FIEA SE Trip

Wednesday April 8, 2009

Technical Director Tom Carbone and I started on a small recruiting trip through North Carolina in search of programmers. Yesterday we visited East Carolina University; today we are in Chapel Hill; tonight we are at Duke and tomorrow at Clemson.

ECU yesterday was a great experience. We touched down in the tiny Pitt-Greenville airport and alas, one piece of my luggage was lost. Bad start.

But we took our rental car and found some food at the Cookout, a burger/hot dog/milk shake kind of place.

We headed through the small town of Greenville and quickly found campus. It is a beautiful southern campus with magnolia trees and dark red brick buildings everywhere. We went to the Career Center and talked with some advisors there about the school and our program. I told the associate director I would send her some information.

We headed to the CS department and met Qin, a professor there who was organizing the talk for 7 that night. She was friendly and introduced us to several of her colleagues and the department chair.

They were warm and welcoming. We ate dinner before the talk at the Japan Inn, a good Japanese restaurant. We also ate with a CS student who had just defended his senior project named Chris. He was tired but happy he had finished.

The talk went well and lasted about 2 hours. We had 35 or so in the audience and both faculty and students asked a lot of questions.

We left ECU about 9 and went back to the airport to find my luggage. Turns out someone else h had taken my luggage. After contacting them, we drove into a nearby neighborhood and picked up the bag. I was very happy to have my clothes for the trip.

We got into the hotel late last night and today we talk at UNC at 2 p.m. and at Duke at 6 p.m. Should be a great day.

Will update again soon, travel permitting.

Todd

Thursday April 9, 2009

Today was a doubleheader of two of NC’s most famous basketball schools, UNC and Duke. After a late night into Chapel Hill, we woke bleary eyed at a Hampton Inn. Tom and I answered email and made phone calls first thing before heading to campus. We came in on a very crowded South Street.

The campus is absolutely beautiful with old, brick buildings and beautiful green lawns and open spaces everywhere. The Weather was so beautiful and the campus was crowded and active. We quickly learned that a major problem on campus is parking. There was nowhere to park. We circled in one public lot and then decided to eat some burritos at a place called Carbabritos. It had a funky, hippy interior and a great menu. We both had fish burritos. Very tasty.

We finally found a parking spot on the Raleigh Street. Tom did an illegal u-turn and pulled off an amazing parallel parking job just to get the spot. Then we had to beg for some change from some unfortunate passersby. Even with all of our efforts, we would barely have enough time on the meter to finish the talk. We walked across the stately campus with all the trees in bloom. Incredible. We finally found the Brooks building after walking through the north quad. Donna met us and took her into her class of 7 or so.

The presentation went well and the kids asked some great questions. Afterwards we stopped at the bookstore and marveled at all of the championship gear. I got some shorts and Tom bought a shirt.

We left and started to head to Duke. I had forgotten an important folder on campus so we had to circle back but then left without finding it.

We quickly drove to Duke. We parked and headed to the building. Tom’s original contact was sick so we were greeted by Susan, who is a professor in the CS department. We finally found the department and headed in. There were only three people in the room with 10 minutes until the talk. Then the room filled up. It was an ACM talk and was nicely run and attended.

The talk went great. Everyone asked questions and they brought in pizza. Lots of student seemed interested in FIEA and the game industry. We both were talking to students for 20 minutes afterwards and we met another faculty member who teaches game classes there. We left Duke and ate at Mcallister’s Deli. Good food.

Now we’re driving 2 hours to Charlotte, where we’re staying the night. Then up in the morning and another two hours until Clemson. Can’t wait to get there.

Saturday April 11, 2009

We drove through pouring rain to Clemson. Clemson is a bit out the way but a beautiful campus. We met our contact Brian at the student union. He took us with the department head Don and Brian’s wife Jackie to the country club for lunch. A nice buffet was served and we talked about all the connections we had in our work. After lunch we were given a tour of Clemson’t Digital Production Arts by Steven, a student who had started at Full Sail. He knew about us and UCF.

The program really does a nice job teaching technical artists and has had some success putting students onto some very well known films. We saw a demo reel and talked to a faculty member about the difficulty of teaching artists and programmers. Our mutual challenges sounded very familiar.

