Spectral Analysis Display
The FFT engine allows real time, synchronous waveform examination.
Displayed is an 8 bit stereo file of a 6600Hz sine wave.
The signal was created to single precision and either Dithered or simply rounded off.
The Left channel (Red) shows the Dithered and Noise Shaped signal, the Right channel (Blue) shows the rounded signal.
Notice how much the peak quantisation noise has been significantly reduced between 0Hz and 12kHz by the Dithering and Noise Shaping.
The Noise Shaping has modified the noise spectrum to place the quantisation noise in the top half of the frequency spectrum where the ear is less sensitive.
This Dithering and Noise Shaping is available in all the Audio Suite components.
For more info on Dithering and Noise Shaping
and a demonstration [866kB] of this Dithering and Noise Shaping in action.
Below is a 1kHz sine wave at 100dB down.
A 16 bit file (naturally quantised) can not store a signal 100dB down, however by using our Noise Shaped Dithering Quantiser
extra dynamic range can be utilised at the expence of more noise at the higher frequencies.
- Nuttall -61dB
- Nuttall -93dB
- Kaiser -40dB
- Kaiser -60dB
- Kaiser -80dB
- Kaiser -100dB
- Chebyshev -40dB
- Chebyshev -60dB
- Chebyshev -80dB
- Chebyshev -100dB
Spectral Weighting Curves|
- No weighting
- CCIR 468
- CCIR RMS
- A weighting
- B weighting
- C weighting
- D weighting
- Psychoacoustic weighting
- Spectral Analysis
- Auto Correlation Spectrum
- Cross Correlation Spectrum (stereo only)
- Temporal Averaging, Averaged Spectrums : 1 - 1024
- Equation: Xout(t) := a.Xin(t) + (1 - a).Xout(t - 1)
- where a := (1/Averaged Spectrums)
Waterfall FFT Display of speech signal
| Mixer ]