A bit late, but here’s my Urørt project (mostly in writing, hopefully you’ll get it).
URØRT STAR
by Lars Marcus Vedeler
Rich Social Media 2008, Oslo School of Architecture and Design
INTRODUCTION
Background
Urørt is a channel for unsigned Norwegian bands and artists where they free of charge can distribute their music on the web. From a listeners point of view this means that an abundance of unknown music is up-loaded every day, just waiting to be discovered. But this amount and accessibility is also a problem for finding the good songs between all the others, when they’re in fact unknown.
Problem Definition
My problem definition for this project is angled towards the listeners and the browsing of music:
How to preoritize and visualize music on urørt for easing and increasing explorativity for the user?
Motivation
When using the Urørt web pages as a new listener the problem with browsing through unknown music seemed as the main concern with the site.
STATE-OF-THE-ART/RESEARCH
Data amount
Urørt’s site contain thousands of songs, mostly unknown for the user. When trying to present that much data in a understandable, informative but fun way, it’s interesting to look at other sites which deals with just as much data, or even many, many times as much.
We Feel Fine (www.wefeelfine.org) shows huge amounts of feelings through extractions of blogpost. The application doesn’t prioritize, but it categorizes the posts in different shapes and colour, depending on their content. Another way of presenting huge amount of quite different items is done in different ways on Etsy (www.etsy.com). In this site the photo of an item is one of the most decisive factors for the user’s search. In some of the search modes, a 3D space also plays an interesting role in giving the user an impression that the amount of content is endless.
Organizing and connecting data
Visualize and animating connections between different data is often shown by drawing lines between them. A project that illustrates this in a refined way is SocioPatterns (www.sociopatterns.org) which deals with moving RFID-tagged people in an office, and their interrelation.
Moma’s Elastic Mind (www.moma.org/exhibitions/2008/elasticminds) exhibition has another presentation of design projects in a large 2D area. The organization of projects in different categories also works as suggested you can connect to, from one project to another.
Navigating
There’s many different ways of doing a zoom user interface (or ZUI). Because navigating in this concept is quite similar in how you navigate in web-maps, I draw quite lot of inspiration from stereotypic ZUIs, like the one in google maps (maps.google.com)
PROCESS & METHODS
Urørt
One of the most important ways of finding shortcomings and possibilities in the urørt web-site, was by using it. Signing up, navigating around and listening to songs was crucial to understand how and why the site was designed as it is.
Surveys
Surveys made by Record and NRK Forskningen gave guidelines for approaching the concept stage at first, but was also to some degree a reminder to return to after some time.
Sketching and listing
When developing the concept, sketching became important as a method. Drawing up everything from the overall structure of the page to smaller components like buttons, as well as animation sequences.
Because of limited time, listing up every step in both the animations sequences and more general site behaviour, was also very helpful.
Mapping
Especially in the categorization of the eighteen genres the mapping of these were crucial for, among other things, the size (in percentage) on the night sky. Also taking random sample within every genre gave an impression on how the genres were linked together for placing and blending them in the right areas.
Lectures
Several lectures assisted in the development of this concept, some of them are: Jon Olav Eikenes’ “Navimation” and Meg Pickard’s “Social Media”.
RESULTS
“Urørt stars” is a different way of browsing through the thousands of unknown songs within the web site. By using different filters the user can emphasize his/ hers preferences in music.
Stars
Instead of presenting the songs through lists with titles, names pictures etc., “Urørt stars” visualizes the songs as dots, or stars- on a night sky. One star will have a specific colour, according to which genre it has been tagged to. They can also be separated by filtering them; the stars will appear lighter and stand out more according to in which degree it fulfils the filtering preferences done by the user. The star metaphor may seem quite cheesy at first glance but holds a lot of good potential.
Genres
The 18 genres is an important, already existing way of roughly categorizing the thousand of songs. In this concept the approach is not to filter these away when searching for music, but instead display them together in groups at the night sky in different colour. This has both pros and cons, but the overall idea is to not narrow down the usage just by genre, but instead encourage listening between genres.
Layers
Moving into the universe of urørt, a feeling of an endless amount of songs is desired (which of course is not the case). The stars are also organized in different layers. A selection of stars are always shown, as a basic set, and the further the user moves into the night-sky by zooming, more and more stars appear. These are stars that may be older or maybe less used, but could be just as good songs as the newer.
This layering is an effective way of displaying the stars; preventing an overload of information for the user, but also for keeping the data information for the web-browser and bandwidth down.
