The Audio Processing Rack
In order to improve the quality of the transmitted audio, I have invested in a good quality microphone and a few items of audio processing equipment. Of course, this is my personal shack setup, and other people will have different ideas about what sounds good to them and how to set up their particular choice of equipment.
My choice of equipment brand is Behringer, some people will hate it, others will love it. The Behringer range is ideally suited to the home studio recording enthusiasts, due in part to it’s budget pricing, and that makes it ideal for amateur radio use.
Let’s follow the audio path from the microphone through to the radio:
The microphone: My microphone is a Neewer NW5200 professional broadcast microphone Read more here. Neewer advertises this mic as a broadcast quality mic, and although it’s not a $3K mic, it certainly does live up to it’s claim.
The microphone is isolated from picking up any noises transmitted through the mounting system by using a suspension style shock mount. Yes, it is mounted upside down, and through copious research, it appears that this orientation makes no difference to a condenser microphone.
Alternatively, I sometimes use an Airlite headset, these are communications grade headsets.
The Mixer: The mics are each plugged into one of the two mic inputs on the small form factor Behringer Xenyx 802 mixer. The mixer is used to provide the 48v phantom power required by the microphone’s internal electronics and also to act as a pre-amp for the rest of the audio processing. The microphone signal is simply too low to use directly in the processing stages. A correctly wired microphone that doesn’t require phantom power, such as my headset, will not be harmed by the voltage supplied to it from the mixer.
Unfortunately, the Xenyx 802 mixer doesn’t have an “Aux send”, but instead has an “FX send”.
So, what’s the difference? Quite simply, the manner in which they are internally connected to the signal path inside the mixer. Normally the “Aux send” is pre-fader, while the “FX send” is post fader. As I required the send to be pre-fader, ie: the level doesn’t vary with the channel’s main level control, I had to modify the mixer as detailed in this post. With this modification, I am able to either use the FX Send control to route the audio through the processing equipment, or by using the Level control, the processing can be completely by-passed. Of course, it is equally possible to mix both processed and unprocessed signals together in varying levels providing a unique sound.
Expansion / Compression
The audio signal emerges from the mixers modified FX send and is sent to the left channel of a Behringer Composer Pro. The first stage of processing here is the noise gate/expander, in this instance it is configured as an expander as opposed to a gate. The expander works by boosting quiet parts of the audio signal. This has to be carefully adjusted, otherwise it would also amplify the background noise far too much.
The expander is then followed by a compressor. No, this compressor doesn’t inflate your tyres, it does however “inflate” the audio signal – sort of. In essence, a compressor analyses the audio signal and lessens the dynamic range between the quiet and loud parts of the signal making it a more constant level. It does this by attenuating the loud parts or voice peaks. Because everybody speaks differently, the settings one person uses may be totally wrong for someone else.
The compressor section has several controls which are described briefly.
- Threshold – how loud the signal has to be before compression is applied.
- SC EXT – routes the signal to any additional external processing.
- SC Mon – switches between external SC signal only or SC and internally processed signal.
- Ratio – how much compression is applied. For example, if the compression ratio is set for 6:1, the input signal will have to cross the threshold by 6 dB for the output level to increase by 1dB.
- LO Contour – Low frequency filter to help prevent heavy bass notes from triggering too much compression which causes a “pumping” effect.
- Attack – how quickly the compressor starts to work.
- Interact Knee – sets how the compressor reacts to signals once the threshold is passed, either hard which may sound “harsh” or soft for a more natural sound.
- Auto – automatically attempts to set the attack and release times dependant on the signal.
- Release – how soon after the signal drops below the threshold the compression stops.
- Tube – adds a subtle effect to simulate an old fashioned style tube amplifier.
- Enhancer – activates a dynamic enhancer which makes the sound appear more natural rather than electronically modified.
- Output – allows you to boost or attenuate the level of the signal output from the compressor.
- I/O Meter – determines whether the LED meters display the input or output signal level.
- IN/OUT – allows a complete bypass of the circuitry.
Unlike a nice quiet studio, my shack produces unwanted noise from the PC fans, PSU fans, and in summer I also have an extractor fan running. All these fans make the audio sound like I’m sitting by a jet aircraft waiting to take off. OK, so it probably isn’t actually that bad, but the noise can definitely be heard in the background, which is something that is most definitely unwanted.
To eliminate these noises, and anything else that may occur, I use one channel (right) of a Behringer Composer Pro which contains amongst other processing, a noise gate. This channel only uses the noise gate, as all other processing is done on the left channel of the unit.
