Description:
Some sound files in WMA format are included to help you calibrate your headphone, loudspeaker, headphone, Hi-Fi and related audio equipment.

"ever_rising1.wma" and "freq_sweep.wma" consists of a tone that increase in pitch endlessly. The harmonics are designed to fade in from the bottom and fade out at high pitch. It is the equivalent of the visual ever rising staircase. An audio perception paradox. Its application is in the checking of audio equipment frequency response. As the harmonic peaks sweep through the spectrum, you can see the locations of resonant peaks. A peak profiling line that holds the most recents spectral peaks for more than 3 seconds will give a stable peaking envelop which is the frequency response curve of the loudspeaker system.

"intermodulation.wma" consists of two sine waves of 1KHz and 1.1KHz. If the amplitude response of the loudspeaker system is perfectly linear and there is no saturation or amplitude distortion in any part of the system, then there are only spectral peaks at 1KHz and 1.1KHz. Other wise, you will find other spectral peaks, notably at 1-n*0.1KHz, 1.1+n*0.1KHz, 2.1KHz, 2.1+/-0.1KHz, +/-0.2KHz etc..

"sine_all522.wma" consists of sine waves of 1KHz, 2KHz, 3KHz, 4KHz and 5KHz having same amplitude. Under ideal condition, they should create 5 equal spaced spectral peaks in an audio spectrum analyzer. For the overall system having flat response, the 5 peaks should have the same height. If it does not, the problem may be due to one or more of the pocketPC output, any external amplifier, speaker, microphone, signal amplifier, and analogue to digital converter. It is also possible for one part in the system to compensate for imperfection in another part, such as the equalizer or equalizing amplifier.

"square.wma" is a simple 100Hz pulse. It can be used to see the step response of the overall system. A stable triggering of the waveform is needed.

To use the files, select it in a MediaPlayer. Also select looping or repeating mode of playback. Then you can begin to play the sound file. Most players create an artificial silent time between repetitions. The ears can detect this easily. But the program can skip over the silent period if we choose the long-time average mode. For accurate frequency measurement, the narrow band analysis should be used. However, the wider shape of spectral peaks for wide band analysis makes peaks easy to identify and measure.

Normally, you will connect your earphone output to the audio equipment under test. Then you bring the pocketPC near the speaker of the system, and run a spectrum or spectrogram display porgram while the Media or MP3 Player is running in the background. Some Media Players amy not be able to produce continuous sound even within the time span of the sound file, which cannot be used.

All sound test files consist of stable pitch tones. They create concentrated spectral energy far above the background white noise level. Therefore, accurate measurements can be made without special noise control. For measurements with fixed pitch tones, long time averaging can be used to provide a very stable spectrum display and provides further noise rejection. A good real-time spectrum analyzer like the HandDee-SA makes this possible.