First Monday

MusicGrid: A case study in broadband video collaboration by Hassan Masum, Martin Brooks, and John Spence


The technical requirements for widespread deployment of broadband video over the Internet are rapidly being met. But a harder challenge remains: how can video–based technologies promote collaboration and learning?

We present a case study: the MusicGrid Project. Running from 2002 to 2004 with partners in several Canadian and international locations, this modestly funded initiative ran over one hundred successful multi–site education and performance sessions. The rationale, development, and operation of the project are discussed, along with general lessons learned. We believe that our experience and the opportunities and issues identified will be useful to all those interested in large–scale, video–based collaboration projects.



Beyond broadcasting
Developing MusicGrid
Running MusicGrid
Toward effective broadband collaboration




Beyond broadcasting

InterCreative media

Until recently, video over the Internet has largely been limited to passive viewing. But as bandwidth increases, it is becoming feasible to use video as a medium of communication and collaboration.

As many have argued (Aldrich, 2004; Bowman and Willis, 2003; Gillmor, 2004; Meikle, 2002), society in general and education in particular will benefit through moving from broadcast–style media to active media that promotes learning, co–operation, and creativity. So the question then becomes, how can group interaction, learning, and collaboration be tailored to best use video? And how can video technologies be developed to promote collaboration?


Figure 1: MusicGrid teaching sessions

Figure 1: MusicGrid teaching sessions.


We’ll answer this general question through the lens of a case study: the MusicGrid project. Conducted from 2002 to 2004 with several partners [1], this collaborative venture was a pioneer in the application of broadband video technologies to music education in a multi–classroom setting. Mark Brazeau of the Ulluriaq School in Kangiqsuallujuaq (northern Canada) had this to say of the program’s impact:

"Basically, without the technology, there would be no learning. There was no music without MusicGrid. The very few musical instruments we had before, like tambourines, were destroyed when the old school was hit by an avalanche ... Now, the violin has been running strongly for a year and a half. The keyboarding program with the younger kids has started up. And in the new year we are going to start up a traditional [Inuit] music class with an instructor in Ottawa." [2]

Through our successes and setbacks, we hope to give others the benefit of our experience — and to suggest fundamental questions, goals, and guiding principles of video–based collaboration.

Why video?

Videoconferencing has made progress since the early days of fuzzy and jerky images, transmitted unreliably by expensive equipment. There are now established standards and companies, PC–based video chat tools, and a growing body of research — see for instance Finn, et al. (1997) and Greenberg (2004).

Despite advances, much work remains in developing technologies and interaction methods for broadband video. Some of the challenges are technical in nature: one–click sharing and publishing of video segments, gaze and speaker detection, and picking out interesting segments of video for indexing, search, and later use. Others are economic: affordable setup and maintenance, quality of service guarantees, and low–cost hardware, software, and bandwidth (especially for remote locations that are connected via satellite). And still others involve group dynamics: sustaining interest, multiple simultaneous group interaction, and organizing and combining contributions of subgroups.

These challenges naturally lead to the question: why video? After all, making good use of video is harder than for mediums with which we are more familiar [3]. Are there opportunities that make it worthwhile to overcome these challenges? We believe the answer is an emphatic yes.


Table 1: What can video offer?
Other places
  • Talking with those in other places: traditional videoconferencing; training, teaching.
  • When seeing matters: musical instruments, dance, arts, speech giving, medicine, motion in general.
  • Seeing other places: nature’s beauty, Web cams.
  • Diversity: interacting directly with people from far away; bringing together ethnic, religious, language, and other groups which may not otherwise get a chance to meet.
Other times
  • Asynchronous collaboration: overcoming scheduling barriers, working when it best suits you.
  • Talking with those in other times: recalling relevant, high–quality speaking and teaching from the past.
  • Conversely, capturing wisdom and achievements for future use, such as great speeches, lectures, and insightful conversations.
Other senses
  • Body language, expressions, and environmental cues can deepen emotional bonds.
  • Enriching our view of reality with information outside human sensory range: telemetry, hyperspectral sensing, processed and augmented reality, haptics.
  • A visual interface to the real world: "browsing reality," monitoring sensor networks.
  • Augmenting discussions, collaborations, and research with analysis of what’s been said, and suggestions of where to go.
Other brains
  • Having a narrow channel between groups can give different approaches a chance to develop, before cross–fertilizing.
  • Opposing parties may feel buffered and find it easier to lower hostility levels with distance.
  • Those who are shy may find interacting through a virtual environment less threatening, especially with larger groups.
  • Different cross–cuts of the same group: by discipline, by locality, by task, by time.
  • It’s easier to get critical mass together, especially for technical and unusual interests, or high skill and knowledge levels.
  • Less need for in–person meetings and travel, leading to time savings and economic and environmental benefits.


Broadband distance education and multi–site collaboration

From a teacher’s point of view, an overriding question is: How can I engage my students? Many students drop out of secondary education or university. Of those who stay the course, a fair proportion are only physically present in class — wishing perhaps for a more active experience than watching the "sage on the stage."

The social affordances of multi–site conferencing seem promising, as there is a natural excitement to interacting with students from a different region or country. However, making good use of these tools requires new roles for teachers, new teaching methodologies, and new ways of structuring student collaboration across sites to be fluid yet productive.

