Audio Performance Transfer (APT) Process
"Revoice Pro allows me to reference the pitch and pacing of a recorded piece of on-set dialog and then impress these timing and pitch features upon a dialog recording from another take or a clean wild line, void of troublesome distracting sounds."
"I use Revoice Pro to create double vocal parts and harmonies. It saves me lots of time and energy, which can be spent elsewhere."
The APT Process transfers user-selected timing, pitch and level characteristics of a good "Guide" audio signal to one or more audio signals - automatically.
Applications for the APT process include:
Vocal and instrumental double tracks: Tighten the timing, pitch and vibrato of “stacked” tracks while keeping them natural-sounding.
Dialog replacement: Lip-sync dialogue (ADR) and vocals by the same or different performers, even when there are noisy guide tracks.
Transfer a performance: Change the inflection in dialog (ADR, voice-overs etc.) with the desired pattern changes provided by recording the director or dialog editor's voices.
Fix corrupted live vocal recordings: Replace and sync problematic vocals with clean retakes.
Description of the APT Processing
The user loads or transfers the Guide and Target (Dub) audio to be processed into the Revoice Pro program from DAWs or audio files. The Guide signal should have the "perfect" features: either timing, pitch or both.
However, if your Guide audio needs further time or pitch modifications, you can use Revoice Pro's new built-in Warp function to quickly and transparently adjust the Guide's (or lead harmony parts') timing and pitch.
In Revoice Pro, you can insert APT function blocks wherever you want them in dedicated "tracks" under each audio track that show where the processing starts and stops. You also chose the initial settings.
The APT process first determines the timing relationship between key features of the Guide and Target (Dub) audio signals using state-of-the-art pattern recognition algorithms based partly on Synchro Arts' VocALign technology.
Next, the APT process measures performance-related features (such as timing, pitch, vibrato and level) of both the "Guide" (which usually has certain preferred performance characteristics) and one or more Targets (Dub) signals.
User-selected Guide performance features are transferred to the Target audio to create a new Modified Target signal with the selected Guide features.
The transfer process usually includes the timing alignment, so Guide features are incorporated at precisely the correct places along the Target. However, this is an option.
If satisfactory, the processed Dub(s) can be instantly transferred back to the DAW via a number of methods including Revoice Pro Link plug-ins, drag & drop or copy & paste.
- The "tightness" of the transferred feature can be controlled to allow natural acceptable differences to pass through unchanged.
- Optional user-inserted "Protected Regions" allow selected sections of the Dub signals to be left unchanged in timing or pitch by the APT process.
- The automatic APT processing can be followed by (or replaced by) manual editing of the output features.
- APT Pitch (and other feature's) processing settings can be made time-varying (automated).
- "Process-based" solo switching, via single keys, provides instant audition to compare inputs and outputs one or two signals at a time.
- The effects of any changes made to APT control panel parameters will be immediately visible (after rendering) in the displayed output graphics of pitch, level and waveform traces.
- Multiple Dub signals can use the same Guide and be efficiently processed at the same time.
- A comprehensive Preset system allows saving and reusing settings
Case Study - Replacing Noisy Dialogue
When José Luis Díaz found himself with location dialogue ruined by the sound of vehicle noise he turned to Revoice Pro to rescue the project.
“I had a take with some words ruined by the sound of moving vehicle noise. At first I tried to clean it with the top audio "cleaning" product but it was not possible. Then I searched in alternative takes and discovered one where those words were clean, but the reading was different and the pitch of the voice was too low. So it was not possible to simply replace the noisy words with the clean words. Nobody speaks in that gymnastic way.
I then tried Revoice Pro, and used the original noisy dialogue as the guide and words from the alternative take as the “dub”. After a few attempts I had clean location dialogue that had both matching time and pitch. This was not previously possible without Revoice Pro.”