We then got to meet with 4 or 5 students and talked to them about the industry. They asked some great questions. Then at 3:30 the talk started and about 30 people watched it in a crowded seminar room. The students and faculty were very sharp and asked some excellent questions. I think all got a lot out of the talk. Or at least I hope so. Clemson had really treated us nice and I appreciate their hospitality.

We drove to Greenville where I would stay with some friends and Tom would stay with family in the area. A long flight the next day and we were home. It was a great trip.

Production Week 1

Last week we took a break from status reports as the Drifters team had just presented the Thursday prior, and Sultans team had a pseudo presentation on Tuesday with the Engineering department in regards to their peripheral. Therefore, the teams took the week to consider the information from their vertical slice presentation and organize their final plans to transition into production. Therefore, this week begins our first update on the actual full production of the games. From here on out, the content delivered during the status meetings is all geared toward the final product. Expect to see more major updates and amazing work coming out within the next 3 months!

Drifters
Two weeks ago the Drifters team held voice talent auditions here at the FIEA conference rooms. Dozens of voice actors were provided with sample scripts to read in order to find narrators, Drifters, and crowd voices for the final game. After a review of all the candidates, callbacks were made and now the casting is being determined. Scripts for cinematics and the voice clips for in-game use are being written, reviewed, and should be finalized by the end of April to begin recording.

The cinematic team has been hard at work and are at the last revision of their script. Once the script is approved, it'll get a storyboard treatment to allow the art team to begin visually rendering the assets needed for completion. A schedule has been developed to keep production running on track, as the cinematic is a large part of setting the tone of the game and establishing the fiction.

Exterior shots of the museum, yes that first one is an actual in-game render!

The design team has also been working on the crowd behavior, namely the makeup of the population and the exhibit placement. The goal is to create a realistic-looking museum environment that has a good population of NPCs for drifting into, while not seeming overly crowded and impede movement. The preliminary layout of the museum is simple and straightforward, allowing for potential expansion with other wings and rooms if the initial design proves to be too simplistic.

To help the design team with the population of the museum main floor, the programming team has been working on some tools to allow an easier placement of objects within the game. NPC pathing tools, as well as scene design tools allow producers to place assets with little to no scripting necessary. Integration of the Python language is also being researched, as it will open more scripting options to the programming team. Alongside the tools being developed, network play is still in the cards and being worked on. Drifting now works across the network, with one player being able to successfully drift into a NPC and then another player in Drifter Vision able to see the effects of that move. The lobby for joining a networked game is next on the development list and being started currently in this sprint.

As previously mentioned, the NPC diversity issue is still being solved with the segmentation of body parts to allow for a mix-and-match system, creating hundreds of combinations. Currently, the bodies are broken up into 4 parts: hair cap, head, upper body, and legs. This structure can be further broken down into 4 hair caps, and the other parts having 3 types of geometric structures each with 2 unique textures - and that's just for a male NPC. This almost guarantees that with the target number of NPCs set to appear on screen set at about 80 and with a blend of male and female models, there's a near impossible chance that the same 4-piece combination will appear on screen twice. A tri-rig system is on the programmers' list of features to begin working on to allow for the randomization of the pieces and combination into a fully functional model.

"Maria" "Agnes" and "Mitch" - some of the new models

With the fictional switch from a normal museum day to the opening night reception of the museum, the lighting has had to undergo some changes to a more dimly lit, dramatic scheme. The rays of light that used to shine brightly from the ceiling are now toned down and tinted slightly silver-blue to simulate moonlight streaming in from above. Orange-yellow sconces on the wall show off a warm glow, while the rest of the interior has an intimate, darkened tone - helping establish that dark feeling of paranoia within the crowd.

The museum interior with some of the basic lighting adjusted for night time

Sultans of Scratch
The Sultans team is showing great progress with new characters, designs, and the arrival of their first peripheral prototype looming on the horizon. They've established an April 26th milestone for a lot of the major gameplay elements to come together in a solid, playable version of the game.

