Accentize has released DeRoom. The plug-in uses an AI-based algorithm to reduce or remove reverb and room ambience in your audio tracks in real time. In addition to music production, this can be really useful for post production and podcasting applications.
Removing or reducing reverb after recording used to be impossible. Plug-ins like Zynaptiq Unveil and Accusonus Reverb Remover have made it a feasible option, and Accentize just added another alternative to this list. According to the developer, DeRoom uses a trained neural network to separate the dry sound from the room reflections.
Accentize says that DeRoom’s neural-network-based algorithm was trained using “many different room scenarios”. The company claims that the plug-in determines the room timing constants automatically, so all you need to do is set the desired amount of reduction. That said, DeRoom does provide three different settings for small, medium and large rooms. It also offers controls for sensitivity, reduction and make-up.
Just like Accentize VoiceGate, which was released only a few weeks ago, DeRoom works in real time. Although the plug-in is specifically optimized for speech recordings, Accentize says that it also works very well on other types of signals, such as drums.
Next time you discover that a recording has too much ambience, you now have another interesting possible solution to try. The arrival of AI algorithms and machine learning in audio plug-ins is opening up exciting new possibilities, and reverb removal is certainly one of them.
Price and compatibility
Accentize DeRoom runs on Windows 7 or higher and macOS10.12 or higher in VST, VST3 and AU formats. The developer recommends at least 4 GB of RAM. Until March 27, 2020, the plug-in is available for EUR 49 from the Accentize website. After that, you’ll pay EUR 69. You can download a free 7-day trial version to check it out.