Wrap Up

I am finishing up the documentation for the Kinect game, as well as packaging the projects I have worked on this summer. The goal is that, at the end of this process, a newcomer would be able to familiarize themselves with the game's scripts and be able to delve into editing and improving it without too much downtime from getting acquainted with the code.

To accomplish this, a set of current achievements, problems, and future goals will be laid out along with explanations of each.

Update 1: At this point documentation of the code is as complete as it will ever be, and finalizing the list of goals, problems and achievements is underway.

Report, Advice for Future Work

I have begun summarizing and synthesizing a report for those who will come after me to work on this project; hopefully I will give them a starting position that helps them be more familiar with the game and its functions/mechanisms than I was at first.

Informal Game Testing

Nick's Kinect game has reached a level where it is usable for play-testing. Some graphical adjustments are needed, and some unintentional gameplay mechanisms should be edited/removed, but the game is usable. Critiques and opinions on the control scheme have been collected orally, as a proper debriefing questionnaire has not been developed.

Opinions are varied, as are the subjects themselves. True conclusions cannot be drawn from such a basic method of data collection but it is good to get a general idea of the impressions the control scheme leaves on the SIG lab's occupants.

State of the Game

As of today, Nick's game has been improved in these ways:

Collision with "destroyed" objectives will no longer reset the course; these objects will no longer be collision targets until the course is reset by colliding with the start sphere. This was achieved by disabling the collider for the relevant game objects as they are triggered.

Current problems are that the timers update each time the end objective is triggered. They should only trigger once. A fix for this is being looked into.

Collider Experimentation

Multiple manipulations of the objective spheres' collider status has yielded various, though undesirable results. More info later.

Minor Progress

Today the projectile objectives have been improved to require each one to be hit in sequence before the timer reports how long it took to shoot them all.

Objective investigation

Today I begin to use the unity-side GetComponent commands and isTrigger booleans for making targets not trigger once "destroyed". The purpose for this is to repair a day one bug where destroyed/passed objectives will trigger an out of order response, resetting the course. With this fix, the player will not accidentally reset the course by hitting invisible objective targets.

Setbacks

Unfortunately an error has sprung up where the course does not properly reset when the starting sphere is touched. This is being looked into, and when fixed, will bring the game to full functionality with its new shot counters and secondary objectives.

Update 1: Reverting to a previous version has brought the game to a point where it works so long as the player does not hit the projectile targets after they've been destroyed. Monday's work will consist of researching and implementing methods to make the unrendered targets unhittable.

Shot Counter Success, Project Milestone

I have found the function that signals when a new shot is to be fired, and attached the shot counter to it. This method circumvents the problem where shots were being counted multiple times across multiple frames. Further, I have added the shot counter in both the Kinect controller's primary fire function as well as the mouse and keyboard fire function; all control schemes will count shots properly.

Sifting Through Nick's Game

Today and tomorrow I will be searching for the section of code responsible for instantiating the primary projectile prefabs in an effort to attach the incrementing shot count integer to it. This will circumvent the issue that has arisen from attaching the shot counter to the frame-updater: sometimes a single shot is counted multiple times.

Bullet Counter

The bullet counter has been added in today. However, there are bugs that need to be worked out: the counter invariably states that more shots were fired than is actually true. This is because sometimes shots are counted twice; this will be looked into tomorrow.

Nick's Game, Yesterday and Today

Yesterday I successfully created a separate loop of timed objectives. Today I have successfully added that separate timed loop into the larger timed objective loop. When played, the game will report your time taken to shoot the three destructible targets.

Nick's Game Restored

As of today timer functionality has been restored. The current state of the game: keyboard controls are active and fine-tuned on the timed-course level. Tomorrow further efforts will be made to add the smaller-scope timer as well as the corresponding shot-counter.

Modifying Nick's Game

Today I've, with help from Nathan, successfully imported the keyboard control scheme onto the Pit Level timed course for testing. Numerous movement speed / mouse sensitivity adjustments have been implemented. Tomorrow's work will focus on restoring the timer functionality associated with the startscript, endscript, and testscript.

Much of the time spent working on the project is spent learning what the scripts, scenes, and objects do and how they relate to each other. More in-depth documentation will remedy this.

Dry Runs, Final Presentations, Asteroids Completion

Today I've reached a point in the Asteroids tutorial where it can be called finished; I now apply my Unity skills acquired during the Asteroids tutorial towards adding the desired increased functionality to Nicks' game. Additionally I have attended a dry run of Joe Kider's defense and Alex's character AI presentation.

Last Tuesday and Thursday

Last tuesday Tuesday I continued to tinker with Nick's game code, but was having trouble with some of the fundamental functionality of Unity. To fix this, I have put myself through a Unity tutorial to make a simple Asteroids game.

