Algorithmix ReNOVAtor

Imagine that during a music festival you recorded a unique live performance in a concert hall or church. After checking the recorded material in your studio before production and mastering you decide the recording was successful in general, except for a few annoying disturbances during some quiet passages: somebody's cough, a squeaky chair, the horn of a passing truck, a bell from the neighboring clock tower. In addition, despite the exceptional artistic interpretation there were a few significant errors: a loud scratch in the part of violin soloist and one too early tone in the brass section.

All this makes your recording unacceptable and, of course, the concert cannot be repeated. As an experienced tonemeister, you know very well that all traditional techniques and tricks fail when you try to remove the disturbances mentioned above. In these cases any kind of traditional equalization or sophisticated editing methods is usually time consuming and causes discontinuities or at least audible changes in level and timbre of the desired signal and ambience. You will say “no way” and close the project.

That was yesterday. Today, we at Algorithmix are proud to unveil the reNOVAtor, our unique solution. In such hopeless situations, reNOVAtor successfully rescues problematic live recordings from extraneous sounds and unwanted audio disturbances. You will be amazed by the high sound quality of the repaired section and how quickly it works.

reNOVAtor is a tool for repairing audio material distorted by unwanted audio events without affecting the audio signal we want to keep, even when disturbances overlap with the signal of interest. reNOVAtor uses several unique denoising techniques developed by Algorithmix between 1997 and 2001. The selected part of an audio signal including disturbing sounds is first converted to its spectral representation and mapped to a color-coded spectrogram. Then the algorithm determines the amplitudes of each spectral beam at the edges of marked area and calculates parameters for a denoisig process working on each single beam within the selection. The user can control the overall behavior of the noise remover using global parameters determining denoising thresholds for the disturbances in each spectral beam of the processed area. It works like an adaptive linear-phase multiband compressor (up to 16384 bands) or an inverse noise gate, following psychoacoustic model of masking. The noise removal may even be restricted to certain gain ranges within the selected area, which is very useful if only certain components of the signal need to be treated (e.g. a specific harmonic). Functions for automatic detection of clicks and spikes, as well as identification of tones and harmonics are also provided.

Working with the reNOVAtor is easy and intuitive. It loads and analyzes the requested part of audio material you've chosen and displays the result as a 3D spectrogram with time on the horizontal axis, frequency on the vertical axis and amplitude of the spectral components color-coded. The color assignment follows the order of the rainbow: red and yellow for low energy; green and blue for middle energy; and finally purple and white for high energy. After getting some experience, this 3D spectrogram representation allows a good feeling for localization and identification of sudden unwanted acoustical events. The spectral area of interest can be precisely marked with a resizable rectangular window. In addition, the whole PlugIn window is scalable and the spectrogram inside zoomable allowing sound repairing with surgical precision.

reNOVAtor has been continuously supported by more and more DAWs. If your editing system does not include the reNOVAtor communication interface yet, we recommend using it with the provided practical stand-alone editor, AlgoEdit. This easy-to-use and reliable editor allows audio repair and enhancement with the reNOVAtor without the need of an external audio editing system. It also provides a unique, very useful function: it can create the difference between original input and processed output signal, i.e. exactly the part of the signal that has been removed. Taking advantage of this feature, allows you to optimize parameter setup and avoid touching the desired part of the audio material. It is also possible to replace any part of the over-processed audio material by the original signal without the necessity to go through all undo steps.