For the first part of the semester, I experimented with the demo Josh had come up with the previous semester. I added the ability for more than one user to connect at once.

After that, I began working on inserting the networking scripts I had become familiar with into the Mermaids build. I was unable to come up with an effective implementation, however. The mermaids either spawned in an unexpected location on the map or simply not appear at all. I was unable, despite the apparently well-designed Photon API, to have the character models rotate at all over the network.  After struggling with the code for the rest of the semester, I ultimately was tasked with providing documentation for Photon so that future network programmers can have a head start if they decide to stick with Photon. Ultimately, it seems that my work this semester amounted to being able to advise Josh on avoiding Photon in the future.

The next big task was simple enough but turned out to be quite the nuisance. Zoom in zoom out. I got it to work pretty quickly, but the idea was that once you zoom in enough you would switch to first person view. that cause A lot of errors and turns out it was just a complication with the inputs that caused it to act weird. So after some work with AJ, we got it to work. Not too bad.

The last big task I was assigned was boost movement. Now I have a contrary-to-popular-belief philosophy with boosts in general. To me, unless its in some kind of racing competitive game, boost tends to lower the aesthetic appeal of a game, taking away from the atmosphere, an attribute I find crucial to good games. So how can I make the traversing of the map less monotonous without belittling the impact of the visuals? I call it the link roll. In Ocarina of Time Link could roll periodically giving him a very short (and I mean very short) burst in the direction he’s facing. I did a very similar thing. I made a small quick force push on the player Mermaid that lasted less than a second but gave the player a button to push to speed up. It has cooldown of 5 seconds which is long enough to eliminate the superficiality of the boost. I think its a healthy compromise between my no boost philosophy and the BOOSTBOOSTGOTTAGOFAST philosophy that is pretty common in modern games.

This semester is my first time joining the Emergent Game Group as one of its 2D artist/3D modeller. I dare to say that it’s the best decision I made for this semester. I re-discover my interest in 2D painting. I learned a lot of tips and tricks to make a good 3D model. I meet a lot of new friends. And did I mention that there’s a lot of cool stuff happening at the lab?:)

I spent the first half of the semester modelling some stuff for Ellis Island game. During the latter half, I did some UV mapping and texturing for 2 of the Mermaid models: the Merbaby and the Eel. While working on those models, I learned the hard way on how UV mapping could be something really challenging. I manage to get by with capturing UV map based on camera with my Ellis model. However, I need to get myself used to the other modes of capturing UV maps when I worked on the models for Mermaid, especially with the Eel model. I spent at least 2 hours playing around with the UV capture tools for each of the models, something that I hope I could improve in the future. Aside from that, I also getting a dip into model rigging and some magical tricks to overcome texture stitches, something ugly that must be avoided at all cost. On the texturing side, I spent most of my time experimenting with color blend and shading, especially when working on the Merbaby.

Merbaby front texture

Merbaby back texture -- the power of shading

Eel texture -- because we need more pink in Mermaid

Although I believe my works can’t be compared to the pros out there yet,  I can’t stressed enough how I learned a lot from spending my time working on the assets for EGG’s projects this semester. The semester is ending, but I’ll be sticking around with EGG on fall as well. I’ll be looking forward to see everybody and to learn something new next semester!

The main interface of Mermaids is a magic spell book, so I worked on the appearance of this book. I iterated on the texture and shape of the book for a few times, so as to make it consistent with the underwater effects. The state of the game and detailed information about player, characters, and skills will be kept inside the book. Besides, it is also a social platform where players could share notes or pages with each other. The shells at the edge are bookmarks which provide players with indexed information.

Afterwards, I worked on the shell houses in Mermaids. Based on Jeanie Choi’s awesome drawing, I created three houses with Maya. The second and third one haven’t been textured yet. They might not look delicate as you expected, but due to the low-poly convention in Mermaids, it is a  challenge to create those genetic looks yet using minimum polygons, namely, under 500. 

My first task was the design and implementation of the notification GUI. The notification GUI was created as a way to deliver messages to players during the playtesting session that happened early in the semester. After the play test, we received valuable feedback about the strengths and weaknesses of the notification system. Future iterations of the notification system will improve in the areas that the users found confusing.

