The search for perfect video-conferencing app

Rishi Tejpal and Kanika Saxena write for ET Prime:

Users have been voicing their distaste for minor yet significant issues popping up during an active session. And there are many. Unable to unmute, audio not available for some, a shared screen showing grey boxes covering the main content, ‘raise hand’ feature getting triggered for participants even when there was no such command among others.

Despite its share of bad publicity, Zoom has kept its game up. The features and tools work just the way they are meant to function — sessions provide high-resolution video conferencing, seamless transition between multiple screens being shared at once, notations are visible to all, and hassle-free streaming of online video over a shared screen during the call. Additionally, there are lesser delays and downtime.

(emphasis mine)

I apologise the write up is behind the paywall, but there is so much inane repetition that I’d choose to highlight only specific areas of interest. I am not interested in the “feature comparison” between applications. I am keen to understand how the plumbing works – codecs, data centres and the works. If Telemedicine must work (for any reasonable stretch of imagination), it must do better than putting up two people in front of the computer. One of the major reasons why Telemedicine didn’t work in the pandemic (to optimal degree) was insufficient integration with the EMR. I don’t perceive anything happening in that space though.

Here’s the comparison spec from the ET Prime:

In the new upcoming Windows 11, Teams will be integrated with the OS and Skype is being depreciated. I am not a huge fan of Teams – it lacks coherence and is a completely “shi**y experience”. The only usage that product managers crow about is the forced integration with other Office products and locked in enterprise licenses. It means that in the pandemic, companies wouldn’t reinvest in a “better solution” like Zoom if they are getting the solution for “free”.

Here’s an interesting explainer from the duo about Zoom:

Zoom uses scalable video codec (SVC) instead of the commonly used advanced video codec (AVC) for transmitting video. SVC dynamically adjusts the video resolution or frame rate in real-time, based on the network performance.

It uses a distributed architecture of cloud servers spread across the globe to deliver optimal performance by using the server nearest to the participant. This is something closely synonymous with edge computing that streaming service providers use to provide a seamless streaming experience.

(emphasis mine)

There are other technical tricks up its sleeve, but it boils down to the data-centres and “edge computing” – bulk of transcribing or encryption which happens locally on the device. Haven’t you noticed how the battery drains after a “Zoom session”?

The write up misses up on locally hosted solutions like Jitsi. I am keenly watching Apple’s Facetime and Telegram’s video calling option (currently limited to beta). Telegram has nailed the voice solution for masses (akin to clubhouse) and is rapidly iterating the video calling for multiple participants. Needless to say, it is incredibly data intensive operation, and the success will depend on codecs and “streaming quality”.

Which is the best option, then? I am impressed with Zoom despite its Chinese origins. I have used it as early as 2012 using BlackBerry Messenger – it was a paid option, and BBM had tied up Zoom then. It was a solution ahead of its time. (pre 3G era and spotty Wifi connections), which limited its appeal. Zoom has come a long way since then.