Traffic
Another way this concept can increase sociability is by showing peoples traffic. Whenever a user listens to set of songs, their will leave footprints in the shape of as playlists (the lines between the songs). This gives the user information about where other users are moving within a given times pan. The traffic is, as the stars, also in layers, so the further you zoom in, the older and more traffic comes to display.
Traffic in the opening window is though different, and not inlines, but displayed as a fog, just to give a hint of the general movement.
Playlists
When listening to music on “Urørt stars” the user select filtered stars. By clicking one song after the other a star map is drawn between the stars, this represents the users traffic, but is also his/ hers playlist.
When logged in, a user can save their own playlists and share them with friends, who themselves has their own playlists. During the making of a playlist users can also add others playlist as an extension to their own.
In this sense, all traffic is playlists, which makes them clickable, and listenable. With other words; all users has access to every playlists.
The Filters
The filters play one of the most important roles in this concept. With these options the users can emphasize the values they want from time to time by lighting stars up.
Although listed together the filters can be divided in three different groups; the first controlled by the Urørt crew, the second more on a basis generated by the content and users, and a third group which help search’s for extra content.
The filters in the first group are:
-Played on P3
-This weeks (and some previous) urørt artist
-Urørt recommends
These options give has a certain promise of quality, since it’s the site’s facilitators who picked these songs out.
The second group contains the following:
-New/ freshness
-Popularity
-Number of fans
-Nearby (geographical)
The new filter in this group is a factor the Urørt site today lacks. Since there’s constantly new song up-loads, it’s important to give these a fair chance and letting the user separate them from old but popular songs, which is not the case in today’s list-system. Popularity and number of fans let’s the user know if this song is a liked by the public and even if this artist has a lot of fans, which indicates a quite good artists, since getting a fan is sort of a privilege. The nearby filter is a tool to find local bands.
The third group:
-Video
-Concerts
These filters are meant to replace sub-sites on Urørt. Since both video and concerts are extra material a band can up-load and post, these can easily be implemented in the filters.
All filters works separately for different searches, but also work combined with each other to specify. F.ex. to combine nearby and concerts or popular and videos.
Moving around
The layering of the stars leads to a quite large area of songs. It is therefore absolutely necessary with a set of tool for moving around. The tools chosen for this concept is a well known zoom and pan system, consisting of a zoom-out, a zoom-in buttons for navigating the layers, and four directional buttons for moving in the x- and y-directions. Mark that these are arranged like the arrow keys on a keyboard which then of course can be used instead of the mouse.
In addition to this there is an automatic zoom and pan option, which is default, but can be turned off. This tool interprets the users choices and behaviour and tries to move around according to this. F.ex. when choosing a song, the auto-mode zooms into this genre. But if the next song is in another genre is picked, it zooms a bit out. The auto-mode does though have a secret agenda; it is constantly zooming in, this is to force a bit of extra explorativity to the user into the star map.
DISCUSSION
Static vs dynamic
In the choice of the content (the dots or stars) being static, as they are, or, dynamic (somehow moving around) I think the three most decisive factors where my wish to keep some degree of order in the system, the layering and categorizing of data, and the playlists. In many sites dealing with large quantities of data (as with earlier referred to We Feel Fine), the graphics often float or bounce around. I think the question on if this should be the case is really how random you want the content to appear to for the user.
Genres
Another fundamental choice in this concept is whether the user should be able to filter out the genres (main categories). By doing so the user will naturally have a much wider rage of results, which can lead to more interesting finds (playlists) and increase exlorativity. The down side is both that the user may see this as irrelevant and annoying data and that he/ she thinks the static genre-groups are located at totally wrong places.
It’s imaginable to solve this by doing things like expanding the night sky panorama view into new windows, but this may lead to an extra degree of complexity for the user.
Graphics
Details in the graphics can have a great deal when using the site. The 8-pt pixel-font and similar icons can be argued being hard to read, but was kept to this size for saving space and keeping clutter to a minimum. Also, the colour of the different genres may be too similar. The colour coding of these was though hard to differentiate because there are only so many illuminative colours stars normally have. A help function which clarifies the grouping may be one solution.
At the bottom of the page a skyline is shown. This is present for giving the user a sensation of standing in a street and looking up on the stars. This could then lower the level of abstraction a bit. If this is necessary could also be a subject of discussion.
(following are images from the web-site sketch I made in Flash)
