The music sounds better with you
The highlight of RX 7 is the new Music Rebalance module (Standard and Advanced), which uses machine learning to divide mixed tracks into four constituent components – Voice, Bass, Percussion and Other – each of which can then be independently levelled in a simple mixer, by up to +20dB or all the way down to silence.
Perfection is never to be expected with deep spectral processing like this, but damn, Music Rebalance is impressive. Not only is it hugely effective for boosting or lowering drums, basses and vocals/voiceovers within the mix, but it confidently handles total isolation or removal of those track elements (vocals, most importantly), too – remixers will love it.
Sure, there are inevitably artefacts and spectral ‘holes’ in the retained material – how could there not be? – but they’re nowhere near as pronounced as we initially expected. It’s genuinely amazing and a remarkable achievement on iZotope’s part.
Being an offline process, it’s worth noting, Music Rebalance is only available within RX 7 itself (there’s no separate VST/AU/AAX plugin) and as an AudioSuite plugin for Pro Tools. It does, however, include a (‘Reduced Quality’, as demanded by the latency of the algorithm) preview function.