The latest plug-in to join Accentize’s collection is said to take a new approach to dialogue processing, and offers a fast and intuitive solution for those working in film, TV, podcast production and audio restoration. Powered by advanced machine-learning technology, dxSplit is capable of separating audio into voice, reverb and noise components, and provides independent level and EQ controls over all three elements.
The company say that while competing plug-ins tend to focus on suppressing unwanted sound, their latest offering provides a new way to rebalance and refine dialogue recordings using source separation. Neural network algorithms are used to split the incoming signal into three distinct stems, each of which feature their own level control and parametric EQ, providing users with detailed control over each element and making it easier to shape the sound of the voice while removing or taming unwanted reverb and noise. Thanks to support for additional outputs, it’s possible to route all three signals to individual DAW tracks for additional processing, if required.
dxSplit also integrates with Accentize’s dxRevive, a speech-enhancement plug-in that is capable of restoring missing frequencies and removing processing artefacts. The processor can be loaded directly into dxSplit’s Voice stem, where it can be used to enhance the natural character of the original speech once the noise clean-up process has been applied.
Compatibility
dxSplit is supported on PCs running Windows 10 or above, and Macs running macOS 13.1 and higher. VST3, AU and AAX plug-in versions are available.
Pricing & Availability
dxSplit is available now, and is subject to the following introductory pricing offers until 31 March 2026:
- dxSplit: €225 (usually €299)
- Loyalty Discount (dxRevive Pro owners): €179 (usually €200)
- Accentize Restoration Bundle: €449 (usually €499)
A free trial version is available.