Even just fitting technologies into the teaching environment can be difficult:

"In the school, teachers are stressed enough as it is. They don’t want to have to deal with broken-down computers and things that don’t work ... they have been burned before when they have tried to deal with technology and it failed them. They are not going to touch it anymore. Who can blame them? If they still feel that computers are a waste of time because they always break down and they are not working and they have to call someone to fix them, then they are not going to deal with them. They’re not going to waste time."

Music education provides particular challenges. Audio issues are far more difficult than for typical videoconferencing, due to exacting requirements such as echo minimization, numerous simultaneous sound sources, and greater fidelity requirements — especially since the goal is a true collaborative and teaching opportunity rather than passive observation. And musical training usually requires frequent demonstration and feedback, as the students play a segment and the teacher suggests improvements or demonstrates a variation.

Students are enthusiastic once they see the relevance of a project, making a good pilot group for collaboration in society at large. Thus, we believe the MusicGrid project has lessons of a general nature for understanding how effective collaboration can happen through a less than perfect channel.

The human dimension is another key driver for development of broadband video education and collaboration. In Canada, as in many circumpolar nations, remote communities are located hundreds or even thousands of kilometers from the nearest sizable population center. Off the grid for communications, they risk being "off the map" for opportunities as well — a situation replicated in smaller communities around the world.

"I find that it has opened up tremendous opportunities to all of our students that we have seen here because every culture they are linked to, they find out something unique. To be speaking to students from the Inuit culture the way they are, presented through MusicGrid, has been tremendous. It opens up the world to these students. We live on an island, and it tends to be isolated as islands tend to be, and you don’t have access like you do in Toronto. So just getting into your car and listening to a class here and there or wherever, is so important. MusicGrid, then, has opened up the world to these kids and you can really tell it has had a tremendous impact."

Using technology to assist underserved communities promises to enhance opportunities for schooling, continuing education, cultural opportunities, trade, and intellectual life. The experience of these communities can be applied worldwide. As the great shift from rural to urban living continues, creating alternatives to moving to large urban areas can make a more distributed lifestyle and human geography viable.



Developing MusicGrid

How it came about

In 2000, Canada’s National Research Council (NRC) had recently initiated its Broadband Visual Communication research program, which was working with Pinchas Zukerman (Music Director of the National Arts Center of Canada) on violin lessons through videoconference. The Communications Research Center of Canada (CRC) had begun the VirtualClassroom Program in 1996, which connected schools at different locations to work together. NRC and CRC joined forces with the LearnCanada program in 2000, which lasted until 2002. Sessions during VirtualClassroom and LearnCanada covered land mines, the Holocaust, water as a scarce resource, Mars, structures for space, and other topics.

With schools and professional musicians already involved and enthusiastic about pushing the limits of various tools, it was natural to continue the collaboration with a new combined program in 2002: MusicGrid. This new program’s primary objectives were to enable, expand and enrich Canadian music education programs in urban, rural and remote communities, to collaborate with international partners, and to contribute to a future where modest investment gives ubiquitous access to global learning communities.

Along with NRC and CRC, partners in MusicGrid included STEM–Net, CANARIE, and a number of school boards, universities, and music organizations. The NRC and CRC labs involved are located in Ottawa, and both labs were interested in exploring the ways in which broadband–enhanced learning could increase motivation, engagement, and collaborative outcomes.

MusicGrid was a pioneer in broadband music e–learning. Deliverables included establishment of a broadband music education community of unprecedented size and diversity, research outcomes on barriers to scaling, and outreach programs. Students and teachers, cultural institutions, industry, universities and international partners shared knowledge, best practices, culture, and passion for music using broadband visual communication tools, including videoconference and video servers.

During the course of the project, multiple sites were connected, in Canada, the U.S., and Europe. MusicGrid was a very active broadband community, with over 150 sessions held — each typically 1.5 hours in length, and involving anywhere from two to eight sites.

"... the big thing is that it has given our students the opportunity to meet other students in other schools. And, as well as, I guess the big thing is meeting other teachers and mentors and to be given the opportunity of gaining expertise about musical performance ... just giving them the opportunity to network and to perhaps learn more than what they would get within the confines of this classroom ... I think what the technology has done is broken the barriers down and just given everybody the opportunity to learn from a wide variety of different sources."

Technical setup

For satisfactory broadband video, commodity high–speed Internet (e.g., cable and DSL) is insufficient — high bandwidth and guaranteed end–to–end availability are needed. Participating sites had bidirectional dedicated bandwidth ranging from a low of 444 Kb/sec (for satellite) up to Gigabit speeds [4]. Sites were linked over high–speed research networks like Canada’s CA*net4, citywide networks for public schools, satellites, and international broadband networks. These dedicated networks give a foretaste of the power that should be available on the public Internet in a few years.

Once bandwidth is available, videoconferencing equipment is required. We used two solutions. First, Polycom units combined the necessary software with hardware, such as integrated cameras with pan–zoom–tilt capabilities. The Polycom–based H.323 solution allowed a maximum of four simultaneous connections, a limitation partially compensated for by the status of H.323 as an open videoconferencing standard [5].

Linux–based PC machines connected to external hardware, along with Isabel (a proprietary application), were a second videoconferencing solution [6]. Isabel allowed many more simultaneous connections, and included software to script interactions and display ancillary information like slides and chat windows. We found that buying standardized Linux computers for each site made setup, maintenance, and remote monitoring much easier.

"At the start, you need all kinds of support, but once you can figure it out, and have the interest in doing it, it becomes relatively easy."

