Providing audio for a livestream can be tricky business. Here are some tips to ensure that your stream has the highest possible audio quality.
Back to the Basics
High-quality audio starts at the source, and this is especially important when livestreaming due to the processing your audio suffers when it’s run through streaming software or a hardware encoding device. One of the most common issues with livestreams is poor intelligibility due to excessive background noise or ambience. Solve or mitigate this problem by making sure that talent is close to the microphone. In the case of a studio podcast stream, use a dynamic mic with a tight pattern (such as the Shure SM7B in Fig. 1) and ask the talent to get right up on the mic. The increased sensitivity of a condenser mic may not be your friend in such applications because increased sensitivity means that a condenser mic is more likely to capture background noise. Where possible, use acoustic treatment to damp reflective walls nearby the talent.
Livestream events on a tight budget might be stuck using a simple web cam to stream video, but don’t rely upon the built-in microphone to provide quality audio. The mics built into most web cams aren’t very good and even in cases when they are, the distance between the person speaking and the web cam reduces intelligibility. Instead, try using a lavalier mic (see Fig. 2) and experiment with position.
Instinct tells you to get as close as possible to the mouth of the speaker, but some lav mics lose intelligibility when you clip them on the collar. Try placing the mic on the chest, 8 to 10 inches from the chin, which can improve clarity and yield a higher sound level from the speaker’s voice. Avoid attaching the mic where it might pick up the sound of rustling clothing and use a windscreen in outdoor situations. If you’re doing a single-camera stream with an external microphone, confirm in the encoder that the external mic (not the mic built into the web cam) is actually being used as the sound source for the stream.
A high-pass filter on the vocal mic channel can remove unwanted low-end. If the filter frequency is adjustable, start with a setting around 80 Hz and slowly sweep the frequency up until you hear the vocal start to thin. Then back the frequency down a bit until the low end is restored and leave it there. Set a compressor on the mic for a ratio between 4:1 and 6:1, moderately fast attack, and set the threshold so that there’s a few dB of compression during normal speaking levels.
Mixing vocal microphones for a livestream is relatively easy but life gets more complicated when the stream includes live music. Ideally, the production would have a minimum of three discrete audio feeds: the house mix, the monitor mix(es), and the mix for the livestream. It’s important to remember that the mix for a livestream only incudes sound that is coming through microphones and doesn’t take into account sound bleeding off the stage from live drums or guitar amplifiers. Those sounds become part of the house mix, so they’ll probably be mixed lower in the P.A. system and if you use the house mix for the audio stream there’s a good chance that the drums and guitars won’t be loud enough in the stream. Closed-back headphones or IEMs can help when checking a mix in this situation but there’s no substitute for monitoring in total isolation from the performance area.
Livestreams featuring live music can be improved by getting rid of wedge monitors and switching to headphones or IEMs, which reduces bleed from the monitors into the microphones and cleans up the mix considerably by minimizing phase issues that result from bleed. If you’re in a situation where there’s one console and you’re running all of the mixes, use a matrix for the livestream. Dedicate an aux send or group bus for the inputs that are loud in the house (drums, guitar amps, bass amp, etc.), and mix it into the matrix along with the L and R buses. This will enable you to use the house mix as a starting point for the stream and push the level of the drums and guitars louder in the stream without the need to create another mix from scratch. There’s always some trial and error involved in the process so plan extra time to sort this out.
Encoding Basics
Audio for a livestream will ultimately be routed to some kind of encoder that combines audio and video into a single data stream and uploads the stream for broadcast. A simple example of such a device is the BoxCaster (Fig. 3), which accepts video via HDMI or composite SD ports, and accepts audio via stereo RCA inputs.
If your mixer has only balanced outputs, use a balanced-to-unbalanced interface such as the LM2U from Whirlwind (Fig. 4), which also converts +4 to -10 to prevent overloading the encoder.
It’s also possible to run encoding software on a computer; if you’re doing so, beware of overworking the CPU by simultaneously running a bunch of other programs (more on that in a minute).
When setting up the encoder, check the audio settings for compression options. Budget encoders may only provide simple “auto gain control” compression, in which case you’ll probably want to turn it off and use compression in the console for the individual channels and mix buses. Don’t be afraid to be aggressive when compressing audio for the stream because the more consistent you make the mix, the fewer surprises in the end product — which will probably be compressed again when it’s sent out by the cast provider.
Regardless of whether it’s a hardware or software encoder, avoid using Wi-Fi to upload the stream. A hardwire connection yields the best possible stream quality and keeps the stream secure from interference.
Sync or Swim
If you encounter sync issues between audio and video there are a few things you can do. A common source of sync error when using a software encoder is an overworked CPU, so run only essential programs and farm out DAW software to a separate computer. Another possible source of sync error can be excessive data on one USB bus, so if possible, use separate buses for audio and video. Most encoders allow you to set the bit rate and video resolution for the upload, and sync errors can often be eliminated by reducing either of these settings. No one wants to reduce video quality but keep in mind that most people viewing a livestream are watching on a phone or other device that may not provide high-res video anyway.
Trying to present a livestream at the last minute is like rolling a boulder uphill, so allow plenty of time to prep your gear. Previewing your stream mix from the encoder output (not from the console’s outputs) is the only way you’ll know how the encoding process is affecting your mix, and that will give you the opportunity to make adjustments. Finally — if at all possible — use a stream preview function from the streaming service to check the mix via the internet before it goes live and make sure that the audio is loud enough.
Steve “Woody” La Cerra is the tour manager and front of house engineer for Blue Öyster Cult. He can be reached via email at woody@fohonline.com.
