Today’s business model for technology tends towards one-stop shopping. What I mean by that is that all software tries to provide all-in-one solutions. While this practice leads to convenience, as the saying goes “if the only tool you have is a hammer, every problem looks like a nail.”
There is no shortage of this in the teleconferencing market. While many companies attempt to do it all, they all really only have strengths in a portion of the market. The bottom line is really a question of how much latency a platform finds acceptable.
Latency is essentially the lag or amount of time between when something is transmitted and when it is received. A few random searches suggest that an acceptable latency for voice communication is in the neighborhood of 50 milliseconds with latency becoming unacceptable when it reaches around 300 ms. To put it in musical terms, that’s a little more than an eighth note at a march tempo of 120 BPM. Also, latency is one way – by the time your friend gets your voice, and you get it back, the latency is doubled. You can test the effect of this rather simply by trying to sing a round with a friend over a cell phone. When video is added, the additional time it takes to compress (and decompress) the video and the additional amount of information that is transmitted, the target reaches “less than 300 ms” or a round trip of quarter note in a march. Although there are some ultra-low latency services (usually requiring Internet 2 with hyper-optimized networks throughout), these latencies pretty much exclude real-time musical collaboration.
Platforms must make decisions between balancing quality and latency. Higher latency means larger amounts of data can be sent and accumulated (called buffering) in order to smooth out transmission and increase quality. Video conferencing like Zoom and Skype typically work in tens or hundreds of milliseconds in order to make two-way communication effective. This is why you occasionally get dropouts, glitches or grainy pictures. On the other hand, YouTube Live streams typically have 30 seconds of buffering – allowing for relatively high quality stereo audio and high resolution pictures.
Beyond latency, features can also be important. Screen sharing and white boards may be included in teleconferencing services, but live streaming services may only have text-based messaging. Service models may include one-to-one communication (like a phone call), one-to many (presentation and audience) or many-to-many (a conference call). For every person contributing to a session, more data is required so a combination of more bandwidth, more buffering and lower quality is required.
Facebook Messenger, WhatsApp, and the like can be considered virtual meeting platforms. They work well at one-to-one and few-to-few connections but if you have more than maybe half a dozen participants, it can get unwieldy. Zoom, Go To Meeting, and Adobe Connect are geared a bit more towards conferencing: one-to-many or few-to-many. While they offer robust meeting platforms, they also have the ability to provide better quality by limiting interaction. YouTube Live, Facebook Live (to some extent), Vimeo Live, etc. are live streaming platforms that provide much higher quality but even more limited interaction so are good for recitals and concerts. Live streaming may also be an option for demonstrations. Don’t discount “old fashioned” text chat and forums. Chats can have lower latency than video – or at least the latency makes less of a difference. Both chats and forums allow for a participant to be part of multiple conversations and it is easy to review what has been said.
One more point: It’s important to realize that services like Zoom, Skype, and YouTube run on massive, virtually unlimited data farms with incredibly high-speed connections between them plus use intricate software allowing them to do complex load sharing. Local institutions, on the other hand, will likely have limited resources to devote to conferencing and will be optimized for local traffic.
In the long run, all of this boils down to using the right tool for the job. Accept higher latency/buffering if you want higher quality with less interaction and setting for lower quality when more interaction is required. With many jobs to do, it may require multiple tools. By taking some time to understand the software at your disposal (and being willing to switch between them), you can make the experience better.