"I had to become somewhat of an expert, to learn the technology, because there were times when our technician would not be in town and I would have to run the video conference by myself. So, the technology, I took an interest in it. I learned how to use the technology and it was fun. I enjoyed it. It wasn’t that difficult."

Videoconferencing units had display screens of varying sizes, ranging from monitors to projection TVs to large plasma screens. Just as with watching a movie, a larger screen size can change the videoconferencing experience for the better, especially with larger groups. Audio hardware included mixers, speakers, and both omnidirectional and directional microphones; the sound setup required some care to avoid echo and ensure high fidelity, especially when large music groups were playing.

Finally, there was also extra equipment at the CRC and NRC laboratories, to manage and store video streams. Asynchronous visual communication was addressed through development and field testing of the PrivateVideo server [7]. The goal of this video storage was to allow review of classes and performances, and reflective criticism of learning and teaching styles (Balachandran, et al., 2000).

All this sounds somewhat complex, and it is at first. But after some experience, it became easy to set up sites fast, both in technical and human resources aspects. Near the end of the project, sites could be actively participating and integrating video into their curriculum within days of delivery of hardware. Both teachers and students became comfortable with the tools — indeed, students often enjoyed the technology–learning aspect.

"The teachers really have moved along with the technology. At first, they were pretty shaky, but now it seems like they can handle pretty much anything."

"I have the students who are very helpful when I really run into trouble, but this doesn’t happen too often. And, you know, it’s been really good for these students who are helping me too — getting exposure to the technology like this. They set all of this up for me ... ."

"We’ve got a fair amount of technology in that room and we’ve seen a couple of students that are not necessarily involved with music at all become quite proficient with the computers, fixing and diagnosing problems as they arise. Students get an opportunity to learn a lot about technology ... It’s not even as if the students’ comfort with the technology was planned as part of the initial proposal or anything. It’s really just been a positive effect, you could say."

Social setup

The core team at NRC and CRC worked very closely together. Always–on connections between the two labs — located 30 km apart — meant that team members were constantly in touch.

But there was another essential element, especially important as the number of active sites increased: enthusiastic and trusted counterparts at each site. For effective broadband collaboration, technical resources are not enough. Making productive use of these tools — especially for large group collaboration — requires a degree of social intelligence, and mutual organizational and personal trust.

MusicGrid was organized into two parallel streams: operations and research. Operations included scheduled weekly instructional sessions, always–on classroom–to–classroom videoconference connectivity, and numerous special sessions. Research focused on development and field testing, and on study of user behavior around asynchronous mentoring tools. We tried to engage real learners in authentic use of prototype applications, in order to discover technology requirements and obtain feedback on technology effectiveness.

We favored always–on connections, for several reasons. First, setup time was minimized, removing one hurdle to tool usage. Second, one could look over and get a sense of what other teachers, classes, and project team members were up to. And finally, since no effort was required to set up a connection, interaction time was maximized and many impromptu conversations took place, enhancing learning and building relationships:

"Apart from the formal things happening in class where there is a kind of structure, I can go down to the music room anytime during lunch and there will be a bunch of kids there who may have dialed up to another school. They’re jamming with the musical instruments, violins or whatever ... So while the formal stuff is great, the informal stuff the kids can just do — just as if they are meeting a friend face to face — [...] just shows how connected they are with the technology."

The time required for new groups to lose their shyness, open up, and collaborate effectively was generally quite short. There were noticeable variations with students from remote areas, but an existing community of good teachers, friendly students, and appreciation for their achievements helped them to gain confidence.

Privacy was generally not a concern of participants: video equipment was located in a single room at each site, partners were well–known to each other, and videos were not distributed outside of the partner network. If and when these restrictions are loosened as broadband video migrates to common usage, resolving privacy concerns will require careful thought and tool support.



Running MusicGrid


The MusicGrid music education community was centered around eight Canadian schools, including an elementary/high school in Kangiqsualujjuaq (northern Québec), a high school in Iqaluit (Nunavut), a middle school in Gander (Newfoundland), a high school in St. John’s (Newfoundland), a middle school in Buckingham (Québec), and an elementary/middle school and two high schools in Ottawa (Ontario). All schools were connected to the Canadian high–speed network, CA*net4.

The remote Inuit community of Kangiqsualujjuaq had no music program prior to MusicGrid. MusicGrid provided a group of ten fifth– and sixth–grade students weekly violin lessons for one and a half school years, and another group of eight children with weekly keyboard lessons for one school year:

"... the violin program and the success of the violin program have intrigued a lot of teachers in the school and a lot of people in the community. When we were asking for participants in the keyboarding program, we had an overwhelming amount of applicants. We turned away like 90 percent of all the kids who wanted to participate in the program because there were only eight spots available. I think there’s a lot of enthusiasm for what’s going on."

Later, a teacher in Ottawa was found to teach a group of six high school girls traditional throat singing and drum dancing. This knowledge had been lost from the community, but is now being rediscovered from experts elsewhere who have kept alive the ways of traditional music. Lessons are still ongoing, and the girls have performed via broadband video for audiences in North America, Europe, and Hawaii.


Figure 2: Trumpet coaching via videoconference

Figure 2: Trumpet coaching via videoconference.