A meeting with their DJ on April 6th allowed the design team to lay out the scratch sounds for the music tracks, meaning note additions can start taking place and create a rhythmic gameplay rather than just random testing notes. The UCF engineering team that the SoS developers have partnered with has announced the first iteration of the custom DJ table controller peripheral will be available to the team on April 11th. Once the team has their hands on it, they'll be able to assign the drivers and have the table controls respond in kind to the proper input commands within the game, giving a much more finished feel to the gameplay.

Artistically the team has been giving their concept artist plenty to do with a female character needing development. After creating concept sketches for over 20 different female styles, the female FIEAns were asked to come by and pick their top three. A consensus was met and the new female lead dancer has been fully sketched and ready to render in Maya.

The unnamed as of yet female dancer

More artistic progress includes a new iteration on the particle effects where they resemble more of an ink swish or streak rather than a spray cloud. Many new turntable concepts have been developed as well to give the user more options on what style of table they'd like to scratch on.

Different trail styles

Various turntable designs

The in-game assets have been cleaned up with a simple turntable layout and note stream areas overlaying the stage. The shader is in place and looking awesome on Mr. Yin-Yang who can perform several seamless dance moves chained together, as the animations have been progressing well through the cleanup stages. Another notable addition was the change to the menu screen, which hits their target of urban and stylized very well.

New start menu

The End of Pre-Production!

"Vertical Slice" is a term we've been kicking around for quite a few weeks now and finally on Thursday, March 26th Cohort 5 was able to see just what that meant for Drifters and Sultans of Scratch. Development documents, printed and bound in a tangible format, were handed to Rick to become the plan of action for the next 4 months of game production. The game demos shown were a representative view of how the completed game will look, feel and play like...literally a 'slice' of the final game experience.

The teams had 6 deliverables: Gameplay Target, Visual Target, Risk Analysis, Pipelines, Design Document, and the game demo.

Sultans of Scratch
The SoS team described their gameplay target as "A rhythm game with non-discrete instructions for input; freedom to input random combo moves." Their visual target was stated as "urban, aggressive and stylized" - a statement they had stuck with since early on in the pre-production of their game. The pipelines are in place for art with a ramp-up document written to help document the export process and file structure. Design has the full design document, plus the programmers have been working on a tool to simplify adding action notes and developing combos, so the design team can make changes to song note data without having to consult the programming team on the actual code changes. The programmers are working heavily with XML and developing tools to streamline processes like the note additions.

As for risk analysis, the design team stated the difficulty of the freestyle system was an issue they were paying close attention to and their current safety designs to combat that was the face that combos are not required, any series of buttons will execute positive results on notes, and also a tutorial feature to allow players to practice learning the combos. Art's major risk was the cleanup time on the large number of animations the game will need for the various dance moves and idle states. Their solution was the addition of Neal, the cohort's resident expert on mocap, to their team as a full time animator. Programming's risk was the custom peripheral as their current build relies on keyboard and mouse controls to play right now and they are unable to test the peripheral controls until it is built. The team is working closely with the UCF Engineering department to establish an early milestone for a functional prototype to be created so it can be implemented and heavily tested. They are also creating a backup plan to control the DJ tables with an Xbox 360 controller in the event the peripheral cannot be realized within the next 4 months.

The demo shown on screen showed a basic menu structure that branched to several choices such as song and character selection, difficulty levels, and other features to be implemented. The environment was fully textured as was Mr. YinYang and they were both fully functional, with lighting and complete rigged animation. The dancer combo system shown was working with certain combinations of button presses on the numbers of notes would execute specific dance moves.

The DJ mechanic saw a final change before being committed to the design document: the turntables now have 4 distinct moves - rotation clockwise, counter clockwise, fader switch left, and fader switch right. The final plan for the game emphasizes the DJ/dancer relationship again with either player performing poorly causing missing notes, lower scores and such. Teamwork is heavily rewarded, though, with good performance creating more 3- and 4-note streams for the dancer to perform combos on, almost like a skill increase for awesome performances. Some of the visual awards being implemented include spotlight effects, the DJ being able to add special particle effects to the dancer, and crowd participation (the crowd shouting for dance moves that if executed will earn more combo points).

