I just created an introductory-level video describing a particular class of audio glitches that frequently show up in digital audio streams on the web. You may hear these digital artifacts in webinars, webcasts, or podcasts where presenters speak over a computer microphone.
I made the video as non-technical as I could. Networking specialists and audio engineers will cringe at some of my half-explanations and oversimplifications of concepts. I also allowed some of my demonstration edits to remain imperfect for the sake of brevity.
If you are reading this in a feed that allows the embedded video to play, you can watch it here (it is less than 12 minutes long):
Otherwise you can click this link to play the video on YouTube: https://youtu.be/T8GycwnAOFc
For those who want to dig a little deeper, here are a few extra insights and amplifications on things I glossed over in the video:
- The types of audio glitches I highlight are NOT caused by the recording process. You will hear them in the live session as well.
- It is practically impossible to track down the root cause when you hear these things in a live session. It could be buffering and congestion on your side of the chain as a listener, it could be a major network problem upstream from the speaker. But the MOST common problem is temporary and intermittent congestion or signal degradation on the speaker's home wireless internet. This can pop up at any time in a session, even when things sounded great earlier. You can help reduce the chance of it occurring by having speakers devote as much network bandwidth as possible to their webcasting/recording session. Have them turn off email, web browsers, instant messenger, and make sure that nobody else in the house is doing things on the internet (especially no movie watching or video conferencing). The safest thing of all is to hardwire your computer to your router with an Ethernet cable.
- Once you start chasing these things down in a post-production edit, it can drive you crazy and use up your life. You can spend five minutes on a one-second glitch. Multiply that by the number of them you can find in a one-hour recording, and it gets hard to cost-justify or time-justify the work.
Anyway, I hope you found it interesting and educational. Feel free to ask additional questions in the blog post comments.