Audio absence (freezing) Go to top

Audio absence algorithm analyzes the audio signal for zero temporal activity. If the audio signal is missing for more than the specified time, the absence is detected. There are two types of absence algorithms in BBT platform: checking for absence in time and frequency domains.

Algorithm supports adding the noise level in dB (default = -60 dB) which is neglected in analysis.

audio_absence

 

Audio level Go to top

Audio level algorithm analyzes audio signal by measuring the level of the signal amplitude. If the amplitude is significantly different from expected, as determined by low and high thresholds, the anomaly is detected. The audio level is calculated in dB and the thresholds are set in dB. Minimum and maximum levels, i.e. thresholds, can be set by the user as parameters.

audio_level

 

Audio discontinuities (click) Go to top

Audio discontinuities are manifested as unexpected amplitude deviations caused by “pops”, “clicks” or temporal signal dropouts. This algorithm, which is performing the analysis in the time domain, detects discontinuities in the audio signal based on the prediction of the value of the next sample. Prediction calculation takes into account last samples and allows for a statistical error in calculating the next sample predicted value.

audio_discontinuities

 

Audio clipping Go to top

Clipping is a form of a waveform distortion that occurs when an amplifier is over-driven and attempts to deliver an output voltage or current beyond its maximum capability. Driving an amplifier into clipping may cause it to output power in excess of its published ratings. Audio clipping algorithm in BBT performs the analysis in both time and frequency domains and detects periods when the signal is clipped due to a very high amplitude level.

audio_clipping

 

Interfering audio signals Go to top

If a non-linear distortion happens or an extraneous signal interferes with the main one, this algorithm detects the audio interfering anomaly. Algorithm performs analysis in frequency domain. For a successful detection, this algorithm seeks help from audio discontinuities and audio clipping algorithms.

audio_interfering

 

Audio compare Go to top

Besides checking the quality of the audio signal, as performed by previously mentioned algorithms, BBT platform allows comparison between two audio streams. This comparison can be used when verifying functionality of devices operating with audio, by comparing the actual audio output of the device with the referent (expected) output. Signals are compared in the frequency domain, i.e. their Fourier transforms obtained by FFT algorithm are compared. Similarity thresholds can be manually set as parameters.