With one character and stage modeled the art team is working on additional characters and potential environments, as well as props and other visual treats to give the game the most authentic club feeling they can. The programmers were able to meet with the Engineering department recently and got to discuss the peripheral design and functionality. Design team is working on music tracks, note streams, and preparing for a focus group test for various visual and gameplay element analysis later on this month. A lot of positive movement coming along, and their momentum is still building!


Drifters
The Drifters gameplay target is described as a "stealth based hunt with tactical gameplay elements where the player is both the predator and the prey with enemy players hiding in plain sight." Their goal is to instill a feeling of paranoia and unpredictability, especially through the environmental lighting and audio. Their demographic is neither hardcore gamers nor casual ones, although they do focus on the online gamer as a target audience due to the networked nature of their game.

The visual target for the Drifters is to create iconic silhouettes with the detail of the drifter form coming from their particular back story. The drifter aura will be a striking, glowing contrast to the rest of the body and environment, centered around the head and upper body. The museum environment is based off the London Museum of Natural History with a very sophisticated, elegant style. The fictional setup is that it is opening night of the museum, leading the crowd present to be dressed in more formal attire and act in a controlled manner, making it easier to blend in with a tranquil crowd. This traditional presentation of a museum scenario populated with ethereal beings creates a blend of the expected with the paranormal to create a compelling story and environment for the game to take place. The goal with drifter vision is to create the feeling of viewing another plane of existence unique to the Drifters. When active, the screen de-saturates except for the drifter auras, creating a sharp contrast between the normal environment features and the glowing, vibrant drifter auras and abilities.

Some of the safety plans they've built into their development plan include extensive QA testing using multiple machine setups, system analysis, and the use of DevTrack for eventual bug reporting and fixes. The programming team set early deadlines for networking and have hit all their milestones so far, allowing extensive stress tests and implementation to take place now so their main mechanic can be preserved. The art team has taken care with their designs to keep their poly count low and load times minimal, considering the depth of detail on their environment as well as the scale of the space and crowd.

The demo shown was played with 4 separate laptops each running the game in a LAN setting, allowing 4 Drifters to intermingle with a crowd of about 30 NPCs. Load screens were implemented to allow for everything to set up and run smoothly during gameplay. When switching into drifter vision, all characters were shown to glow white with NPCs staying that way and fellow Drifters slowly starting to re-saturate with detail, eventually giving them away to the player. The environment was completely base textured with the final texturing at about 50% completion at that point of development.

The team stated that their basic game cycle was a process of scanning the environment and getting a lay of the land, drifting into a body, switching to drifter vision to get an ability, finding and stalking a target, using the ability to make a kill, drifting again, and blending back in to the crowd to start the cycle anew. Of course, the game is open enough to allow for any sort of player strategy to emerge, as some more aggressive players will want to make multiple kills with little drifting, while others will want to play a bit more stealthily and use their abilities to their maximum potential. Drifters is shaping up to be a very open-ended sandbox of play, with no clear advantage to playing strategically or spur of the moment, allowing for a great expansion of player types. Their progress has been steady and shows to continue to bigger and better things within the next months!

Pre-production week 9

This is it! The final week of pre-production before the vertical slice presentations are made and the teams transition into production. This week was spent revising, reviewing, analyzing, testing, and iterating on the major points of the games – there were major changes made, upgrades, and in some cases total character switches. The teams are confident in their choices and move on to get everything implemented for a playable demo that will capture the feel and playability of their title to the point where production can be greenlit and we’ll start to see some major assets coming down the line!

Drifters
When the new team members were integrated into their roles, they were given documentation on all aspects of the game, and one major point of confusion was the actual game mode, “capture the kill”. After much discussion, iterations and analysis, the mode has been adjusted to create a more simplified and easy to grasp concept. While the new game mode is still point-based depending on kills, making a kill will now earn you another ability slot while being killed will cause you to lose one (you always have one static slot). After the 10-minute play period, players enter a sudden death mode as the museum closes and the NPCs filter out, leaving only the Drifters behind to enter a melee battle for maximum points. This creates the chance for 2 potential winners – most kills during the match, and most kills overall. This simplified combat system, with 1-hit kills, lets the player focus more on stealth and strategy rather than just “running and gunning”.