Timer and Shot Counters, Nick's Game

Of the people here in the lab and with Nick away in Germany, I am the most familiar with Nick's game project. Because I am to add in a timer and shot counter to the game, I must familiarize myself further; today I am searching for the timer function in order to replicate it elsewhere. Specifically, it will record the time interval between the first green target being hit and the last green target being hit. The shot counter will be active in between these events, recording the amount of shots fired in order to hit the three targets.

Previously I have been working with the Asteroids tutorial, which will be described in later posts.

Yesterday and Today

The SIG presentation went well on Tuesday, and Nick's project continues to be worked on, now through Unity. Efforts to migrate to Adobe After Effects have been impeded by various issues. Yesterday my troubleshooting of the front monitors has so far proved effective; VLC is not consuming as much processing power or as much memory as it was earlier this week. The front monitors' computer has run stably, looping for several hours.

However, the interesting problem of multiple instances of VLC running has not happened again; these multiple instances seemed to be the chief cause of increased consumption by VLC (and perhaps the cause of the repeated crashes). The methods I employed to fix the computer's repeated crashing did not focus on preventing multiple instances from running, so I will continue to monitor the front computer to see if its looping starts up multiple instances.

Nick's Kinect Game, changing video editing software

Today Nick's game was successfully built and run through Unity. The control scheme is still unwieldy and will be worked on.

I plan to switch away from Camtasia to Adobe After Effects.

Shorter Version

The smaller version has reached a level of quality similar to the longer version. At this point it will be sent out for review. Attempts will be made to fix a possible memory leak error that is causing the infinitely looping video playlist to eventually crash.

DMD and CCGT Selected Technical Work

The official name of the DreamWorks demo reel, which is now completely finalized, is ready for distribution. It is a good thing it is done now, because the trial version of Camtasia is ending soon. Future projects will use  different video software.

Update 1: I will create polish the short version and bring it up to the same standards as the longer 4 minute version.

DreamWorks Demo Final

At this time, 100% of those who have gotten back to me with critique for the latest version of the Demo have had nothing but good things to say.

Camtasia Difficulties

After several hours of troubleshooting, the cause of rendering run-time errors for Camtasia was identified: Camtasia's trial version will not directly render things into the mp4 format any longer (it used to, similar to how it used to accept more video formats). With only a few more days left on the trial version I am going to advocate purchase of Camtasia second, and purchase of a more proven software first. If the trial version is this unusable, what can assure me the purchased version will be different?

Update 1: Attempts to find a workaround have been successful; the most recent project version was successfully rendered into the avi format and then converted into an mp4. Tomorrow I will check to see that all the recommended changes have been properly implemented.

DreamWorks Reel Finalization

Today I begin the final accompanying audio track for the DreamWorks demo reel. This track will be in sync with the visuals.

Today's reading group discussed the Georgia Tech's swimming simulation paper.

Nick's Game

Today I continue to look at Nick's code in an effort to better understand it. The main objective of this familiarization is to find how to activate the in-game GUI for recording new gestures and mapping them to actions. Another objective is finding possible causes of error when attempting to run the game on adjacent computers (which show a lack of received input from the Kinect motion sensor).

In addition, I have created a 2-minute version of the DreamWorks demo reel.

DreamWorks Demo

The DreamWorks demo is wrapping up, and will soon be approved in its current state. At that point I will add an accompanying audio track, in sync with the visuals.

Camtasia continues to be troublesome, and efforts to replace it with Adobe's equivalent are underway. As far as this project is concerned, migration to the other production software might not be sensible given the approaching deadline.

DreamWorks Demo Reel, Troubleshooting Camtasia

Editing continues, with the new clips of Dan Knowlton's work being added in. Addition of a soundtrack is still on hold pending finalization of the visuals. At this point the new clips do not have titles, but research will be made towards titling them appropriately.

Overall it seems that Camtasia (or at least its trial version) fails to import media properly unless it is in a select few (.wmv, .mov) formats. Further troubleshooting is necessary; I doubt this limited usability is by design. Camtasia continues to seem unstable as a program, where the slowly degenerating functionality can be restored by restarting the program.

My optimistic assumption is that this is due to my using of a trial version, or due to the limitations of this PC running a large project through 32-bit emulation with other programs running in the background; however, there is no indication that the PC is limiting Camtasia's capabilities. In any case, having these unstable trial-version experiences is unimpressive.

Demo reel milestone #1

The third rough-draft has been sent out and critiqued: a list of suggested edits will be implemented. These include:

Shortening some clips for improved pacing 
Reducing the amount of Karl
Improving text placement
Addition of title slides
Overall improvement of pacing
Addition of some clips

Updates will keep track of editing progress.

