As a producer, it is important to understand phonetics, a branch of linguistics that looks at how sounds are pronounced and classified. Plosive sounds, such as “p” and “b”, are taken into consideration when editing and recording by an audio engineer.
This article, however, will look at sibilant sounds. These are sounds categorized based on their whispery nature. Some of the basic examples of sibilant sounds include “s,” “ci,” and “ts”.
Listening back to a recording with multiple sibilances occurring is quite irritating to the human ear. Here is an example of a statement with a few sibilant sounds:
Sherlock sold facial products in Charleston.
Adobe Audition has a DeEssing functionality which is used to reduce or eliminate the hissing-like sounds that occur due to having sibilant phonetics within your recording.
The first step is to check the spectral view of the file. The spectral view is handy as you get to identify areas affected by sibilance quicker. Drag your mouse pointer over the section highlighted in the image below.
You will notice the mouse pointer change to a row resizer pointer.
This will allow you to drag windows and resize them in the Adobe Audition interface.
Hold down the left mouse key and drag so that your audio waveform and the spectral view, each share half of the screen.
You can also access this feature using the keyboard shortcut Shift + D or from the View menu option as shown in the image below.
This is important in helping you determine how much sibilance is affecting your file and the where in your file it occurs.
Here is an example from my file highlighted with a green rectangle.
In my audio, the sibilance is from the word fear caused by the f sound.
Once you are done identifying these areas, close the spectral view using the keyboard shortcut Shift + D.
Select the portion of the audio file that needs DeEssing. If this operation is being applied to the whole file, select all by clicking Ctrl + A.
Click on the Effects option in the menu tab followed by hovering your mouse over the Amplitude and Compression option.
A new dropdown list will appear on the right. From there, click on DeEsser.
This will open the dialogue box in the screenshot below.
There are different buttons, elements, and entries in this dialogue box. The first one is the Presets entry. In this entry, you can pick a preset based on the properties of your file. One of the presets is Low Voice DeEsser for low-level recorded sounds.
Always choose Default when you have to use custom settings
The next section you will follow up on is the Mode settings. You have two options here: Broadband and Multiband.
Broadband is used to uniformly compress all frequencies, while multiband is used to compress within the sibilance range. The former generalizes the process to all levels within the file, not as ideal as the multiband mode, which focuses on the range where the sibilance occurs. Pick the latter, and note that the processing time is a bit longer in multiband mode.
The next setting to adjust is the Threshold, represented by a slider. This is used to indicate a mark where the audio levels, beyond the level set, will have the sibilance reduced.
The CentreFrequency setting is used for checking where the sibilance is highest. As you play your audio, adjust this to get the range.
The Bandwidth is the frequency range that will trigger the compression of the sibilance.
Please note that these last two settings are applied in form of ranges. This is the importance of the graph section within the dialogue box.
Alternatively, you can adjust these two settings on the graph by hovering your mouse over the edges of the clear part and dragging.
The Output Sibilance Only checkbox allows you to hear the removed sibilance. This is key as it helps you identify if the rest of your audio will be affected by the compression. You can mark this box, playback the audio, and make further changes if parts of your audio are removed with the sibilance. When you are done adjusting, uncheck this box.
Lastly, the Gain Reduction setting found on the right side of the five described settings is used to indicate the level of your processed file. This will help with knowing if the levels are reduced considerably and generally if they are audible.
Once you are done making all these adjustments, click OK in the dialogue box to apply the changes.
While it is important to know how to remove and reduce sibilance when editing, it is also important to try and avoid it in the first place. You can achieve this by:
- Having some distance between your mouth and the microphone: Having the microphone closer to the narrator or vocalist gives prominence to the sibilance sounds.
- Having a pop filter: While a pop filter is primarily useful in reducing plosives such as p and b, it also minimizes hissing by filtering it from the diaphragm of the microphone.
- Using the right amount of gain on your console or mic: Adding a lot of gain gives prominence to these sounds which will be hard to get rid of when editing.
- Knowing your voice talent: This is a bit rare as compared to the tips mentioned previously. You want to look for speech impediments such as lisps, as well as issues articulating statements with multiple sibilant sounds. Knowing the capabilities and weaknesses of your talent will help with the execution.
- Avoiding cheap gear: Cheap microphones do not have good workarounds when it comes to getting good headroom while recording. With cheaper gear, your recordings are prone to distortion.
- Put in the work upfront: ensure you capture the best sound possible. Sibilance comes up as a result of overprocessed audio. If you have a good setup, there will be no need to apply multiple effects to correct a bunch of problems.