To follow with their overhauled game mode, the drifter abilities were also looked at with a new light. The focus is to have the abilities be more of a lure or trapping mechanic, rather than a damage-style advantage. One ability they demonstrated was to have a Drifter able to place a normal-looking object within the environment that would have a radius of effect, trapping any other Drifters that entered it for a period of time and rendering them immobile. In normal vision, the object (a chair, banister post, etc.) would appear normal, but in Drifter Vision it would have a faint glow to it, hinting that it was a trap.

Drifter Vision is also undergoing scrutiny to make sure it is balanced and effective as a major mechanic of gameplay. The tells given off in non-Drifter Vision have been de-emphasized to allow for better stealth operation. Players in DV will be able to see kills being made by others as particle effects surrounding the killer, while in non-DV they’d just see one person lunging at another. The de-possession of a body is also a new mechanic being looked at. A Drifter who possesses a NPC who was recently possessed by another Drifter will begin showing tells more quickly than if they had possessed a more dormant NPC. While this hints to other Drifters that there’s now a Drifter in that body, it hints to the Drifter who just entered it that there is another Drifter nearby who just jumped bodies.

The programmers are working steadily toward their goal of networked play and are on a good track with message passing functional out of engine in another test bed and ready to be implemented in engine for further testing. Their secondary priority is activating the Drifter Vision during gameplay. The artists are working on making two full-body Drifter meshes for the Drifter Vision states, as well as more textures to create a more diverse crowd. The next mocap shoot is scheduled for today to create more walk cycles, idle animations, and individual quirks of various NPC classes, which will also help in the diversity of the crowd behavior. The current build of the game shows a variation of 3 walk cycles on various combinations of textured NPCs, all running full screen at 60FPS.

As the vertical slice comes together, so does the documentation for production. Once things make it into the end of pre-production, they are set into the production documentation that will become the team’s guide that drives the entirety of production. For next Thursday, the team will have their programming guide, final design document, and art style guides ready to be committed to the final gold master state of the game. The team has posted their 4-month calendar of production with milestones for various deliverables, alpha and beta stages, and other important dates so everything can be kept on track for the remainder of Spring semester and Summer!

The 4-month plan

Sultans of Scratch
The major change on the Sultan’s side was the analysis and redesign of the GUI. Several options were mocked up and tested with each version having its pros and cons. Some designs allowed for ease of note reading, but blocked the views of the dancer, while others were quite the opposite. The challenge was to find a compromise that allowed the notes to be easily seen while not cluttering things up too much and allowing the players to enjoy the visuals of the dancer and their moves as well as the environment. The team finally settled on a design where the turntables were at the bottom of the screen with the DJ notes streaming top to bottom and hitting on the records to be scratched. The dancer’s notes would stream up the middle with everything stopping at a mid-way point so the dancer could still be seen performing their moves. Camera movement with the GUI is also a point of note, as some of the dance moves get a little wild and need to be followed with the camera. While the turntables will stay anchored at the position on the bottom of the screen, the main visuals of the game will be followed by the camera.

Scene design concepts

The note streams in place

Lots of the current art pieces have also gone through some changes. The current turntable design was also improved, adding more flair to the inside of the case including some nice flashy LEDs.

The improved turntable design

The new character concepts from last week were improved upon, one so much that he was not only given a full profile, but then deemed the new main character of the game, replacing Mr. Hat as the ‘mascot’ character.

Some of the character designs, improved and detailed

Mr. Ying-Yang's full character profile

Mr. Ying-Yang's various action poses with graffiti sprays

“Mr. Ying-Yang” has now been modeled and rigged and even without normal mapping, UVs, or texturing, he performs much better than Mr. Hat did. His body structure is a bit closer to Nelson (the break dancer who provided the mocap data) and therefore works more fluidly with the animations. Mr. Hat’s less-realistic proportions caused deformity issues during some of the more complex moves, and therefore Mr. YY has replaced him for the vertical slice.

Mr. Ying-Yang's design sketch

The team plans to have all their schedules and pipelines for the 3 teams in place for Thursday, as well as a finished development plan written in draft form by the 23rd and finalized then. They are also doing another mocap shoot to generate various behaviors and movements for the crowd that will be surrounding the dancer’s stage.