Site design and hosting by
OARnet Copyright 2001 MegaConference III.
All rights reserved.

From: Hans Pfeiffenberger <hpfeiffenberger@AWI-Bremerhaven.DE> (Germany) compared to MCII: better sound and video quality, at last from those sites I saw through presenters MCU (I my memory serves me well) My ViGO could not be contacted from the MCU despite a working rehearsal on the day before ( I think I panicked Arif and everbody else by supplying wrong IP addresses, being panicked myself 15 min before my presentation. Please send my excuses) Later I learned that "nobody" could call the ViGO.

Solution: After some guesswork, I switched from Netmeeting- to builtin-Appsharing (reboot) and back to Netmeeting (reboot) and voila - it does work again (without changing any other configuration settings)!

Experience was fine - lacked audio which I will address below. Quite an experiment but the more we do these things the better we get at making large scale conferencing happen Liked the over conference - went well. Changes in schedule seemed to affect participation. Would recommend adhering to schedule closer unless technical problems dictate other. Good to be on the dist list and to see who else has the capabilities to VC.

We never were able to distribute audio from our site. We were using Polycom FX with most current upgrade. It works great with the FX MCU capabilities and great one-to-one. We could have been using an audio codec that was not supported by the MCU. I heard another conference attendee state that they had the same problem and went to another MCU, which worked better.

Publish more details about codecs, compatabilities, ... keep up the good work.

This was my first conference. I found it exciting. I found your organization and the way this worked fascinating. I really liked having the people able to talk to each other. it is too bad that everyone did not have wonderful band width and the ability to send speech clearly. Even tho speech was a bit unclear for one of the sessions. I was excited that so many people could and did participate.

The site I was at did not connect the way I thought it was going to. This was my fault for not following up with the people at my site in the right way.

I would like to be involved. I would like to have present a session using one some of the k-12 schools I am working with to present what they are doing to enhance learning while using the Illinois Century Network and The Collaboratory Project's Collaboratory. :)

PS: Thanks so much for your energy and leadership with the Megaconference III and H.323 deployment.

I thought the presentations this year were much better than last year. The presenters really prepared for their presentations, and it showed. I don't think the virtual picnic worked. With this many sites I think you need some structure. Now if we could figure out how to play picnic games that would be fun. I always like the water balloon toss. Sometimes we use eggs. NetMeeting didn't work for me. I connected several times, but I would get disconnected for no apparent reason. When I was connected I did see the PowerPoint slides at times. My ViewStation worked fine all-day long. I was on until about 3:00CDT

I don't have any great ideas for next year. If I think of any I will send them to you.

We participated as leastener in the Megaconference and I think one of the benefits from it could be a contact list with all the points participating in it, Organization, Contact Name, Phone, Address, IP, etc. This can be published in every R&E network website for example, so when some one wants to make a H.323 videoconference knows who to contact.?.

My report on the Megaconference - it was totally awesome to think we could see and hear all those sights from around the world, even from areas like Uganda.

At times the audio was poor but I know that there was a great deal of interference and we are very new at this. I have been in computers over 30 years and am always amazed at how far we have come, especially with an event such as this. Also I was delighted to see people in follow up emails eager to contact partners in other areas of the world - hopefully that will open up possibilities for sharing expertise as well as cultural understandings.

BISSNET is a consortium of 19 Independent High Schools in the Buffalo area. The schools joined together to implement technology. In June 1999 they obtained a $4 million grant from a local Foundation (Oishei - Trico). I was hired in August, 1999 and since then we have wired the schools with electrical and data wiring; installed LAN at each school; installed a Verizon WAN with minimum T1 capacity to each school; contracted with an ASP to host several applications and to manage the high speed Internet and associated equipment; and now we are installing h.323 videoconferencing with Zydacron units. We have done extensive inservices for teachers on various aspects of technology and applications.In many cases my schools are competitors and we have walked a fine line in maitaining each school as being unique. The schools are starting to realize the huge advantages of a consortium in sharing expertise and in buying power.

As a result of the Megaconference I have made contact with Kris Skurtvald from the Vancouver Schools and we hope to try an AP Physics or AP Chemistry project together. I plan to actively contact other members of the megaconfernce just to connect and to find out what they are doing. Buffalo has a very large Polish population and one of my schools and a local college are very interested in the Polish traditions - perhaps we can work with the Technolgy Institute at Lodz.

