Theory and Practice – By Dale Shirk
Today’s electronics can handle many of the tasks ordinarily performed by a sound system operator but not all of them.
Today’s sound systems can be quite complex. Many times that complexity exceeds the requirements for simple a talking head meeting. Fortunately the same system can usually be set up for simple automated operation with no operator present.
Today’s electronics can handle many of the tasks ordinarily performed by a sound system operator. An automixer can turn mics on and off as they are being addressed. Leveling amplifiers can adjust the volume for louder and quieter talkers, and for different microphone techniques. Feedback chasers can take care of minor feedback problems that may arise.
What the electronics cannot do very effectively is actually mix. Sure, it could set two signals to the same level, but that’s rarely desirable. Setting the balance between two voices, or between a voice and an instrument is an artistic judgment best left to humans who understand the context and desired intent.
Political debate shows notwithstanding, members of polite society take turns speaking so that each may be heard. In this format the electronics can do an excellent job controlling the audio. Much has been written over the years concerning automixers and the advantages and disadvantages of different operating principles. I won’t repeat that here, but instead discuss what else goes into making an automated system work really well.
Today’s digital signal processors can make the addition of automated operation a low cost option, sometimes requiring only a little programming to set up. Generally you want the system to either operate under full control of the operator or completely automated with only a few speaking mics available. The switching is handled by selecting presets or configurations in the DSP. It is highly recommended to mute all the inputs from the console, including monitors, during automated operation, and to mute all the automixer inputs during manual operation. This prevents unintended signal paths, and confusion.
DSPs with full mic preamps can have mics connected directly. DSPs with only line inputs will need a small automixer ahead of them. The mics that will be used for automated operation will split to both the automixer and the regular console. Another option is to have a few specially marked jacks just for automated operation. The sound operator or setup person will have to make sure the mics are plugged in where they should be.
Leveling or AGC processing can be tricky to set up, but it really is the key to a successful automated system. Again, there are several different types, each with advantages and disadvantages. They should make slow gain changes over several seconds to smoothly compensate for different signal levels. They should avoid sitting at maximum gain when no one is talking, the target gain should be several dB below maximum. They will only go to maximum gain when a quiet talker is addressing a mic. They can also incorporate a gate or downward expander to keep the system quiet. If a compressor block is used it should be after the AGC.
Another important feature that helps the AGC work well, is to not let it detect and ride low frequencies. The reason for this involves the microphone proximity effect. If an AGC is set to output a consistent level, regardless of frequency content, then when a hand-held cardioid mic is held close to the lips the boosted low frequency dominates, and the windscreen-muncher may come out quieter than the microphone-shy person. If the AGC is only looking at 500 Hz and up, then the intelligible parts of speech will come out the same level for both. Yes the mic muncher will have more bass, and more electrical level, but will not lose the important speech frequencies in the process.
When a sidechain input to the AGC is available, add a 500 Hz high pass filter to it. Where no sidechain input is available, you can get there by adding an EQ ahead of the AGC and another one after it. Set the first EQ to a LF shelf at full cut starting around 500 Hz. Set the second EQ to LF shelf at full boost at the same frequency. If you have the same frequency and cut/boost, you should end up with flat response, but the AGC will see less bass. Of course the normal high pass filter you would apply to any speaking mic should be before the AGC as well.
In setting the system up, be sure to check for any possible feedback. Coax the system up to maximum gain by talking quietly. Move the mic to any location it is likely to be used. Also be aware of differences in microphone sensitivity, a more sensitive mic can throw the system into feedback. Typically you’ll have to specify that a single microphone model or type be the only one used for automated operation.
Automated systems can work extremely well if set up properly. They can be far better than lazy, unskilled or unmindful operators. More than one church has removed the console because the automated system worked better. Give it a try, but be sure to attend an actual meeting with a real live audience, to be sure it is working properly. ds