The participating schools in the rural communities of Gander and Buckingham have music programs, but their community size necessarily limits the programs’ breadth. MusicGrid expanded their programs: Doug ("Pace") Sturdevant, former principal trumpet player with the National Arts Center Orchestra, worked weekly with individuals and groups of brass and wind players for one and a half school years. These sessions were of great value to the teachers, as well as the students:

"Every Thursday, Pace, from Ottawa, gives a trumpet lesson to my students. This has a very positive impact because the students are able to see the difference in how he plays versus how they play themselves. He gives them ideas about what they can do better or differently. Then, after this, when they play, they are able to think about what they’ve learned and try and apply it to their own music within the group."

"I see Pace almost every week. Students play for him, and he gives them feedback and provides tremendous insight into music that the students are playing, whether the music is German, French, Italian or whatever. So, he not only talks about playing, but also the aesthetics of what your responsibility is as a musician and the piece of music you choose to play."

The participating schools in the urban communities of St. John’s and Ottawa have strong music programs. MusicGrid enriched these programs, providing advanced mentoring, cross–cultural musical exchange, and opportunities for accomplished students to assist with broadband teaching.

International sessions exposed the MusicGrid community to Chinese music from Beijing and Hong Kong. David Cooper, principal of Holy Heart of Mary Regional High School in St. John’s (the capital city of the province of Newfoundland and Labrador) reminisced about one session:

"I think one of the real highlights was using the MusicGrid technology when we had a session with China and the previous principal, Bob Pitman, addressed the graduates at Holy Heart, but also he had the students in China doing a program that allows them to not only graduate with a Chinese diploma, but also a Newfoundland high school diploma. And so these kids watched our graduation and Bob did the address on the big screen in the auditorium, and I’ve got to say I was really impressed. There were close to 1,000 people involved, doing it live, and for me that was one of the highlights ... ."

Another session included pop and avant garde music from Geneva. Student conductors in St. John’s and Potsdam, N.Y., participated together in a masterclass for conducting. Violin students in Finland participated in trials of new asynchronous mentoring tools with the NAC. And a piano student in Finland received coaching from Ottawa on tone production, using two cross–connected Yamaha Disklavier pianos, where playing on one piano caused the other to also play.


As with every new technology, a number of difficulties had to be overcome. Scheduling was one difficulty:

"Generally, the time of the network sessions was somewhat awkward this year. Our music classes are arranged so that one day is ‘performance’ based and the next is ‘theoretical’ based. Most of our network sessions fell on days which were not performance–based days. It was difficult to co–ordinate the use of students for performance–based activities. Most of the students had to be pulled out of other classes — either other music classes or from academic classes. The kids usually didn’t mind but some of the other staff did. Sometimes the session times (which had been agreed upon earlier) went way overtime."

Another difficulty, especially in the initial stages, was finding or making time to use the technology. A potential new participant must learn the tools, figure out how to use them productively in their environment, and find others elsewhere with whom to work and collaborate. So getting involved could take a while:

"It’s a slow process for the teachers too. They’re used to doing their regular routines and I have to be patient with them and allow them to come on board when they’re ready. This has to be done even more so for the other schools. I can’t push other schools to collaborate with our schools, because most of the other schools don’t have a resource person like me, so I usually have to deal with the teacher directly and they are very busy people. So, when we are doing some work, we have to pace ourselves so that it is manageable for everybody involved."

"It’s fine to have the technology, but, as well, you need the time to be able to devote to organize and facilitate any of the sessions. I’m finding, I guess, in terms of the other teachers involved, I’m half time dedicated to this, but I’m finding, to be honest with you, that this is even evolving beyond something you can do with a half–time position."

Most participants managed to get involved in an active and ongoing way, and take advantage of the technology’s opportunities without investing too much preparation time. Indeed, one difficulty near the end of the project was the unwillingness of participants to stop:

"[Y]ou develop the momentum — and it’s not just me, personally — I think everyone who has been a partner in this has learned an incredible amount, including the teachers, technicians, as well as the administrators, and anyone who has been involved with the project. The growth that has occurred in the past three months, even, I mean the momentum has been incredible."

"I think one of the priorities as we work towards the end of the project is finding some way of maintaining or continuing it because of what has been accomplished ... I look at the closure of this particular project wondering, my God, you’ve developed all this incredible momentum, you learn so much about the technology — for it to end, God, you wonder what you’re going to do after that, you know."


A variety of positive learning outcomes took place during MusicGrid, which are best conveyed through the voices of those who were directly involved:

"Often what I will do is sit down and watch a rehearsal that is taking place at, say, the School for the Arts in Ottawa, and be able to, just by watching the rehearsal, pick up certain ideas ... about how I might go about ... teaching myself. I think that, while looking at mentors that specialize in specific areas, such as brass — like, for instance, my thing is woodwind — you pick up on certain things that you probably wouldn’t know yourself. And that’s great. So, from a personal perspective, I pick up an awful lot of information because we have access to that kind of mentorship without having to go and hire someone or bring someone in. You know, that person just happens to be there and through the use of the technology, it’s very easy to gain access to that. I’d have to say that, as a teacher, it has a pretty profound impact on my own teaching."

"As a music educator, my overall goal would be to improve the students’ overall musical skills, their general appreciation for music, and also, as a teacher, to help me improve my skills as an instructor so I can better communicate to them how to do some of this and I certainly see this project as a fantastic way of doing that. Students will get to see what’s happening in other parts of the province, country and other parts of the world. Are they on par?"

"I am inspired that people can actually see through a TV screen and correct problems and anticipate certain outcomes with respect to the performance of the kids, and as a former volunteer myself, it’s such an incredible resource for students. To have that feedback, but also to have somebody who is knowledgeable in so many areas brings the kids that much more to a higher level."