I am also working with Jim Whitlock on Regional issues of Connectivity and Videoconferencing. A group of us will be attending the CANARIE conference to learn more about dark fiber.

In conclusion - I can't wait for next year and Megaconference IV !! Keep up the good work. Please feel free to contact me if you need more information.

PS. My Mimio arrived - what a treat to win a prize on top of such a fabulous day!

Idaho State University learned several valuable things. We discovered needed contacts in Taiwan and at Stanford and we are making plans to have Benjamin at small farm networks, present to our rural sites.( I like the idea of having a "unique" presentation like his break up the more routine presentations)

The event itself also helped convince me we are on the right path moving to H.323 as bandwidth allows. I enjoyed the rollcall and feel it was valuable. It gives folks a much better appreciation for who is involved and what this technology can do. I also learned from your patience and positive attitude when silly people didn't mute mics or did other stupid things.

I have a number of perspectives that I think would be excellent for your report, provided you are interested in slightly non-technicals -- for example, we recommend a "closed circuit (members only) onscreen chat window" for our distributed presenting group, because we were unable to give our participants key notes (as to volume or camera issues) while it was on-going.

Also, we found that those of our presentors who used rear projection (as opposed to direct video input through vc units) had great success, provided that they controlled the lighting on their faces and had lower light in the room.

We found that standing figures (such as behind podiums or in front of their projection screen) were more compelling than seated speakers. We found that telephone conferences between presentors and mega technicians, well in advance of rehearsals, significantly affected how the presentation would proceed -- because presentors better understood the capabilities and the technicians better understood what the presentors had in mind. We found that use of the mega list e-mail system was an excellent community builder as we saw what was being done elsewhere and could ask for or offer help as a result.

We found that the streaming (which Mike at Livermore Berkeley arranged on our behalf for the Office of the UC President) gave those on the outside a very powerful impression of what was done which created an important additional component.

We found that key points such as, understanding how long it takes to refocus to another individual in the same room - or between parties in different locations this simple experience in real time TIMING made the difference between whether a speaker was on camera by the time the vc came to their site or not...

Liked the topics but would have liked more details. For example, when talking about going to I2, one of the K-12 examples was virtual field trips, yet it sounded as if they were gatewaying out to sites via ISDN. For us, who are new, just hearing about a standard dialing plan (vide.net) was an eye-opener (we are going to join). Even though we participated in two pre-session tests, we had problems connecting. The calls timed out as if the gatekeeper did not get our request, as opposed to during the tests when we got a call rejected when a given MCU was full. After about 90 minutes of re-dials (and trying to watch via streaming video) we did get in.

>From a personal point of view, at the end of the day I was more tired having watched the conference than I believe I would have been if I was at a live conference of the same duration.

In the discussion following the Megaconference I have seen no discussion about the quality of the signal received during the conference. We are brand new members of I2, and had high hopes that the quality of the video and audio would be quite high, and in fact it was so low as to be almost incomprehensible. We are trying to isolate the cause for this, and one obvious question is whether these problems were experienced by other members of Internet2 (and therefore non-local) or not (and therefore due to our pipeline to the gigapop in Chicago). No need to take up extensive bandwidth on this, but if others experienced severe video and audio dropouts, and just haven't mentioned it, or if we are the only ones with the problem, I would appreciate hearing about it off-list.

My own general (and simple) view is that quality is a complex function of 1) IP path quality in terms of packet loss, jitter and out-of-sequence delivery error rates, 2) endpoint and infrastructure component (MCU, GW, GK) hardware & software immunity & robustness, 3) event scale (larger -> lower quality), and 4) the variety/heterogeneity of endpoint and component manufacturers & models. With that as a backdrop, my own expectations for quality during edge-of-the-envelope large scale events like MC-III are not high. Too, large scale events always introduce human factors (lack of discipline, lack or preparation, ignorance, rudeness ...) that are all but impossible to control.

I was very careful to promote the MC-III within our region as an exercise in and demonstration of scope/scale and not one of quality. Our regional early adopters are well aware of (and motivated by) the exceptional quality they see all the time in our local regional demonstration projects and trials where the scale, component mix, paths, etc can all be more carefully managed. I'm also *very* careful to aggressively discourage early adopters who do not seem to appreciate (and are not prepared to deal with) the underlying complexities and support requirements for practical deployments today outside of controlled private-network homogeneous modest-scale contexts. While sales reps may want to have it otherwise, I don't think the technology is anywhere near the comsumerization levels of, say, telephony and people need to be aware of that to avoid both tactical and strategic planning and deployment failures.

