Skip to Content

Test signals

Along with listening tests SoundExpert performs objective measurements of audio devices and technologies. For the purpose a new audio metric is used. The core of the method is testing with some input signals and measuring level of their waveform degradation at the output. While test signals can be of any kind the measurement procedure stays the same and waveform degradation is measured with a single parameter - Difference level (dB). Below is a list of actual reference signals along with their time domain representation, frequency domain representation and Matlab code for their generation. All signals are 30s long, -10dBFS(rms), stereo, 16bit and have 44100Hz sample rate.

Sine wave 1 kHz

Sine wave 1k in time domain Sine wave 1k in frequency domain
% Sin wave 1000 Hz
% 30s, -10 dBFS (rms), 2 channels, 44100 Hz, 16 bit
Fs = 44100; % [Hz]
L = 30; % [s]
G = -10; % [dB]
F = 1000; % [Hz]
Ls = round(L*Fs);
t = (0:Ls-1)';
ref = 10^(G/20) * sin(t*2*pi/(Fs/F));
ref = ref .* (2^15);
ref = round(ref);
ref = [ref,ref];
audiowrite('ref_sine1k.wav', int16(ref), Fs, 'BitsPerSample',16)

Sine wave 12.5 kHz

Sine wave 12.5k in time domain Sine wave 12.5k in frequency domain
% Sin wave 12500 Hz
% 30s, -10 dBFS (rms), 2 channels, 44100 Hz, 16 bit
Fs = 44100; % [Hz]
L = 30; % [s]
G = -10; % [dB]
F = 12500; % [Hz]
Ls = round(L*Fs);
t = (0:Ls-1)';
ref = 10^(G/20) * sin(t*2*pi/(Fs/F));
ref = ref .* (2^15);
ref = round(ref);
ref = [ref,ref];
audiowrite('ref_sine12.5k.wav', int16(ref), Fs, 'BitsPerSample',16)

DFD 12.5 kHz / 80 Hz

Difference Frequency Distortion (DFD) signal is described in the standards IEC60118 and IEC60268. It consists of two pure tones: 12460 Hz and 12540 Hz and is useful for assessment of intermodulation distortion.

DFD 12.5k in time domain DFD 12.5k in frequency domain
% DFD 12.5 kHz mean, 80 Hz difference (12460 Hz + 12540 Hz)
% 30s, -10 dBFS (rms), 2 channels, 44100 Hz, 16 bit
Fs = 44100; % [Hz]
L = 30; % [sec]
PdB = -10; % [dB]
F1 = 12460; % [Hz]
F2 = 12540; % [Hz]
Ls = round(L*Fs);
t = (0:Ls-1)';
ref = sin(t*2*pi/(Fs/F1)) + sin(t*2*pi/(Fs/F2));
Pr = sqrt(mean(ref.^2));
Pd = 10^(PdB/20)/sqrt(2);
Kp = Pd/Pr;
ref = ref .* Kp;
ref = ref .* (2^15);
ref = round(ref);
ref = [ref,ref];
audiowrite('ref_dfd12.5k.wav', int16(ref), Fs, 'BitsPerSample',16)

Square wave 1 kHz

Band-limited version of square wave is used. It is generated by means of Fourier synthesis (http://www.dspguide.com/ch13/4.htm):

Fourier synthesis equation Square wave coefficients
Square wave 1k in time domain Square wave 1k in frequency domain
% Fourier synthesis of square wave 1000 Hz
% 30s, -10 dBFS (rms), 2 channels, 44100 Hz, 16 bit
Fs = 44100; % [Hz]
L = 30; % [sec]
PdB = -10; % [dB]
F = 1000; % [Hz]
Ls = round(L*Fs);
t = (0:Ls-1)';
n = [1 3 5 7 9 11 13 15 17 19 21]; % harmonics for 44100Hz sample rate only
A0 = 0;
Bn = 0;
ref = zeros(Ls,1) + A0;
for N = n
    An = sin(pi*N/2) * 2 / (pi*N);
    ref = ref + An*cos(t*2*pi*N/(Fs/F)) - Bn*sin(t*2*pi*N/(Fs/F));
end
Pr = sqrt(mean(ref.^2));
Pd = 10^(PdB/20)/sqrt(2);
Kp = Pd/Pr;
ref = ref .* Kp;
ref = ref .* (2^15);
ref = round(ref);
ref = [ref,ref];
audiowrite('ref_square1k.wav', int16(ref), Fs, 'BitsPerSample',16)

Triangle (odd) wave 1 kHz

Band-limited version of triangle wave is used. It is generated by means of Fourier synthesis (http://www.dspguide.com/ch13/4.htm):