Often, just knowing that there is an audience — that your peers or a wider community will be looking at your accomplishments — is enough to boost motivation, interest, and performance:

"... [I]f they know I’m videotaping, like the other day when I videotaped, they were like trying to play everything at the correct time ... trying to look really professional and things like this, and I must say, when they played the other day they were pretty good. And so I think that’ll be really good for them, you know performance–wise because they’ll ... they’ll have that little bit of a boost there."

"One parent said: ‘My child is always practicing for Thursday.’ It’s really great. So, all of a sudden, this Thursday event that we do is really motivating the students to practice. We have a session with one of the trumpet players from the National Arts Center coming up, and the students are practicing for him, ... and so I mean, that makes it even better for me because they’re playing some of the music that they’re learning from me and they’re practicing for both of us, really. It’s a wonderful program all the way around. ... [Y]ou’re enthusiastic about getting in there and learning what’s going on."

Both teachers and students can find motivation through interaction with a wider community:

"This has also been really positive for the teacher involved. I really notice how this has just given her this extra bit of motivation — not that she wasn’t motivated before, but sometimes we all need that extra spark to get us interested again after so long and this project has really seemed to do that."

"Some students have been helping me with the Kangi violin classes on Wednesdays. They are really benefiting from this exposure. It is making them more self–assured and they are getting hands–on experience with teaching younger students."

The interaction between different music groups provides a chance to self–evaluate, and to learn from and appreciate high performance levels of other students:

"We’ve actually listened to other kids perform and I’ve had a chance to evaluate where we are in comparison with where other people are."

"They’re sharing musical involvement and musical interpretation and they’re simply making comparisons of where they are and they’re allowed to listen to each other without fear of reprisal or anything else."

"Canterbury’s fine arts school, for instance, has musicians of such high caliber, and my kids are able to see this. They’re learning so much from them, and they’re only in grades seven, eight, and nine ... I think just the exposure to listen to kids, say, from Canterbury school, which is a school very similar to ours, I think is very much eye opening. just the other day we had a kid from Canterbury playing the flute and I had a whole bunch of flute players down here watching her... she was quite exceptional ... [It was excellent] to have the ability to hear fine players and to work with them."


Figure 3: MusicGrid performance sessions

Figure 3: MusicGrid performance sessions.


And the chance to experience cultural diversity firsthand can be an effective way to develop understanding and empathy for those from elsewhere — and to simply enjoy playing together:

"... just the opportunity to play and perform and to be exposed to each other’s culture, to gain an understanding of what it is like to live in Iqaluit as opposed to what it’s like living in St. John’s, Newfoundland. I think all these things are really important."

"There was one time where there were four screens, a choir, a fiddle player and a bass player. But what struck me was that the fiddle player was being accompanied by a bass player in another region of Canada, and students were jigging. It was amazing to me how the virtual ensemble was going on and covering mass regions in Canada. And really there wasn’t too much of a delay. It was incredible."



Toward effective broadband collaboration

MusicGrid and similar efforts elsewhere in the world have prototyped broadband video collaboration. But these are just starting points. In this final section, we first summarize lessons learned from MusicGrid. We then look at tool development opportunities, and at the long–term potential of these technologies as enablers of large–scale collaboration on the Internet.

Lessons learned

MusicGrid encompassed three effective principles for scaling of broadband visual communication for learning:

  1. On–demand access: At most MusicGrid schools, the videoconference equipment was permanently installed in the music room. With at least 10 Mb access via CA*net4, schools could initiate sessions or store and retrieve video segments from the library at will.
  2. Very frequent use: Every school was involved in at least one broadband music session per week. Often schools were involved in one or more sessions per day, facilitating strong community ties. Additional point–to–point and multipoint visual communication for discussion, planning and problem solving was initiated multiple times each day.
  3. Diverse community: The MusicGrid community spanned differences in musical genre, musical expertise, grade level, culture and language for both students and teachers. Grade levels ranged from grades K–12, plus university. Genres included classical strings, concert band, folk, jazz, avant garde, Inuit throat singing and drum dancing, African drumming, and Caribbean steel drums. Languages included French, English and Inuktitut.

The three principles of on–demand access, very frequent use, and diverse communities enabled two phenomena that are essential for large–scale broadband learning:

Easy inclusion of new community members: Two schools joined MusicGrid in the last three months of the project. They were fully functional and contributing within several days after joining. In short order, they were leading broadband sessions with mentors present at their schools, demonstrating the ability to rapidly become engaged in global learning communities for a modest investment in bandwidth and videoconferencing tools. Being constantly connected visually, they quickly became familiar with other partners’ musical activities, and could get immediate response to technical questions. The MusicGrid community diversity provided each partner with multiple opportunities to both provide and receive knowledge.

Low session overhead: Broadband access in the music room and frequent use of broadband visual communication drastically reduced the need for session–specific technical preparation. The resulting familiarity between partners greatly reduced pedagogical planning overhead. Consequently, only a low level of effort was required for planning and execution of high–quality broadband learning sessions, creating a virtuous cycle whereby very many sessions were carried out with increasingly lower overhead. Examples of low–overhead sessions included a five–school "Morning with Mozart" session, featuring Pinchas Zukerman and National Librarian Roch Carrier, and a six–school Christmas session that was shown on a national Canadian TV channel. With the exception of student musical preparation, these sessions each required about one person–day of session design and pre–production work.