Yes, there were quality problems at times during MC-III and the VIMM sessions. But I, for one, and our regional constituency for another were vastly impressed with both the presentation content and the event quality evident at the tearing edges of the envelope. We could safely conclude that many of the trial and deployment plans that we're working on now are clearly manageable -- provided that appropriate support staff and technology are available. I have not yet and may not have time to produce another post-mortem for our regional experiences as I did last year (which I'll be happy to forward as an example to anyone who asks directly). To Bob Dixon and all of the myriad other institutional and commercial whizzes who made the events possible, raucous huzzahs, kudos and huge thanks from Western New York for another hugely successful Megaconference. We all learned a lot from both the experience and the event content. Please don't even consider giving this activity up. We really need this type of activity to keep moving forward. Great job! Huge thanks! Please don't stop.

I would second Jim's comments on Megacon3 quality and add that one way to compare what 'we saw' vrs 'what others saw' is to review the streaming archives. Granted that streaming introduces other problems as well, but it is a way to check if your link quality was at least in the same ballpark as what the streaming site had. As I recall there were a couple of sites that appeared to have link problems that most other sites did not seem to have. Perhaps a check of the streaming archives would help ---- but you may need to get the local link fixed to be able to see the streaming archive. :-)

Stan [regretting that we were not able to participate in the streaming this year]

I agree with the two previous persons. However, saying this I also think the next MC should have a serious look at the video/audio quality issue and the over all order. Bob did a fantastic job and I don't blame him. He made something out from an almost impossible equation.

If we will ever be able to sell this to other persons than ourselves, that are already converted, we have to be able to provide better video/audio quality. I just wonder what these people thought who came as new into MCIII and saw poor video and audio from where it was difficult to judge from where it came. Perhaps we had the connection to the space station after all? This serves as a very good testbed for us and the HW/SW companies but does this help us to sell the consept to the normal doubtfull endusers? I really don't know.

We have to have a lot better tools available to handle this complicated MCU structures. Shouting out kind and less kind requests for NetMeeting users to drop off doesn't sound very professional. Before MCIII I was kindly asking about the main purpose of MCIII because it wasn't clear to me. From the point of an engineering exercise I can get the idea but from a normal enduser's point it's still not clear to me.

This was the first time that the University of Delaware participated in the Megaconf. Our overall impression has been positive, but we also did not have the expectations of high quality audio/video. We were looking at feasibility/useability and this is definitely great work going on.

There are many factors involved but most disturbing were the human factor errors (not muting mics). Even here, my colleagues forgot to mute the mic on 2 occasions, mostly because making changes to the gatekeeper to get to different room, the system restarted and it comes up with the mic unmuted. This was during I2 and not the Megaconf. Take this factor and multiply it by 200 and you can begin to see the magnitude of such "innocent" errors. I would also point out that the connection speed was very important. If you tried to connect at anything other than 384K, you would get very poor performance. We accidentally saw this during the I2 meeting, when changing phone numbers, etc. in the addressbook, we got a default speed of 768K. I agree with the others, that we were constantly checking the real stream to determine how our H.323 was performing. In general (when we weren't dealing with the unmuted mics), we saw the same performance.

Because of the cancelation of the I2-Meeting there was interest to participate using VC. But due to holiday absence we were not able to participate in MIII with our own MCU. But we kept informed our H.323 users about the events.

Even all our users are expected to understand English and our gatekeeper has configured the ViDe Zones we needed to translate every announcement of the conferences.

We needed to translate the time into GMT and we needed to find out the dialing number for our users. We don't like the idea to change the gatekeeper config at every terminal and to get into trouble at the next local conference because a part of the terminal still use the gatekeepers configured for I2 conferences.

It would be useful to find/define a format for such announcements and to get the end users used to it. (If a format is defined und generally accepted there is a little step towards a XML encoded document which could be easily used to automatically generate the specific parameters for a local user.) This conferences was a great experience for new users and an exellent occasion to test our equipment and to train our support staff.