Issue link: https://digital.copcomm.com/i/1546135
C I N E M A A U D I O S O C I E T Y. O R G 61 Dialogue edit before treatment of the noise floor The goal was to consolidate all automated tasks into a single morning of prep work and then use that time productively— answering emails, reviewing ADR requests, or handling other production responsibilities—rather than spreading repetitive processing tasks throughout editorial and mixing. Once all renders were complete, we generated short equal power crossfades globally on the noise tracks only. The material was then organized so that each scene occupied one of the A, B, C, or D packs before creative dialogue editorial began. What Changed During Editorial? The workflow immediately began affecting editorial decisions in ways we had not fully anticipated. Some changes were practical. Others fundamentally altered how we interacted with the material. Voice and Reverb Become Largely Fade-Free One of the first things we noticed was how little fading was required on the voice and reverb tracks. Once the noise component is separated onto its own track, natural silences emerge between words and lines of dialogue. As a result, leveling dialogue often requires nothing more than clip-gain adjustments on the voice and reverb tracks, without the constant need for fades or crossfades unless they are creatively motivated. The ability to choose different clip boundaries for each of the tracks allows for more precision when it comes to abrupt cuts between takes. Voice can cut right where a line ends and a new one starts, reverb can cut or fade later dependent on the location's reverb time, and noise might have yet another transition that best serves its purpose. Managing Noise Becomes Dramatically Easier The benefits became even more apparent when shaping and matching noise. Imagine a scene built from numerous cuts, each carrying a slightly different noise floor, now crossfaded neatly on a scene's single noise track. Simply looking at the waveform provides immediate visual feedback about where inconsistencies exist. It's usually best to resist the urge of reducing this track altogether or simply attenuating any outliers. More important than the general amplitude will be each take's level of harmonics caused by lights, generators, or cameras. First, we often batch-loaded all clips into iZotope RX via clip- by-clip to remove such harmonics, for instance, via spectral repair and vertical attenuate with 4,096 bands. Crucially for any dealings with harmonics is that none of these cleanups affect the dialogue itself, which also naturally consists of its own set of harmonics that might normally be affected, making the process far more forgiving and intuitive. After those spectral issues were addressed, smoothing a scene often became little more than clip gaining the noise clips until everything flowed naturally. The Noise Track Becomes Room Tone's Natural Descendant Historically, room tone that is added to the dialogue edit has always been something of a compromise. Whether captured on set, assembled from snippets between lines, or generated through tools such as Undertone and Ambience Match, it often functions as a patch for an edit that would otherwise reveal Dialogue edit after treatment of the noise floor