For many productions good audio is even more of a requisite than good video.
Most (all?) digital cameras have some kind of microphone, or can have external microphones added. The advantage is that the microphone points where the camera does, will be in sync with the video, and will (should) be available over the same HDMI connection. Disadvantages include that the mic can’t be placed independently, and can’t be used without the camera.
For flexibility and so that audio is a separate system, this project is going to focus on non-camera based audio.
Recording (capturing) from a single microphone (or source) on a PC is trivial.
However the complexity quickly increases if you want to record more than one microphone or audio source at the same time.
The underlying issue is that most analog-to-digital devices have their own clock for determining when to sample the incoming signal. No two clocks are going to be exactly the same. One may have slightly longer pauses between ‘ticks’ compared to another. Over a long enough period of time the same “time stamp” in recordings from two microphones will be at different real times in the audio. There will also be drift between the audio and video if they are being captured by different devices.
For this reason, most (all?) software sound APIs, and most applications, will only capture sound from a single sound capture device.
If you want to capture sound from multiple microphones or other sound sources there are a few ways to make it work. In order of most-to-least expensive:
- Use a specialized sound card that has many inputs (and potentially outputs) such as the Focusrite Scarlett range
- Use an analog mixer that gives live control of multiple inputs- eg microphones- and outputs a single stereo pair
- Use a virtual sound card that wraps up several real sound cards and makes them appear to software as a single sound card
For dedicated multi-input sound card or analog mixers, look for devices that support “phantom power” for XLR microphones. Phantom power should be switchable on/off for each of the XLR jacks on which it is supported.
The first option should “just work.” Look for devices that explicitly state “low-latency.” ASIO drivers are advantageous- this should provide the lowest latency input.
An analog mixer is a straight forward and relatively easy to use solution to record multiple mice at the same time. You can get the final master stereo mix and connect that to the analog sound input of a PC. Even better, some include a stereo USB sound card built-in. For only two speakers you can pan one person to be only on the left track, and the other, only on the right so your final recording can be separated by speaker. For more speakers this won’t be possible.
The final option is to use a virtual sound card to combine several sound cards (eg microphones with USB) together to appear as a single soundcard to applications. This software may, or may not, deal with the potential issues of drift. It’s the cheapest option, but can be fiddly. E.g. things breaking, or mapping of device to virtual channel changing, if devices aren’t plugged in, or are moved to different (USB) ports.
It’s easy to do on MacOS as the operating system has built in support for aggregate sound cards. (Use the Audio MIDI setup app in
/applications/utilities). It’s also relatively straight-forward with Linux- via
jack. There are both text and graphical configuration tools.
For Windows, there are 3rd party applications. The most famous is Voicemeter (free for personal use.) Synchronous Audio Router (SAR) is free and works very well. This Reddit post has a very good list of audio routing apps for the PC. This is an introductory how-to article.
Aside: controlling the audio routing of computer software (where sound from different applications goes) can be frustrating. There are addons/extensions for most popular web browsers that will allow you to direct sound from different tabs/windows to different sound cards. For other apps on Windows, EarTrumpet is the sound control that should have been builtin to windows. Rogue Amoeba has some excellent tools for MacOS. There are many tools for controlling Linux routing (for both ALSA and JACK.)
Now that we have an understanding of how audio can be captured by a PC DAW (digital audio workstation) we can briefly consider microphones.
A good starting point would be to have 2 mics (presenters, or presenters + audience) that can be handheld or shock mounted, and a wireless lav microphone. Dual USB and XLR outputs is a nice to have feature- allowing the mic to be used stand-alone with any PC, or connected to a mixer/advanced capture card. An example mic is the Samson Q2U.
Mics that are USB only should be avoided, simply because they are more difficult to record simultaneously with one computer.
Every recording/stream will need to be monitored by the director so that they can hear the sound in realtime and make adjustments. Basic analog headphones can be used with built in sound cards. However USB headsets (“gaming headsets”) have some advantages- they will appear as either one or two (separating “game” sound from “voice chat”) sound outputs in production/recording software. Many have a built in microphone- usually not of great quality, but good enough for the director to “break into” the stream/recording.
- Separate audio from video
- Prioritize microphones/audio capture devices with USB + analog (eg XLR) connections
- Don’t forget headphones to monitor output!
- Don’t forget to include a “director” mic for the person running the recording/stream
- Note that many vlogger/streamer targeted hardware is specialized to be used by a single person; they can be difficult to add/use with extra microphones, cameras, etc
- If dedicated capture cards (eg Scarlett) or mixers are purchased, be VERY careful re speccing units. E.g. does it have enough inputs to cover ALL inputs? What about mix-minus? A recording mixer really benefits from having a second output bus- see this explanation of difficulties with the Behringer Q802USB. Here is a list of recommended mixers.
- Mic holders: full size and desktop mic stands; shock mounts
- XLR and/or USB cabling of appropriate lengths