The third usability test has been performed. Mockup screens can be found here.

The most important changes are:

• More intuitive navigation
• The user doesn't have to confirm as many actions as before
• The music view now shows a list of songs that are going to be played in the near future
• Neighboring pubs can now be selected on the map view as well
  

Overall, most testers found this version to be a major improvement. There were however still a few remarks, such as:

• Lack of social networking
• Facebook/Twitter messages should be published automatically. Some others don't want this. This could be fixed by allowing the user to configure this behaviour in the settings panel.
• Facebook/Twitter login information should be stored for auto login

The lack of social networking features remains the biggest problem. While at this point it isn't the main purpose of the application, some interesting social features could be:

• An online user profile containing the user's achievements, visited pubs, etc. This could be turned into a facebook application.
  
• The possibility to see other users in the area and invite them for a drink at a certain pub.

Another interesting idea mentioned is to make a study about how much people value their privacy regarding social applications. The social features could be based on this information.

Currently, I'm creating the basic screens of the application. The plan for next week is to implement the possibility to receive a list of pubs together with info about the music they're playing.

A second usability test has been performed based on the new application. The mockup screens can be found here. The most common remarks were:

• Too much navigation (too much clicking on back buttons, etc.)
• There should be a direct link between a pub and its route description, without the need to view the pub info in between.
  
• Whenever an (not incredibly important) information message is required it should pop up and disappear automatically, without the need for the user to manually confirm every message. An example is the message that tells you your achievement was successfully published on Facebook.
• Use icons to indicate sharing on sites such as Twitter and Facebook.
• Show the playlist of a certain pub instead of just the current and the next song.
• While viewing the route directions, show other pubs on the map and allow users to switch to another pub if desired.
• Achievements should have a better purpose. For example, they could be shared with other users.
  

The final topic of the iphone application has been chosen. The application will be partly based on the Stella Artois 'Le Bar Guide' application, but it will focus more on music. The features of the application will be the following:

• Receive a list of pubs in the area. The radius of the area can be configured. The default radius could be up to 2 km. Instead of a list, you can also use a world view (camera view) showing your current environment overlayed with labels showing pubs in the area.
  
• You can select a bar from the list and hear the music that's currently playing in the bar.
  
• You can ask for route directions to the bar. The route will be presented on a map. The music will play in function of the route. When nearing the destination pub, the music volume will grow louder. When moving away, the music fades away. Instead of the map, you can also switch to the 'world view' which will show you an arrow and the remaining distance towards the pub.
• You can filter the pubs shown by the genre of music they are playing or by the similarity with a certain song. You can also rate pubs and order them by rating or by the average amount of visitors they get.
  
• You can indicate your visit to a certain pub and share this on Facebook or Twitter. This way, you can invite your friends over and in turn they'll know where to find you.
• You can gain achievements and share these on Facebook or Twitter as well.

The achievements aren't fully decided upon yet, but some possibilities are:

• City Hopper: Visit pubs in at least 5 different cities.
• Pub Hopper: Visit over 5 pubs in less than 5 hours.
• Restless: Visit more than 10 pubs in a single city.
• Recognized: Visit the same pub over 10 times.
• All the king's men: Visit at least 50 different pubs.
• Home sweet home: Visit the same pub over 50 times.

Another idea, based on this presentation about combining music and colour synaesthesia and Moodagent:

The idea is to assign colors to songs. The color you pick can be based on intuition, a memory of the videoclip, a color in the song title, etc. The typical features of the application could be the following

1. Browse the musicotheque (Tunify music library and assign a color to each song you play
2. The songs in a playlist are annotated with the color you picked.
3. Users can annotate each color with tags (to express what they feel when thinking of it)
4. Playlists can be generated by selecting one color or a range of colors.
   
5. The song - color combination could be stored in two ways (maybe both ways can be used?):

• Locally: each user has a personal database for storing his song/color information. The advantage of this is that the used colors and the song genre will be more consistent and the user has no songs from others interfering with his music library.
  
