ToneBoosters Sibalance v4: 3 de-essing modes and loudness detection
ToneBoosters describes its well-regarded de-esser plug-in, Sibalance, as a Swiss army knife (of sorts) for vocal tracks. It throws fairly elaborate tech, such as spectral processing and loudness detection at those pesky ess-es. The goal is clean and transparent results for the needs of your vocal and speech recordings. Thus, Sibalance is practical not just for music production, but also post-processing.
Now in its fourth incarnation, Sibalance offers three modes for de-essing. The List mode is tasked with finding sibilants independently via analysing the perceived volume (loudness). Different timbre models also play a role. Unfortunately, the developer does not offer much detail in the product description, but the approach certainly sounds interesting.
With the available Range, Attack and Release controls, you can adjust further settings for the Listen mode. You will also get visual feedback from the detailed display, which shows the waveform of the audio material, the gain reduction, but also the frequency spectrum.
The de-essing is done with spectral processing and, according to ToneBoosters, the process brings transparency. With vocals, this tends to be a good thing. After all, legibility has a major impact on getting the song across.
As an additional extra, Sibalance offers a Mid / Side editing mode. This may not be as important for vocal recordings, but it can be very useful for stereo instrument tracks. The plug-in can also help tame harsh resonances from individual tracks or entire mixes. Furthermore, sample rates between 32Hz up to 384 kHz are supported, along with low-latency processing.
Sibalance also runs on iOS devices, and presets can be imported and exported between sessions. The cross-platform concept and the fairly affordable price are things all plug-in makers should strive towards.
Price and availability
ToneBoosters Sibalance works in VST, VST3, AU, AAX and standalone formats on macOS, Windows, and iOS/iPadOS. The price is EUR 19 (or EUR 5 for iOS).
Will it de-ess a vocal in a mix, without too much harm to the top end? That would be very, very useful. RX7 will do ‘freq range de-essing’, but it still alters other things. Maybe some sort of AI util could do it? I find that when you EQ a mix that’s dull, the top end boost can cause essing. Something that could de-ess a vocal in a mix would be great.