Fourier synthesis equation Triangle wave coefficients
Triangle wave 1k in time domain Triangle wave 1k in frequency domain
% Fourier synthesis of triangle wave 1000 Hz
% 30s, -10 dBFS (rms), 2 channels, 44100 Hz, 16 bit
Fs = 44100; % [Hz]
L = 30; % [sec]
PdB = -10; % [dB]
F = 1000; % [Hz]
Ls = round(L*Fs);
t = (0:Ls-1)';
n = [1 3 5 7 9 11 13 15 17 19 21]; % harmonics for 44100Hz sample rate only
A0 = 0;
Bn = 0;
ref = zeros(Ls,1) + A0;
for N = n
    An = 4 / (N*pi)^2;
    ref = ref + An*cos(t*2*pi*N/(Fs/F)) - Bn*sin(t*2*pi*N/(Fs/F));
end
Pr = sqrt(mean(ref.^2));
Pd = 10^(PdB/20)/sqrt(2);
Kp = Pd/Pr;
ref = ref .* Kp;
ref = ref .* (2^15);
ref = round(ref);
ref = [ref,ref];
audiowrite('ref_triangle(odd)1k.wav', int16(ref), Fs, 'BitsPerSample',16)

Triangle (even) wave 1 kHz

This is an even-only harmonics companion of standard triangle signal. It is generated by means of Fourier synthesis (http://www.dspguide.com/ch13/4.htm) like triangle signal but using even-only set of harmonics :

Fourier synthesis equation Ass train coefficients
Ass train 1k in time domain Ass train 1k in frequency domain
% Fourier synthesis of triangle (even) wave 1000 Hz (ass train)
% 30s, -10 dBFS (rms), 2 channels, 44100 Hz, 16 bit
Fs = 44100; % [Hz]
L = 30; % [sec]
PdB = -10; % [dB]
F = 1000; % [Hz]
Ls = round(L*Fs);
t = (0:Ls-1)';
n = [1 2 4 6 8 10 12 14 16 18 20 22]; % harmonics for 44100Hz sample rate only
A0 = 0;
Bn = 0;
ref = zeros(Ls,1) + A0;
for N = n
    An = 4 / (N*pi)^2;
    ref = ref + An*cos(t*2*pi*N/(Fs/F)) - Bn*sin(t*2*pi*N/(Fs/F));
end
Pr = sqrt(mean(ref.^2));
Pd = 10^(PdB/20)/sqrt(2);
Kp = Pd/Pr;
ref = ref .* Kp;
ref = ref .* (2^15);
ref = round(ref);
ref = [ref,ref];
audiowrite('ref_triangle(even)1k.wav', int16(ref), Fs, 'BitsPerSample',16)

White noise

Generated with APx Waveform Generator Utility v4.2

White noise in time domain White noise in frequency domain
% White Noise
% 30s, -10 dBFS (rms), 2 channels, 44100 Hz, 16 bit
PdB = -10; % [dB]
ref = audioread('Noise_white_44k.32_1.wav','native');
ref = double(ref);
Pr = sqrt(mean(ref.^2));
Pd = 10^(PdB/20)/sqrt(2);
Kp = Pd/Pr;
ref = ref .* Kp;
ref = ref .* (2^15);
ref = round(ref);
ref = [ref,ref];
audiowrite('ref_whitenoise.wav', int16(ref), Fs, 'BitsPerSample',16)

BS EN 50332-1

A special test signal "Program Simulation Noise", whose spectral content is representative of music and speech. It can be created by passing pink noise through a special filter network defined in IEC 60268-1. BS EN 50332 adds an additional requirement - that the crest factor of the test signal (the ratio between the instantaneous peak level of the signal and its RMS level) be between 1.80 and 2.2. Here it is generated with APx Waveform Generator Utility v4.2.

BS EN 50332-1 signal in time domain BS EN 50332-1 signal in frequency domain
% Pink Noise (BS EN 50332-1)
% 30s, -10 dBFS (rms), 2 channels, 44100 Hz, 16 bit
PdB = -10; % [dB]
ref = audioread('Noise_bs_44k.32_1.wav','native');
ref = double(ref);
Pr = sqrt(mean(ref.^2));
Pd = 10^(PdB/20)/sqrt(2);
Kp = Pd/Pr;
ref = ref .* Kp;
ref = ref .* (2^15);
ref = round(ref);
ref = [ref,ref];
audiowrite('ref_bsen50332-1.wav', int16(ref), Fs, 'BitsPerSample',16)

 

Test set of music material "Variety"

The new audio metric allows to measure waveform degradation using music as a test signal. The following 35 tracks of various genres are used for objective measurements of audio equipment. You may see results of such measurements in the form of histogram on Portable players page and in the form of diffrogram (for each track) in corresponding articles.

