De-Esser –
Audio Media

Auto Dynamic Processor
Jayson Chase

The new Auto Dynamic De-Esser from SPL is, as the name suggests, a unit designed to remove the destructive level increases caused by sibilance. The Auto Dynamic De-Esser uses a new and unique circuit, designed to search the frequency spectrum for sibilance or “S” frequencies as SPL call them. Once these narrow band frequencies are detected, they are automatically removed.

Remove The Sibilance

The method used by SPL to remove sibilance involves feeding a phase inverted band of sibilant frequencies back into the signal path, thus the offending frequencies cancel each other out without any audible deterioration of adjacent frequencies. The accuracy of the SPL De-Esser is remarkable, because only the offending frequencies are inverted. There are virtually no audible side effects on the character or timbre of the voice, indeed, the process is so subtle, it’s almost unnoticed.

Layout

The Auto Dynamic De-Esser is housed in a 1U rackmount case; the unit is built around two identical but separate channels, which cannot be linked, but can be set to the same parameters. Both channels are equipped with electronically balanced XLR and stereo jacks for both inputs and outputs. The output XLR and stereo jacks are connected in parallel. This enables the output to be split to both connectors, so that you can record two media simultaneously. There is also a ground lift switch to help avoid ground loop problems. The power supply is based around a toroidal transformer, which is designed to minimize induced hum and noise due to the non-existence of an air gap.

How do traditional de-essers work? The traditional way to remove sibilance would involve using a compressor. This method requires the splitting of the signal; one split is fed to a compressor then on through the channel path, the other split is fed to an equalizer or filter.

Once fed to the equalizer, the sibilant frequencies are detected manually using a parametric EQ with a high Q setting. Once the Q is set, the boost must be adjusted to increase the level of sibilance. This increased level is then fed back to the compressor via the sidechain. The threshold is adjusted so that the compressor becomes active when a high level is detected at the sidechain, and the ratio is used to decide how much reduction is required, without becoming destructive to the rest of the audio. This method is time consuming and needs a good ear and plenty of patience. It is also difficult to set a threshold level if the voice level is continually fluctuating, as say, during a live performance. For pre-recorded material, the vocals will have to be monitored through several passes to ensure the settings at the compressor are adequate; and even then, traditional de-essers produce side effects on vocals such as speaking through the noise or lisping.

How does the SPL Auto Dynamic De-Esser work? The de-esser works best when it is positioned after the microphone preamplifier and before any variable gain amplifiers, such as compressors or expanders. The front panel is sparse, because the unit is very simple to operate and it takes very little, if any, setting up. There are only four controls per channel, which probably make the Auto Dynamic De-Esser the easiest piece of equipment to use in any setup.

“S-Reduction” is the control for adjusting the amount of gain reduction. Its range is marked from zero to ten, and I found a setting between four and seen created the best results for most practical applications. As you move the S-Reduction knob through the increments, you can instantly hear the difference in the quality of the audio. The range is wide enough for you to make the effect of processing audibly destructive, this makes it easy for you to detect the over use of reduction. There is an Led display which shows the amount of gain reduction in 2 dB increments from 0 dB to -20 dB. The S-Reduction control is the only adjustable parameter – turn the knob and the Auto Dynamic De-Esser will do the rest for you.

A very useful option is the “Auto Threshold” function. This allows the input gain to automatically adjust itself to any variations in level created by the vocalist moving in front of the microphone. This function becomes invaluable when used in live situations, especially with an untrained vocalis where the distance between them and the microphone can vary dramatically.

This function also does away with constant monitoring and manual gain riding normally associated with de-essing. The gain reduction will vary with the input level, no matter how much the levels vary. Each channel can be switched in or out by using the “Active” button, allowing quick and easy comparisons between processed and unprocessed material. The “Male/Female” switch is a preset Q designed to quickly detect the different characteristics of the male and female voice. The frequency band for female voices is set at 7 kHz, and for male voice 6 kHz.

Summary

SPL have built a simple and most effective method of removing sibilance, making the Auto Dynamic De-Esser a unique and versatile product. Its ability to remove sibilance without affecting the adjacent frequencies is very impressive, as is its ability to react quickly to variations in input levels. When working either in the studio or in live situations, this can prove invaluable. I found the uses for the Auto Dynamic De-Esser were not restricted to vocals; it also worked well with cymbals and other bright instruments that I tested. It simply smoothes the rough edges off anything that is sibilant or contains “S” frequencies.