 After completing the notification system, I then began work on the Spellbook. In mermaids, the Spellbook is one of the most important UI features of the game. It houses information on spells and items that the player has acquired, and holds maps, clan information and messages. I began design work with Rahul, a fellow {egg} member with a background in UI design. After going through many iterations of wireframe designs, I began implementing the Spellbook in the game. After a basic skeleton of the system was created, art assets provided by Fengbo were integrated into the Spellbook.

 As demo day approached, I switched from working on Mermaid’s UI to working on the spell system. Using the new point and click system designed by Jeanie, I reworked the build spell. The build spell now uses point and click controls, and is integrated with the rest of the converted spells.

Over the summer I will be focusing my efforts on Ellis’s quest system, but I look forward to returning to work on Mermaids in the fall!

With this goal in mind, through the semester, we were able to iterate through 3 versions of the interface. With each iteration we  tried to settle for the best possible middle ground between usability and visual appeal. The work we put in to the new interface truly was a collaborative effort. I was amazed at the skill and commitment that members of the Mermaids group. I worked in tandem with Avery whilst building the wireframes and he used his Unity skills to build working prototypes from the wireframes. Later in the semester when we were working on ways to make the interface  seem more like an extension of the world it was set in, Fengbo came up with some great ideas of mapping items that are commonly found under the sea to the components of the Book. The Book being the main UI element through which a player can navigate the game’s interface.

Through my involvement with Mermaids I got to meet a lot of really talented people and got to view my job as a UI designer in a new light. I also got to learn first hand the reasons why the process of building game UI is completely different from traditional interface design.  I feel that this last semester was only the start of my journey with the EGG. I am really thankful to Celia and the rest of the group for making me feel right at home on the Mermaids project and hope to contribute to and learn more from working with the group in the future.

As the Lead Designer, I headed up our weekly team meetings, where each member of the team would show their progress on their assigned task. I would then assign each member a task suited to their capabilities. We made a lot of progress on the game this semester, starting with a very productive playtest early in the semester that provided valuable feedback that guided our development throughout the semester. Thanks to the feedback from that test, we decided to revamp our spell-casting system to use a new point-and-click system that felt much better, and we also provided new features to increase the smoothness by which the player interacts with the mermaid (including a zoom feature and a boost feature).

Our efforts culminated in a Demo Day where we showed the final product to several groups touring our lab. The changes we made this semester greatly increased the playability of the demo, and the people we showed the game to seemed genuinely impressed.

On the Capstone team, I took a more technical role as well as my role as Project Manager. The system we implemented was primarily a backend system, such that a lot of software engineering was involved in the design. Additionally, we were sure to implement a front-end for the system such that the effects could be immediately seen. Through the use of GUI element in Unity and a large variety of systems, we were able to show our project at Demo Day and it was a great success.

I’m looking forward to continuing my work with the Mermaids team into future semesters!

Spire Concepts

Kelp Spire Concept 3D Render

After reconstructing these early spire, I created a spire that more accurately encapsulated what the mermaids would build. With that, I created the first large asset for this semester.

I learned many thing just going through this first model that would be for mermaids. I worked on properly mapping simple geometries  but more importantly I achieved a lot of insight on workflow of modeling, UV mapping, texturing, rigging, painting skin weights  and animating. Getting to through all those phases withing one model was really satisfying. Another thing I had to take care of for the model was making sure that looping animation I made using Inverse Kinematics Spline Handling tool was seamless.

Mast Spire

After that I my task began focusing on bringing other spires to life. The first building I reconstructed from concept art was Rachael’s rendition of some mermaid homes. One was a a styled home that had a nice curvy shape to it. When making that I made sure to get a good balance between a nice poly count, and maintaining the natural curves of the building. The second building that I created was another one by Rachael. This seemed a lot bigger, so I had a chance to get organizing conventions going for when many assets would be in one huge model, specifically the many windows were around all sides of the building.

Second House with Concept Art Beside

A side task that I got to learn from was a tutorial video that Erica got some good tips from. It taught about creating auto-swim fish rigs using expressions in Maya. With this I took my first few steps into learning about expressions and learned to make a fish rig with nice sinusoidal motions and options to modify and control it.

http://lesterbanks.com/2011/08/creating-an-auto-awim-fish-rig-using-expressions-in-maya/

The last major task I had was Texturing Richard’s shield turtle that he created.  From mapping and texturing the shield turtle, I got better at mapping organic objects. I was very happy to take on this job since I definitely want to develop skills in anything that can be useful in character modeling.

