://www.youtube.com/watch?v=CtyNXjN_HFk

I was quite pleased with the emissive lighting procedure and think it added an interesting cinematic element to the scene. Above is the material layout with a simple glow (before I added a light neon blue colour to the emission).

In addition to the emissive light panels, I added in some spot lights and also changed the time of day by rotating the skybox to create a dramatic effect for the sci-fi feel.

Above is a screenshot in-game of the final scene. A last minute feature I added into the build, was to make the gun look like it was levitating via a magnetic levitation display. I would have liked to have had more time and include things like a script that rotated the gun slowly around 360° and wobbled slightly over time; for a realism touch.

Overall, I am happy with how I have recovered this project from its major complications and issues (and also not knowing my own limitations before). On the other hand, we learn a lot from our mistakes and I definitely found out some these and tried to the best of my ability to rectify them.