CAM2: DNS with LEARN™
The world standard for background noise suppression
Noise suppression is needed to clean up noisy dialogue for film production, to suppress ambient noise for live TV and radio broadcasting, to revitalise sound effects libraries, and to enhance speech for forensic audio investigations. Until recently, this meant using conventional filters, noise gates, dynamics processes, or processes developed from analogue encode/decode noise reduction systems. But these proved to be inadequate, so we developed the Academy Award winning DNS process to remove the rumble, the hiss, the whistles, the broadband noise and the shot noise from contaminated sounds. DNS splits the signal into a large number of well-defined bands, and sophisticated digital filters then analyse each of these, suppressing the noise independently in each. The innovative design of the filter bank allows you to control the DNS process using relatively few controls, making it simple and quick to use in all situations.
At first glance, the latest incarnation of DNS appears to be identical with previous versions, but its secret is revealed by the LEARN button that has appeared at the bottom left of the control panel. A further development of the groundbreaking algorithm contained within our award-winning DNS 8 Live dialogue noise suppressor, LEARN allows DNS to calculate a remarkably accurate adapting estimate of the background noise and determine suitable noise attenuations at each frequency for optimum suppression.
LEARN is not a noise fingerprint and you do not need to find a section of the audio that contains little or no wanted signal to take a noise measurement. Yes, you can use it to take a snapshot of existing conditions and then fine-tune the parameters, but its real power lies in leaving it switched on so that it can adapt to changes in the background and surroundings. It not only adapts in a fraction of a second to changes, it differentiates between the wanted signal and the noise so that you obtain superb noise suppression at all times.
If you work with film dialogue, the speed, flexibility, and ease of use of DNS provides solutions to audio problems that you could not previously solve. And, with eight, 96kHz channels of the DNS algorithm in a convenient, automated format, it cannot be bettered for multi-channel post-production in the film, video, and TV industries. Elsewhere, in the audio forensic laboratory, DNS is ideal for removing motor noise from small covert recorders, for eliminating electrical interference, and for helping to clean up recordings suffering from unfavourable acoustics and poor microphone locations.
Uniquely effective broadband noise reduction
CEDAR has been enhancing speech for more than 30 years, and one scenario has proved to be stubbornly problematic, and often impossible to solve. This is the problem of revealing and increasing the intelligibility of voices recorded in noisy and rapidly changing environments such as moving vehicles, crowded cafés, and other public spaces. The forensic audio community encounters this time and again and, while the venues and scenarios may vary greatly, they all share a common difficulty. How do you reveal a wanted speaker when the signal to noise ratio is very poor, or if the perceived noise on the recording is actually other human voices?
FNR is an automated noise reduction system for speech recordings suffering from very poor signal/noise ratios, and is capable of performance that would have seemed impossible just a few years ago. Because it's adaptive, it reacts - and reacts quickly - to changes in the background noise, and is capable of revealing much more of the wanted signal without having anything like as much effect on the wanted voices as filtering systems from elsewhere. This also makes it of great value in post and broadcast wheh there's simply no other way to reveal the wanted speech in interviews and news reports recorded in challenging environments.
Speed and simplicity
FNR is extremely quick and simple to use. It requires no noise fingerprint and constantly adapts to the noise present in the input signal. Its controls determine the nature of the noise removal by allowing the user to adjust the spectral resolution of the process as well as factors relating to its adaption and noise reduction.