Other innovative achievements in MusicGrid included:

The ability for new participants to come up to speed quickly was in large part due to the mentoring and community wisdom provided by previous participants. There are many examples in organizations, newsgroups, Internet communities, and communities of practice where living communities act as repositories of knowhow that is taught informally and efficiently to new participants (Rheingold, 2000). In MusicGrid, this aspect was clearly seen to be present in multisite collaborations linked through broadband video.

Tool development opportunities

A basic requirement for synchronous broadband video is a broadband connection. MusicGrid was designed to make use of bidirectional 10 Mb from the workstation to the backbone for all fiber–connected schools, and 1.5 Mb for satellite–connected schools. MusicGrid would therefore not work on today’s commodity high–speed Internet, especially since bandwidths are typically lower upstream than downstream. Using multiple simultaneous connections or emerging technologies like HD–quality videoconferencing will require far more bandwidth.

Scaling up network management is a challenge, especially for non–technical users. By its nature, broadband video places high demands on the network architecture: guaranteed quality of service and low latency are essential for a satisfactory experience. Session management software must also be both easy to use and powerful — a particularly tough pair of criteria to meet when many groups are interacting.

Because cameras are usually not placed inside display screens, most videoconferencing has an "eye–shifting" effect: the people on the other end appear not to be looking directly at you (since they are looking at your image on their screen, instead of at their camera). Though most users can adapt to this initially disconcerting effect, a solution would be preferable. We also found that the size of the display makes a significant difference, as does having high–quality audio hardware.

Expert mentors are particularly useful for advanced students, but can be hard to schedule. "Video e–mail" that is as simple as text e–mail to use would let students record presentations or performances, with the mentor giving feedback when time permits.

Asynchronous video use could allow an expanded range of applications beyond traditional synchronous videoconferencing, just as e–mail, mailing lists, newsgroups, and blogs provide many more kinds of interaction than synchronous text chat (Bargeron, et al., 2002). In the context of educational initiatives, development of a library of lessons and best practices would be useful for future students. Indeed, even recording failures could be useful, especially if they were systematically analyzed to develop better learning strategies for the future. Developing an open video library of "teachable moments," available under a Creative Commons ( license, could be one step toward a global learning object repository (Downes, 2004).

MusicGrid studied student use of video for reviewing musical coaching sessions, as part of the National Arts Center’s Young Artist Programme. The study contrasted group versus individual viewing, and determined that group viewing focused on ensemble playing whereas individual viewing focused on verbal instructions — an observation which could lead to improved metadata and user features in future systems.

"This project has moved beyond the technology in that it has already shown that the potential exists and it actually resulted in some very significant learning. My feeling is that we are moving in the right direction — more as technology fades into the background. Because the technology is by no means the end, it’s only a means by which to teach; we need to do what we can to make the technology completely disappear."

"Video blogging" started in 2004, as a combination of traditional weblogs with supporting video segments. Some challenges in moving from text to video include: making video–based production and weblogging tools as easy to use as the text–based equivalents; securing and storing permissions for video segments, especially if minors are involved; adding metadata and ratings to video segments to improve searchability, quick scanning, and the identification of particularly important portions of a video. Considering the effect that weblogs are having in increasing news and discussion channels, enabling news and discussion around video segments could broaden private, educational, and public discourse.

We have seen that the tremendous increase in textual material available online necessitated ever–more sophisticated search technologies. The same will happen with video; as we move from a situation of relatively few producers to millions of producers, we will need effective search, filtering, and recommendation tools. Producers of video can help by including metadata to easily mark video segments of interest, so later viewers and learners can zero in on key segments. One possibility is to automatically run video through speech–to–text software, to create an approximate commentary track.

For effective team collaboration, video is just one part of the toolkit. Integration with other tools like chat, Wiki, weblogs, wireless feedback, synchronous text editing, groupware, and so forth is necessary. Based on MusicGrid and subsequent activities, we feel that the best way to do this is, simply, by doing: identify important problems, gather a group that’s keen on solving them, provide tools and background advice, and watch what happens. Then improve and iterate.

A human–centered interface and tool development process is essential for broad deployment — one which considers the true needs and concerns of users (Shneiderman and Plaisant, 2004). Outside of public spaces, a logical approach to privacy is that video should be private, unless participants explicitly give permission otherwise. Roughly speaking, there are three natural access levels, which can be viewed as concentric spheres:

  1. Personal — only you can see.
  2. Community — only your community can see.
  3. Public — the whole world can see.

A fourth natural possibility is to explicitly give permission to specific individuals, for each separate video segment. This is more labor–intensive, but allows more fine–grained control.

In all cases, implementing privacy requires a thoughtful combination of software development and policies. On the production end, video segments should include their producer’s desired privacy or access levels. Securing these preferences becomes nontrivial with larger groups, or when a participant has a relatively small role. With regard to enforcing desired privacy, the easier case is where all members of the community will respect the stated desires of other participants. This is a reasonable assumption for cohesive communities with repeated interactions. The harder case is when videos are released to the wider world, at which point the past decade’s experience with file sharing suggests that control is effectively lost. For this reason, we expect that people will be relatively careful about what they choose to release to the wider world, and therefore that it should become customary to ask the permission of everyone on a non–public video segment before moving it into the public sphere.

