When I started my career in the audio industry, believe it or not we used analog electronics to implement mix-minus systems.  We painstakingly mixed phase-inverted microphone signals into the overall mix (thus the ‘minus’) at different ratios, delicately tweaking potentiometers for hours to get larger systems sounding just right.  It was a labor of love, and I remember being so satisfied when we got the perfect, natural sound we were looking for.

In today’s world of digital signal processing, the ‘minus’ has been removed and in general we strategically ‘mix’ microphone inputs to outputs at different cross-point levels to give each microphone the right level at each speaker.  Unfortunately, the openness of DSP platforms allows for many ways to implement mix-minus incorrectly, leaving the end user with a subpar audio system in their conference rooms or facility.  Let’s go over five ways you can create a better mix-minus system.

1) Focus on quality over quantity
One common problem I see is audio levels being set too loud. The goal of a mix-minus system isn’t to make the audio loud, but to make it sound normal and intelligible to everyone in the room no matter where they’re sitting.  Ideally, the users shouldn’t sense that there’s voice lift in the space at all.  With gain before feedback being the central problem with all reinforcement systems, the higher you go, the closer you are to catastrophe.

2) Install the right number of speakers
Many rooms are not designed for quality mix-minus. For example, you may have a microphone at each seat around a long conference table, but not enough speakers to tailor the individual mic levels to each seating position.  The optimal setup is one microphone and one speaker per seating position, but of course that’s not always possible due to cost.  Remember, the AudiaFLEX PA-2 card is perfect for this application: you can drive up to 24 individual speaker circuits at just the right power directly from a single AudiaFLEX chassis.

3) Use automatic mixers
Automixers (gating or gain sharing) are instrumental in ensuring the level in the room is constant for any number of talkers.  With mix minus, one uses the direct outputs of the automatic mixer to feed the mix-minus matrix.  In gating automixers, you’ll want to make sure the direct outputs are set to post-gate, post NOM.  And speaking of NOM, remember that limiting the number of open mics is also best practice with gating automixers.  Of course the joke in the audio industry is you should allow three open mics, one talker to propose an idea, another to argue, and a third to tell the other two to shut up.  For the most natural sound, I would also recommend that the last mic hold should be set as well.

4) Remember that all talkers are not alike
Employing AGC or leveler objects in the signal flow is a good idea too.  Depending on the way people sit (how close or far from their mic) and how loud they inherently are, there can be a lot of variation in talker level.  For example, if one talker in the conference room is really loud, another is quiet, and another sits back in his chair away from the mic, we must be ready to compensate for all three cases.

Tesira employs the AGC object to increase or decrease the level of any talker to the target level. Employing SpeechSense™ technology to make sure the level is only being adjusted in the presence of human speech is possible as well.

In Nexia/Audia, we recommend the use of a leveler object that limits a talker’s level to about -10dBu and then apply 10dB of makeup gain with a level control ahead of the input to the automixer.  This levels-off louder talkers and leaves the quiet ones alone.  The subsequent 10dB boost in gain presents a nice 0dBu signal to the automixer for proper gating.

5) Maintain control
If the above principles have been applied, although they may ask for it, end users have no reason to control microphone levels in the room.  After you, the integrator, have demonstrated your prowess in the setup of mix-minus systems, handing over mic level control only gives the end users the tools to negate your work.  Explain to your users that you have accounted for a wide range of talker levels and positioning, and that this expertise is a big part of what they’ve paid you for.  You can then move the discussion to the signals you can allow them to control, such as program inputs, telephone/VTC receive, etc.

In summary, implementing a great mix-minus system starts with remembering the goal of speech clarity and natural sound in all listening positions.  The DSP toolbox gives us many great tools to control circumstances, keep our levels in the safe zone, and provide a great listening experience for the users.  I hope these pointers can help make your next mix-minus implementation a complete success.

If you have additional questions about mix-minus, sign up for one of our online trainings or call us at +1 503.641.7287.