Dreamworks Demo

Today I have been working on the Dreamworks Demo, rearranging its clip order as well as working on syncing the accompanying music to the visual action. Additionally I am troubleshooting a neighboring computer which is not properly reading Kinect input, despite having the Kinect SDK, drivers, and runtime installed. I will look at Nick's input script to see if I can find anything there; updates will keep track of editing and troubleshooting progress.

Extreme Editing, BBQ Lunch

Today, a three-minute version of the ninety-minute demo reel was requested. A rough copy has been produced and will be reviewed and improved based on the pending critique. At 12pm I attended an indoor barbecue lunch.

Update 1: The reel will be further streamlined, and text-boxes will be added in place of title-slides to give credit to contributors.

Demo Reel Editing, Nick's Game

In between changes to Nick's game, I will continue editing the demo reel. The audio track will be completed, and pacing will be improved through cutting and the addition of transitions where appropriate.

Update 1: The first round of editing is complete. Editing will resume after the current reel is reviewed, if it is thought necessary. Attempts have been made to bring Nick's game to a neighboring computer, but there are complications with the Kinect software ( drivers are installed, but there remains a lack of input, or interpretation of said input by Nick's game) which are impeding progress.

Update 2: Peer review of the demo reel compilation is in progress. Voiced-over PowerPoint will be removed and replaced with single-slide text introductions. The resulting demo reel will strike a healthy balance between technically impressive and visually interesting.

Motion Capture, Nick's Game

Today Nathan will be the subject of motion capture research. Joe Kider will demonstrate proper practice which will be used in future testing. Biometric data such as heart rate and body temperature will be collected and analyzed in order to aid the development of realistic modeling practices. Later it is expected that I will continue with the familiarization, preparation and eventual utilization of Nick's game. Further updates will enumerate the specific actions taken in order to ready his game for objective and subjective testing.

Update 1: The GUI for the "Level 4" control scheme (which uses hips and leaning forward or back for aiming, as opposed to moving one's right arm) is not showing, and I am currently looking through the relevant sections to find what I can change to make it visible.

Update 2: Observation of Level 4's control scheme shows that while it is effective if used properly, its learning curve is a tad steep. The recommended improvements should be carried out, though it is encouraging to see improvement of game playablility after only a few sessions.

Demo Reel #2, Nick's Senior Project

The focus of today's work will be the reel's audio track; this will include the addition of music to silent sections as well as editing of voiced sections. Steps will be taken to ensure that those who contributed their final projects to the demo reel will be credited; text revealing the creator of each work will be added where necessary.

Update 1: Equalization of existing audio is complete, addition of music accompaniment in progress.

Update 2: Met up with Nick and practiced using the Kinect control schemes to play his game. Testing will begin after the controlling gestures are improved, and the users are more familiar with these gestures.

Demo Reel #2 Prep

The first order of business for today was to determine which file types Camtasia Studio likes; eventually all video media was converted into .wmv or .mov files. Afterwards the Camtasia project's presentation timeline was manually populated by me to include all pictures and final demo videos. Later updates will keep track of minor audio-editing progress (some editing progress was lost due to scrapping the earlier Camtasia project that did not have a more homogeneous set of video file types, meaning only .mov and .wmv for video file types) as well as video editing.

Update 1: Some video clips have been sped up and/or cut; the same goes for their respective audio tracks. Transitions have been added in between each segment when possible; the speeding up and cutting cancels out the time added by the transitions (3 seconds per fading transition). Equalization of the overall audio is a work in progress.

Demo Reel #2 Start

Today I started collating the various pictures and videos that will become the second demo reel. Through the use of Camtasia and Photoshop I will combine photos into single slides and then present them to music, alongside the final project videos.

Update 1:
Many of these videos have voice-overs set to music already; however, their levels are different and need equalizing. Similarly, certain audio tracks will be edited in order to minimize annoying white noise. In addition, several videos of varying formats cause error messages to be returned by Camtasia upon attempted importing of these videos. Attempts to download a file converter and then convert the offending files without quality loss will be performed tomorrow.

Update 2:
As of this update, all compatible videos have been imported and added to the Camtasia project timeline.

First Weeks

One of the first tasks I undertook was the setup of the front monitors' looping video playlist. Given a lightly edited demo-reel without audio, I finished up the editing and added music tracks. This demo reel was then added into a playlist with another video, the playlist being automatically started up by a small .bat file.

After this editing and minor coding were concluded, I began exploring the Python language through a script associated with delegating work to the render garden's servers. Eventually we narrowed the error down to a single line: when describing the arguments in a call to os.spawnv, an error message, explaining that no such file or directory was found, appeared.