Large–scale broadband video collaboration, and the future of the Internet

Broadband video collaboration is fertile ground for research. But given the large numbers of studies trying to show that videoconferencing is "good" or "bad" for educational quality, and the difficulty of separating out causal variables, we agree with the point of view put forth in (Greenberg, 2004):

"We do not need further research that attempts to prove the pedagogical worth of videoconferencing for education — that is, that videoconferencing improves or degrades the educational experience. What we need is research that achieves the following objectives:

A similar point holds for broadband video collaboration in general. The basic questions to ask are: What additional benefits can it provide? Under which conditions? And for how much time and money?

Rural and remote communities can reap obvious benefits from this technology. Indeed, the educational and training opportunities, and the harder–to–quantify "feeling of being in touch with the world," could be important policy tools for slowing the brain drain from rural to urban areas that affects so much of the world. This could be important in developing world communities. India, for example, is rolling out cheap broadband in many areas. But it is equally applicable to rural and remote areas in much of the world, and in sparsely populated regions such as the Arctic circumpolar area.

To the extent that video can help reduce travel for face–to–face meetings, it offers cost and environmental advantages. Even more gain will come from those meetings and collaborations that never would have happened in the absence of the technology. E–mail has made rapid correspondence with friends and colleagues from around the world an everyday reality — a futuristic dream for most people only fifteen years ago. Broadband video holds the promise of doing the same for a wider range of activities — of promoting collaborative teaching, learning, and problem–solving across boundaries of many kinds.

One of the earliest worldwide examples of video collaboration was Megaconference (at, which used video technology to showcase a variety of applications from participants around the world:

"Each year, the Megaconference occurs as a result of a tremendous amount of volunteer effort and good will, with the goal of connecting people together everywhere on Earth where someone chooses to participate, in order to further the use of videoconferencing in education and research and to advance the state of the art in videoconferencing technology."

As the technological challenges are overcome, as they surely will be, the bigger issue will remain. Once anywhere–to–anywhere video is as ubiquitous as text is now, how can we best make use of it? How can it be a participatory medium? How can that participation be scaled up to larger groups, and to larger numbers of groups interacting?

We would like to suggest three general themes that we feel confident will be important, out of the many that will doubtless be developed. First, videoblogging (in combination with consumer–level video tools and P2P distribution mechanisms; see has the potential to become a significant news channel in the near term, challenging television for certain niches just as blogging is challenging newspapers. The collaborative development, dissemination, and filtering of video–based stories will open up many public conversations which do not now take place.

Second, dealing with a diversity of political, cultural, religious, and intellectual viewpoints can only grow more important as the world grows more interdependent. Travel and direct constructive contact with those from elsewhere can broaden the mind and increase understanding and co–operation, but these opportunities are not part of most people’s everyday experience. However, they can be through online channels. Once those channels grow rich enough, any who wish will be able to understand their global neighbors better, and to work with them on problems of mutual concern and interest. Unlike text where all interaction flows through the same set of symbols, seeing the greater diversity of those we are interacting with will require more give and take — but may lead to more true understanding and tolerance.

Third, there are many big problems that are hard for any individual or even traditional group to solve alone. Scaling up video–mediated collaboration will be one part of the toolkit that makes the Internet a true "meeting of minds," letting us discuss, debate, and collaborate with people worldwide. Tough, civilization–critical issues demand the most intense possible search for solutions, and those who go searching will need to partner with the best colleagues — no matter where they are [8].

"It’s still amazing to me that we can all meet, despite our geographical distances, and chat about what we think and our experiences. It’s easy to feel the connection between the sites and to see how we’ve all been working together to implement the technology."

"Anytime you have students from one part of the world talking with others in a whole other place, good things are going to happen, I think. How can they not? Music is a great place to start, but as I saw with the space event, it’s not really limited to that. I think there’s a lot more we could do." End of article


About the authors

Hassan Masum is a postdoctoral researcher affiliated with the National Research Council of Canada, the Communications Research Center of Canada, and Carleton University.

Martin Brooks leads the Broadband Visual Communication Research Program at the National Research Council of Canada’s Institute for Information Technology, in Ottawa. Dr. Brooks was Research Manager in MusicGrid.
E–mail: martin [dot] brooks [at] nrc [dot] ca.

John Spence is Program Manager for Applications Technologies at the Communications Research Centre Canada in Ottawa. He was the Project Manager of Operations for the MusicGrid Project.
E–mail: john [dot] spence [at] crc [dot] ca.



We would like to thank all those who made our project a success — the partners in Note 1, the participants listed below, and all the other students, teachers, and helpers whose commitment and enthusiasm made MusicGrid such a worthwhile learning experience. We also acknowledge the tools developed by the Open Source community which were essential to the technical end of the project.


Management and Co–ordination Team:

Technical Team:

Research Team:

External Evaluator:

Expert Mentors:



1. MusicGrid Partners included CRC, NRC, STEM–Net, Canarie, the University of Ottawa, Memorial University, Ottawa Carleton District School Board, La Commission Scolaire au Coeur–Des–Vallees, Avalon East School District, Ecole Ulluriaq (Kangiqsualujjuaq), Lewisporte–Gander District School Board, the National Arts Centre of Canada, Telesat, and the City of Iqaluit. There were also a large number of associate partners, many from outside Canada.

2. This and later quotes are from Murphy (2004), unless otherwise noted.

3. See Majchrzak and Malhotra (2003), in which the authors of a guidebook to managing dispersed teams suggest that videoconferencing is best avoided in favor of other tools. Objections of team leaders included both technical issues (expense, bandwidth, availability) and human interaction issues (slow and disruptive, more difficult to scale up to many participants than audioconferencing or groupware, looking at people rather than shared documents, needing to use special meeting rooms). Sonnenwald, et al., (2001) discuss some practical troubles with video collaboration in an academic environment, and the protocols and solutions which arose in response.

