RENDERING WIDEBAND ULTRASONIC SIGNALS AUDIBLE

Abstract

The invention relates to a method for rendering ultrasonic signals audible that is characterized in that the temporal dynamic range of the ultrasonic signal is maintained. The amplitude profile of the ultrasonic signal picked up in the time domain remains unaltered. The frequency shift from the ultrasonic range to the audible range is possible up to a factor of 32 using the present invention.

Claims

1. Method for rendering ultrasonic signals audible, characterized in that the temporal dynamic range of the ultrasonic signal is maintained.

2. Method according to claim 1, characterized in that only that component of the ultrasonic signal whose amplitude variation is in the audible range is processed.

3. Method according to claim 1, characterized in that the ultrasonic signal has its frequencies compressed.

4. Method according to claim 1, characterized in that an ultrasonic signal is detected using a suitable microphone or a suitable sensor, the analogue signal is converted into a digital signal using an A/D converter, the digital signal is transferred to a suitable computation unit, a continuous data stream is transferred from the computation unit to a D/A converter, the acoustic signal obtained is output.

5. Method according to claim 4, characterized in that a computation unit performs block-by-block breakdown of the time signal, performs block-by-block transformation of the time signal into the frequency domain, performs block-by-block back transformation of the time signal into the frequency domain, and performs synthesis of the signals transformed block by block.

6. Method according to claim 4, characterized in that a frequency shift for the ultrasonic signal up to a factor of 32 is performed.

7. Method according to claim 1, characterized in that an ultrasonic signal is digitally sampled, 1/n octaves are computed from the original signal by a filter bank, a time-dependent level value is computed for each narrowband octave, bandpass noise from the target frequencies is produced in the audible range, the acoustic signal obtained is output.

8. Method according to claim 7, characterized in that the narrowband octaves are scaled using a suitable factor, scaling being able to be effected in linear or nonlinear fashion.

9. Method according to claim 1, characterized in that the original time signal of the ultrasonic signal is registered, a narrowband spectrum around a carrier frequency registers a narrowband signal, the carrier frequency is automatically or manually varied, the narrowband signal is reproduced for each carrier frequency in the audible range, each band of the ultrasonic range being allocated a band in the audible range.

10. Method according to claim 9, characterized in that the signal is additionally compressed.

11. Method according to claim 1, characterized in that the acoustic signals obtained are output substantially in real time.

12. Method according to claim 1, characterized in that the acoustic signals obtained are output onto a storage medium.

13. Method according to claim 1, characterized in that an inverse A-rating is implemented in the time signal of the output channel.

Description

[0058] The invention is explained in more detail below with reference to 3 drawings.

[0059] FIG. 1 shows a depiction of the power spectrum for application of the method using the Fourier transformation.

[0060] FIG. 2 shows a depiction of the power spectrum for application of the method using a filter bank.

[0061] FIG. 3 shows a depiction of the power spectrum for application of the method using the evaluation of a narrowband signal range around a carrier frequency.

[0062] The sound pressure values in FIG. 1 to FIG. 3 are not referenced to 20 μPa. They are only relative indications of the sound pressure in dB.

[0063] The aspect of rendering ultrasonic signals audible relates to signals in the time domain. The illustration is provided in the frequency domain here, for reasons of better comprehensibility. The depiction in the frequency domain illustrates not only the aspect of rendering ultrasonic signals audible but also the requirement of data compression. The level values are not referenced to a reference value.

[0064] FIG. 1(a) shows the frequency spectrum (logarithmized representation of the power spectrum of a real ultrasonic signal produced by technical means). FIG. 1(b) depicts the spectrum that has been computed by means of vocoder methods and Fourier transformation. FIG. 2 shows the frequency spectrum (logarithmized representation of the power spectrum of a real ultrasonic signal produced by technical means) and the rms values of these bands, produced by means of a filter bank. FIG. 2(b) depicts the digitally produced noise curve that is used for the weighting with the intensities of the narrowband octaves. FIG. 2(c) shows the narrowband octave spectrum that is used for the output. The values represent the intensity of the signal to be rendered audible in the respective band that is used to weight the noise function.

[0065] FIG. 3(a) shows the frequency spectrum (logarithmized representation of the power spectrum of a real ultrasonic signal produced by technical means). Highlighting denotes that portion of the spectrum that is influenced by the mixing and that is available after the transformation for rendering the signal audible. In this method, the frequency axis is not scaled. Frequency differences are maintained. FIGS. 3(b) and 3(c) relate to the two mixed frequencies (40 and 60 kHz) and the different intensity, conditional thereon, of the respective output. The lower intensity in the range around 60 kHz in comparison with 40 kHz is also reflected in the spectrum of the down-converted signal.