So what does the noise gate do? In a nutshell, the noise gate simply turns off, or mutes, the audio signal until a pre-set level is reached. Because the background noise is much quieter than my voice, it doesn’t trigger the gate and is muted. Although it can be possible to hear the noise while I’m speaking, I have the mic and gate levels adjusted to maximize the wanted signal (my voice) while keeping the unwanted signal (the noise) to a barely audible level.
This control simply helps to reduce the “ssssssssssss” when pronouncing words with an “s” in them.
Prevents the output from rising above a pre-set level. Under normal circumstances, the limiter should only operate very briefly on very high signal peaks, and ideally, not at all.
The audio signal then passes to the left channel of a Behringer Feedback Destroyer Pro. This unit allows me to monitor the audio signal while preventing feedback from occurring. I have to say that the unit is very effective at quickly identifying and filtering out any feedback which may occur. The feedback destroyer may be configured in such a way that it operates as two completely separate devices, as is the case here.
I use the left channel of the feedback destroyer to prevent feedback, this uses a bank of 12 filters to locate and stop the feedback. I have these all set to auto mode which continually scan for any new frequencies which are about to cause feedback. The right channel is used for equalization as explained in stage 8 (below)
At this point in the audio chain, the signal exits the left channel of the feedback destroyer and re-enters on the right channel.
Each of the 12 filters on the right channel of the feedback destroyer are configured to be used as parametric equalisers. A parametric equaliser is very similar to the graphic equalisers that were in common use in home stereo systems in the 80’s & 90’s. The major difference being that the graphic equaliser operates on fixed frequency bands and fixed filter widths, while the parametric equaliser filters may be tailored to suit the frequency bands and filter widths as required.
The final stage in the processing chain is to send the audio signal from the parametric equaliser into a Behringer Virtualizer Pro which is capable of producing many different effects.
As the signal up to this point is mono and the Virtualizer requires two channels to successfully create a reverb effect, therefore the mono signal is split and fed into both channels.
The Virtualizer is used to add a very minute amount of reverb, in fact such a minor amount of reverb it is almost inaudible.
At this stage the signal is now pseudo-stereo due to the manner in which the Virtualizer operates. This stereo signal is now routed back to the mixer’s Aux return channels.
Isolation & Attenuation
The signal leaves the mixer after passing through the master level control. This is then fed to two independent 600ohm audio isolation transformers. The reason for using an audio isolator is outlined below in the section headed Eliminating Ground Loops.
Following isolation, a simple resistor network is employed to attenuate the line level signals down to a level suitable to be fed into the radio.
The left hand signal is routed to the Icom IC-756 Pro3 via the rear accessory socket using the pins that would normally be used for data modes, while the right hand signal is routed to the Kenwood TS-850. A data interface is connected to the accessory socket on the TS-850, thus requiring the signal to be fed into the microphone socket.
When feeding a signal into most Icom radios through the microphone socket, it is necessary to install a DC blocking capacitor in the mic audio line due to Icom supplying power on this wire for the standard microphones. DO NOT confuse this with the 8v supply that is also present on the mic socket!
Eliminating ground loops
You may be asking what is an ground loop. Ground loops are caused when you have two or more pieces of equipment connected together and in doing so creates more than one electrical path to ground. A quite detailed explanation is detailed on Wikipedia here.
How do we manage to eliminate ground loops in the audio chain?
A ground loop in the audio chain when using external processing equipment can be quite a pain to resolve. Luckily, as we are dealing with low level audio signal paths rather than supply voltage cabling it’s made simpler. In essence, we need to avoid connecting the ground (or shield) of the audio cable where it may create a loop path.
The very easiest way is to simply leave one end of every audio cable shield disconnected. Using this method will ensure that it is impossible for a ground loop to be made. The major downside with this method is that you may find the audio cabling picks up interference or hum is induced from nearby power cables. Another possible downside is that most audio processing equipment uses balanced connections and this will upset the balancing of the signals, thus leaving a shield disconnected at one end may prevent the equipment from functioning correctly.
A second method, is to disconnect the shield at one end of the cable, but this time simply put a 100-200 ohm resistor between the shield and the ground connection. This low value resistance, is just enough to prevent a ground loop, but at the same time it keeps the cable grounded at both ends.
WARNING: DO NOT EVER use either of these methods in any power, RF or other safety ground cable, ALWAYS respect the safety grounding on your equipment. It is there for a reason.
Finally, the best method is to use an isolating audio transformer. These can be purchased as ready made units for ease of insertion into the audio cabling, or it is possible to buy the transformers and easily connect them into your audio path and mount them in a small box.