4. All schools were connected to CA*net4, either directly or via the Telesat satellite network. Bandwidth from the classroom to the backbone varied from 444 kb (Kangiqsualujjuaq, via satellite), to 1 Mb (Iqaluit, via satellite), 6 Mb (Buckingham), 10 Mb (Gander), 100 Mb (St. John's), and Gb (Ottawa). Broadband technology included videoconference and the PrivateVideo server. Videoconference included H323 (384 kb–2 Mb bandwidth used) and Isabel (2 Mb–10 Mb bandwidth used). Note that satellite bandwidth, although relatively modest as compared to other bandwidth connections, is by far the most expensive to operate.

5. See the Videoconferencing Cookbook at for more details on standards and tools.

6. See Quemada, et al. (2004) and the Isabel home page at

7. The PrivateVideo server is a platform for seamless storage and retrieval of video segments. Video off the server streamed between 300 kb–5 Mb, and at project close the server contained 548 videos. Teachers, students and the research and education community made 2463 distinct video viewings; total data transferred from the video server was 17.9 Gb.

8. See, for instance, the Science of Collaboratories Web page (at — a portal for computationally enabled and geographically dispersed scientific collaborations. The inherent multidisciplinarity and sheer difficulty of scientific challenges makes the "collaboratory" concept a natural and necessary one for the scientific community, as espoused in a recent NSF report (Atkins, 2003).



Clark Aldrich, 2004. Simulations and the future of learning: An innovative (and perhaps revolutionary) approach to e–learning. San Francisco, Calif.: Pfeiffer.

Daniel E. Atkins (chair), 2003. Revolutionizing science and engineering through cyberinfrastructure: Report of the National Science Foundation Blue–Ribbon Advisory Panel on Cyberinfrastructure. Arlington, Va.: National Science Foundation.

Anand Balachandran, J.J. Cadiz, Jonathan Grudin, Anoop Gupta, Gavin Jancke, and Elizabeth Sanocki, 2000. "Distance learning through distributed collaborative video viewing," Microsoft Research Technical Report MSR–TR–2000–42, at, accessed 24 April 2005.

David Bargeron, Jonathan Grudin, Anoop Gupta, Elizabeth Sanocki, Francis Li, and Scott LeeTiernan, 2002. "Asynchronous collaboration around multimedia applied to on–demand education," Journal of Management Information Systems, volume 18, number 4, pp. 117–145.

Shayne Bowman and Chris Willis, 2003. "We media: How audiences are shaping the future of news and information," at, accessed 15 March 2005.

Stephen Downes, 2004. "The learning marketplace: Meaning, metadata and content syndication in the learning object economy," at, accessed 15 March 2005.

Kathleen Finn, Abigail Sellen, and Sylvia Wilbur (editors), 1997. Video–mediated communication. Mahwah, N.J.: L. Erlbaum Associates.

Dan Gillmor, 2004. We the media: Grassroots journalism by the people, for the people. Sebastopol, Calif.: O’Reilly.

Alan Greenberg, 2004. "Navigating the sea of research on videoconferencing–based distance education," Wainhouse Research report, at, accessed 15 March 2005.

Ann Majchrzak and Arvind Malhotra, 2003. "Deploying far–flung teams: A guidebook for managers," at, accessed 24 April 2005.

Graham Meikle, 2002. Future active: Media activism and the Internet. New York: Routledge.

Elizabeth Murphy, 2004. "Final summative evaluation report of the MusicGrid project," Research Report, Memorial University of Newfoundland, for CANARIE, Ottawa.

Juan Quemada, Tomás de Miguel, Santiago Pavon, Gabriel Huecas, Tomas Robles, Joaquín Salvachúa, Maria José Perea, Eduardo Moro, Diego Andreas Acosta Ortiz, Juan Aantonio Fernandez, Fernando Escribano, Alvaro Diaz, Juan Luis Fernández Rodriguez, Juan Antonio Sanchez, Javier Sedano, and Miguel Gomez, 2004. "Isabel: An application for real time audience interconnection over the Internet," TERENA (Trans–European Research and Education Networking Association) Networking Conference (7–10 June 2004, Rhodes, Greece), at, accessed 24 April 2005.

Howard Rheingold, 2000. The virtual community: Homesteading on the electronic frontier. Revised edition. Cambridge, Mass.: MIT Press.

Ben Shneiderman and Catherine Plaisant, 2004. Designing the user interface: Strategies for effective human–computer interaction. Fourth edition. Addison Wesley.

Diane H. Sonnenwald, Reto Bolliger, Paul Solomon, Noriko Hara, and Tom Cox, 2001. "Collaboration in the large: Using video conferencing to facilitate large group interaction." University of North Carolina at Chapel Hill, National Science Foundation Science and Technology Center for Environmentally Responsible Solvents and Processes, White Paper, at, accessed 15 March 2005.

Editorial history

Paper received 6 April 2005; accepted 20 April 2005.

Contents Index

alt="Creative Commons License" border="0" src="" This work is licensed under a Creative Commons License.

MusicGrid: A case study in broadband video collaboration by Hassan Masum, Martin Brooks, and John Spence
First Monday, volume 10, number 5 (May 2005),