Retouch 7 declared ‘huge leap forward’ by CEDAR Audio
Boasting a redesigned user interface and enhanced processing capabilities, Retouch 7 has been unveiled by UK-based audio restoration specialist CEDAR Audio. Making up part of the CEDAR Studio 7 suite, Retouch 7 has already been hailed by its developer as ‘a huge leap forward in spectral editing technology’. Also recently revealed as part of the suite is the latest incarnation of the DNS One dialogue noise suppression system, which now includes a ‘Learn’ function.
Alongside a simplified GUI, two processes have been added to Retouch 7, the first of which is named Cleanse. According to CEDAR, it was previously difficult to isolate sounds such as wanted speech in the presence of strong, but relatively short-lived background noises such as gusts of wind blowing across a microphone. Previously, users could draw around the wanted signal to suppress the unwanted audio, but the process was both time consuming and laborious. In contrast, the Cleanse function can reportedly separate wanted from unwanted signals, allowing the suppression of noise ‘at the touch of a button’, by modelling the signal outside of the region to be restored, then using this model to isolate the wanted signal. CEDAR adds that it has already applied for a patent for the process.
The second new process is named Revert, and as its name implies, it allows users to define any part of the spectrogram and return it to its original, unprocessed form. The function is described by the manufacturer as ‘much more powerful than stepping backward and forward through a list of actions’, because it allows the re-initialisation of any part of the audio, regardless of where it came in the process history, thus leaving later work untouched.
Meanwhile, the DNS One’s Learn function is a development of the algorithm contained within the DNS 8 Live dialogue noise suppressor. Based around a new Learn button on the GUI, it allows the software to calculate an adapting estimate of the background noise level and determine suitable noise attenuations at each frequency for optimum suppression.
CEDAR stresses that Learn is not a noise fingerprint and that users do not need to find a section of the audio with little or no wanted signal to take a noise measurement. Instead, it works most effectively when left on to adapt to changes in the background and surroundings. Crucially, the software retains its zero latency performance.