#. (year) Album - Artist - Track name

1. (1932) Russian Religion Chants - Theodore Chaliapine - Twofold Litany. Glory to Thee, O Lord
2. (1959) Kind Of Blue - Miles Davis - Blue In Green
3. (1960) Elvis Is Back! - Elvis Presley - Fever
4. (1962) The Nutcracker (Complete Ballet) - LSO, Antal Dorati - Act II - Waltz Finale and...
5. (1967) Sgt. Pepper's Lonely Hearts Club Band - The Beatles - A Day in the Life
6. (1968) Ogdens' Nut Gone Flake - The Small Faces - Ogdens' Nut Gone Flake
7. (1971) Hunky Dory - David Bowie - Oh! You Pretty Things
8. (1975) In Los Angeles - James Last - Bolero '75
9. (1977) Exodus - Bob Marley + Wailers - Jamming
10. (1977) I Remember Yesterday - Donna Summer - I Feel Love
11. (1977) Marquee Moon - Television - See No Evil
12. (1981) Nightclubbing - Grace Jones - I've Done It Again
13. (1982) Thriller - Michael Jackson - Beat It
14. (1983) Swordfishtrombones - Tom Waits - Frank's Wild Years
15. (1984) The Pearl - Harold Budd & Brian Eno - The Pearl
16. (1986) Graceland - Paul Simon - Under African Skies
17. (1987) Solitude Standing - Suzanne Vega - Tom's Diner
18. (1990) Violator - Depeche Mode - Halo
19. (1993) Bach Oratorios & Contatas - Collegium Vocale, Ph.Herreweghe - Choral. Der Lebensfurst, Herr Jesu Christ (BWV 43)
20. (1994-96) Alfred Schnittke. The Complete String Quartets - Kronos Quartet - String Quartet No. 4 - IV. Vivace
21. (1994) Grace - Jeff Buckley - Corpus Christi Carol
22. (1996) Richard D. James Album - Aphex Twin - Logan Rock Witch
23. (1996) Unchained - Johnny Cash - Rusty Cage
24. (1997) Homogenic - Bjork - Hunter
25. (1997) OK Computer - Radiohead - Exit Music (For a Film)
26. (2000) Cult - Apocalyptica - Fight Fire With Fire
27. (2000) Tourist - St. Germain - What You Think About...
28. (2003) Tour De France - Kraftwerk - Chrono
29. (2004) Amassakoul - Tinariwen - Amidinin
30. (2004) Careless Love - Madeleine Peyroux - J'ai Deux Amours
31. (2005) Takk... - Sigur Rós - Með Blóðnasir
32. (2007) The Trentemoller Chronicles - Trentemoller - Klodsmajor
33. (2008) This and That - Benjamin Grosvenor - Concert Study No.3 - Toccatina
34. (2013) Psychic - Darkside - Paper Trails
35. (2015) To Pimp a Butterfly - Kendrick Lamar - For Free (Interlude)

 


 

For almost 20 years SoundExpert is the home for innovative audio measurements and research. Newly developed listener-centric audio metric combines instrumental measurements with listening assessments and helps to perform both more efficiently.

Since 2001 more than 10 000 volunteers participated in distributed blind listening tests on Soundexpert website. Sound quality ratings of 100+ codecs were computed thanks to their time and efforts. You can still take part in them.

Recently we developed a new method for instrumental measurements of audio equipment. It is based on the new audio metric and helps to examine technical performance of an audio device in great detail. Furthermore all audio devices can be grouped according to their sound signatures. The measurement procedure was fine-tuned during two years of beta-testing with 30+ portable devices. Now it is mature enough for production environment. First few tested players can be found on Portable players page.

I hope this new audio metric will attract more audio engineers and enthusiasts as it opens new field of audio research where objective and subjective are not separated anymore. Further advance in the research requires many more different audio devices to be tested - a serious challenge for our non-commercial research project. Fortunately portable players are good test subjects for this research, even the old ones. That's why I need some help from our volunteers again. If you have such old/previous/ex digital portable audio player or a smartphone that was marketed as "HQ audio inside", if it is fully functional but not in use anymore, please, consider donating it to SoundExpert. Its df-measurements will appear on the portable players page. Then it will be either kept for research purposes or sold off supporting the project additionally (or given away to some music lover if it can't be sold for some reason). I'm going to refund your postal expenses but can't confirm this right now. Please, contact me for further arrangements.

Manufacturers of portable devices can order df-measurements for their products. The cost of the work will be equal to retail price of the device. I will buy it and perform standard df-measurements. If you decide I will not publish results during a year.

Today building a comprehensive audio quality map of various audio devices on the market can be done by ourselves - audio consumers. And it's not hard indeed. In the end this will free up some time for listening and discussing Music, not technology )).

Thanks,
Serge Smirnoff
SoundExpert Founder