• Globally: This means that when the user generates a playlist by color (red, for example), there might be songs in it which he personally would tag with another color, such as green. This means that he might not get the music he expected, as people in different areas of the world experience different feelings with certain colors. Nonetheless, it might be an interesting way to discover new music and to see what other people assign to colors. Also, this way statistics could be generated which could show the most popular songs for a certain color, or a diagram of all the colors ever given to a certain song and their impact.
  

I think the information should definitely be stored locally, to give the user full control over his music. The possibility to access color/music information on a global scale is optional, but could be very interesting.

Wednesday was the second thesis presentation. The purpose was to talk about the usability testing performed with a first prototype of the application. The slides of the talk can be found here (PDF).

The created prototype for the usability test was a basic and lightweight tunify player. The pencil & paper sketches can be found here. However, the current version of the application is too much of a standard iTunes application, with the typical playlist, song and library features such as creating playlists, browsing and adding songs, etc. So the task for the coming weeks is to come up with a few more interesting ideas and make the application revolve around these ideas.

A few thoughts for ideas are:

• Use the iPhone GPS functionality to get the current position of the user. Present him a map showing the pubs using Tunify in his current neighbourhood. The user can then select a pub and listen to the music playing there in real-time. He could also ask for directions.
• Annotate the world with music. User A could be taking a walk in a park and suddenly think about some song which he thinks fits with the environment. He could then take out his iPhone, and 'annotate' the location with that song. The song will then be coupled to his GPS location. Later, user B might walk near that same spot while listening to the iPhone application. When passing the same location as user A, the song user A annotated will start playing. A location can be annotated with multiple songs. Users could for example configure the application to play only classical songs annotated with locations. The songs could be coupled with google maps, so you can 'see' the music of a certain location and maybe even pick the path you'll walk based on the music you'll pass.
• Use the GPS to get the current location of the user. Then gather information about all the artists within a certain radius. The application could then show a 3D sphere with the latest songs or albums of these artists, each placed in the general direction of the artist's location. The user is positioned inside the sphere and while rotating in the real world, the sphere rotates around him and shows the work of the artist located in the same direction the user is facing. This information could be overlayed on images of the user's environment made with the camera.
  

I've recently got more info about the Tunify API, which can be accessed through the use of SOAP requests. There are about 40 methods available, such as 'loginUser', 'getMusicCollection', 'saveCalendarItem', etc. So far, I've got most of them figured out.

I also got some practice with iPhone development in XCode. I only managed to do the very basics so far, such as entering text and responding to buttons. Next week I'll try to get a lot more practice. I'm also on the look for a good SOAP library for the iPhone, as apparently there is no native support for it.

Finally, I'm now in the process of creating the (possibly final) versions of the application user interface and views. The idea is to perform a usability test somewhere during the course of next week with about 5 to 10 people, after which the design can be refined where necessary.

Recent highlights

Once again it's been a while since I last updated this blog. I tend to forget I have one.

Some of the highlights of the past few weeks:

• The basic idea for the iPhone application is known. At a basic level the purpose of the application is to create a lightweight version of Tunify. So the application could allow you to browse and edit your playlists, browse the music channels, look for similar music and maybe allow you to schedule your playlists as well. Besides that, the idea is to create some interesting extensions. Possible ideas could be to include a social networking component, offering users to share songs and collaborate in the creation of playlists or basically anything that sounds interesting and has value for the user. It could also be nice to do things with the built-in GPS of the iPhone, the time of the day, the weather outside... I'll think about these things the moment I've finished the literature study.
• The first presentation has been given. The purpose was to give an idea of the work that has been done so far and the future planning. The slides are available here (in pdf format). Note that there is a little error on the sixth slide: Ocarina does not allow you to listen in real-time to performances of other users. It does however, try to select the most recently played tune.
  

The plan for now is to finish the literature study by the end of next week. After that, I can focus on the application development part and start tinkering with both the Tunify API and the iPhone SDK. This is just for research purposes. Before I can start programming the actual application, I need to specify the exact details of the application and evaluate (together with others) the usability and usefulness of the desired features, application screens and general flow of the application.