BRIL NOISE REDUCTION ALGORITHM


BRIL is a high quality blind noise reduction algorithm. It takes as its input a noisy audio or voice signal and tries to estimate and reduce the noise in the input signal without distorting the audio. In most cases the input signal is recorded by a single-microphone. The source of the noise in such recording may vary from the environment (background) noise to the machine, electronics, or even the microphone self noise.





What professionals say about BRIL

Tony Bartosiewicz from Dynalec Corporation after testing BRIL integration on Dynallec LPC2387FBD100 microcontroller device

"I just wanted to let you know that I've had a little time to test the software you sent. Everyone that I've demoed it for has been very impressed with the noise reduction. The processing delays appear to be within an acceptable range (subjectively speaking)."





BRIL PERFORMANCE

BRIL is an efficient and effective block-processing frequency domain algorithm. It has already been deployed in many challenging applications, including tank and military vehicles communication systems, with excellent results.

CODE PERFORMANCE

The small optimized BRIL algorithm is implemented entirely in standard ANSI C, available in both floating-point as well as fixed-point implementations, and can be easily compiled to run on any DSP or CPU that has a C compiler. It has been tested and deployed on many embedded and mobile platforms with or without an operating system.


QUALITY PERFORMANCE

The sample audio files below demonstrate the audio quality of the processed speech signal in different noisy environments. The noise recordings have been obtained from NOISEX-92 database. The processed signals have been obtained by processing through BRIL with noise reduction level set to 20 dB. This allows sufficient noise reduction to be achieved without distorting the speech and at the same time preserve the environment noise characteristics. The images show the spectral contents of each sample.


input output

Volvo 340 Car at 120 km/h

BRIL Output


input output

M109 Tank Cockpit

BRIL Output


input output

Coctail Party Noise

BRIL Output


input output

F16 Cockpit

BRIL Output


BRIL FEATURES

The following is a partial list of the most important BRIL features:

  • Efficient, small, and effective block frequency domain algorithm.
  • Block length is user adjustable therefore giving the user control on processing delay.
  • Noise reduction level is user adjustable.
  • Provides up to 25 dB of background noise reduction with negligible speech distortion. Additional noise reduction above 25 dB is possible with moderate gradual degradation of speech.
  • Works at any sampling frequency without any calibration or modification. It has been already deployed at sample rates ranging from 8kHz to 48kHz.
  • Trivial to integrate due to its simple Application Programming Interface.
  • Already lab and field tested on several fixed-point and floating-point processors and DSPs with and without an operating system.
  • Supported on all major desktop, mobile, and embedded platforms.
  • Floating-point and Fixed-point implementations optimized for several general purpose processors, microcontrollers, as well as digital signal processors are directly available.

EVALUATING BRIL



Several demonstrations are available that allow you to evaluate BRIL in real-time. However, the easiest way to give BRIL a quick test to see whether it is suitable for a specific purpose is using the file processing test application. This simple command line application reads an audio wav file (noisy speech), processes that file through the BRIL algorithm to reduce the noise in the input file by the specified noise reduction level given at the command line, and writes the result to the output wav file. For instance if you have a recorded noisy speech signal named "NoisySpeech.wav" and would like to reduce the noise in this recording by 20dB you may give the following command at the command line prompt

bril.exe NoisySpeech.wav BrilOut.wav 20

Click this link to download the Windows version of BRIL test application.

Please note that only uncompressed PCM wav format is supported. If you have a recorded file in any other audio formats, including ADPCM wav or any other encoded or compressed wav file you will need to convert it to PCM first using any available audio file editor. Simply open the recording in the audio file editor of your choice and select "File->Save As" and then select PCM wav format.

For more information on the BRIL algorithm we appreciate taking the time to contact us using the web form.