General Questions

Q: When will musicians be able to sing together again, in person?

As we learned from a choir in Mt Vernon, socially distancing 6ft is not enough when singing. Singing requires moving a lot of breath farther, and scientists suggest this can transmit the virus as far away as 16ft. The problem becomes more challenging when you consider that sound moves slowly. Research suggests it becomes increasingly difficult for musicians to stay in sync at distances as small as 25ft.

While no one can really answer this question, singing together again in person may not be safe until after a vaccine is made widely available. Although we all hope and pray for this to happen sooner, most experts predict that it will take 12-18 months.

Q: Why can’t we just use Zoom or similar solutions?

Zoom and similar solutions have helped greatly alleviate the burden of social distancing, allowing for many activities to continue with some degree of normalcy. But these are designed to facilitate human conversations, which have a much higher degree of tolerance for latency (the time between when a sound is made by one participant, and when it is heard by other participants). Unlike conversations, timing is extremely critical for music. If two musicians are unable to hear each other at the same time, they have trouble staying in sync. Research suggests the threshold for this latency is about 25 milliseconds, while latency in the hundreds of milliseconds is normal for Zoom.

Q: Why do you believe this is possible?

The advantage we have is that sound travels very slowly compared to the speed of light, and the latter is our real constraint. Luckily, all digital communications travel roughly at the speed of light (up to 2/3 depending on medium), and the digitization of sound has come a long way in the past few decades. Even relatively inexpensive technology can make it indistinguishable from an analog in-person experience for the vast majority of people.

Of course, the speed of light is not infinite. The farther the distance, the higher the latency floor. Someone in California singing live with someone in China may never be possible. But at least within the context of regions like the San Francisco Bay Area, the distance can be overcome. Sound travels at roughly 1.125 feet per millisecond, which is about 874 thousand times slower than light. Accelerating sound to the speed of light should easily enable someone in San Francisco to sing live with someone in San Jose, with an audible experience that is nearly identical to standing less than a foot away.

Q: What are the challenges we need to overcome?

Here are a few challenges that we know need to be overcome:

  • Most people do not own professional quality digital audio equipment, with latencies low enough to stay below the 25 millisecond threshold
  • Not all computer operating systems and hardware drivers are configured for low latency audio, and configuring these can be challenging
  • Most people do not have high quality Internet connections at home. Even gigabit class services that many consider to be very good typically average in the 30-50ms round-trip range. However, good fiber-to-the-home and point-to-point WiFi connections can range under 10ms.
  • Low latency audio requires the use of a server that is fairly close (from a network routing perspective) to all participants. Internet latency by nature is highly unpredictable. As your traffic traverses more network links, you are likely to experience higher jitter. This requires using larger buffers, which further increases latencies.