Shield turtle with texture attached

Future Work:

I definitely want to get a lot more into animation and figure out techniques that are used to help biological beings animate smoothly, which would involve expressions, constraints, skin weights, rigging, but also I want to get into proper edge placement that fits the contours on animals, especially humanoids. Retopology would also be useful to learn too. The tasks that I think will be most relevant will be diving into the auto swim fish rig, and figuring out ways to apply that knowledge to other creatures that will be made for “Mermaids”.

The first task was to build the Shield Turtle for the Mermaids environment.  My duties for the Shield Turtle was to build from Ashu’s concept art, rig, animate, and export to Unity.  The textures for the turtle were provided by Jay Belmon.

My first task for the Mermaids Design team was to gather all past lore documents and build a library on the EGG Drive for future years to pull from.  The files can be found under Mermaids/Mermaids Design/Lore in the Egg Drive.  This assignment took roughly one month to accomplish after going through all previous mermaids sites and downloading the various documents.

The second task for the Mermaids Design team was to design the Glyph Naming System for the player.  To understand how the family system worked, I spent the bulk of the semester sitting in on weekly meetings between Celia and AJ’s capstone team that were designing the family/tribe/clan system for Mermaids.  The goal of the naming system is to provide the player with interchangeable glyphs based on their family, clan, and tribe.  Each part would contain a predetermined set of glyphs for the player to choose from, as well as modify color schemes so that no two players could have the exact same glyph name.  To design the glyphs, I collaborated with Keith Johnson in the EGG Art Team.  The basic mechanics for the system are completed and all that needs to be finished are the glyph designs, which will be completed over the summer.

I felt that the original implementation of spells such as the bubble and build spell worked fine for testing purposes, but not in the manner originally intended. Typically when we think of casting a spell we think of some wand being waved around (which I think was captured well with the spellcasting UI system), and then pointed or tapped on the object to cast the spell on. To better reflect this I implemented a cursor system that changes the cursor to reflect when the wand is loaded, and when it can be cast on an object. To further clarify what objects could be cast upon, I also made the objects ‘light up’ upon hovering. There is also the option to cancel a spell now as well, by clicking on the target.

I also altered the bubble spell quite a bit, to make it fit the new system better. Originally the spell would generate a bubble in front of the player, which the player could then bump into objects to hold. The bubble spell’s primary purpose is that of a levitation spell, though, which typically in fiction would cause an object to float, usually towards the player. The bubble now grows around the cast object, and then floats to the player. One of the planned features that would have used the original cast-then-collide method involved multiple merfolk gathering together to create a conglomerate huge bubble that could lift an object they normally wouldn’t have been able to lift alone. I am thinking that with the new bubble system a set size limit could also be imposed on how large the bubble could grow: if an object the player wants to lift is too heavy, it could merely hang suspended in the air in a tiny, frail bubble for a period of time before the bubble pops. If other players cast their spells on the object within that period of time, though, then the bubble would grow/fortify, and the object could finally float towards the median of the casters’ locations.

Future Plans

My modifications to the spell system still need a lot of fine-tuning. For example, my original design did not account for spell-dependent spells (like the build spell), which means implementing more complex ‘stacked’ spells requires a lot of convoluted maneuvering. Props to Avery for successfully maneuvering the build spell into the system! There are also a few scenarios my design did not cover. I mainly covered two types: the common ‘click to affect the target’ spell and the ‘find and lead me to target’ spell. There are still others, such as spells cast on oneself (cures, boosts) and generative spells (making things out of nothing). I hope to tweak my current implementation to be flexible enough to easily accommodate these numerous types. There is also the matter of how objects know what spells can be cast on them. I am thinking that in the future, rather than drop numerous offshoots of the ‘spellObject’ script on each item, a single script listing the kinds of spells that can be cast on it could be placed on it instead. Items could then generate and access the appropriate spellObject script if a spell had been cast.

There are also a few tweaks I’d like to make as well. During testing I found there were times when I wanted to cancel my spell, but couldn’t without casting/canceling on an object. Some universal gesture like shaking the spell off the cursor or something would be nice. I also found that I’d forget what spell I’d cast; at the moment on mouseover the cursor changes to a generic ‘you can cast magics on this object!’ icon, but I hope to alter this to an icon based on the spell itself.

I also hope that over the summer I can look over tutorials for nifty CG effects, so that I can better implement underwater ambient effects, such as caustics and murk. I also plan on doing more concept art next time, as well. Yay for aesthetics! I look forward to Fall.