mirror of
https://github.com/cookiengineer/audacity
synced 2025-04-30 15:49:41 +02:00
17ef5b1c75
* Calculate loudness for short or silent selections as well. In case of selections shorter than 400ms (one momentary loudness block), take what we have got and divide only be the actual length. Abort loudness normalization silently if the selected audio is all silent. * Fix loudness effect bug when selection includes a gap. If the selected audio in a track contained a gap between two clips, an incorrect amount of samples was processed.
292 lines
10 KiB
Matlab
292 lines
10 KiB
Matlab
## Audacity Loudness effect unit test
|
|
#
|
|
# Max Maisel
|
|
#
|
|
# This tests the Loudness effect with 30 seconds long pseudo-random stereo
|
|
# noise sequences. The test sequences have different amplitudes per
|
|
# channel and sometimes a DC component. For best test coverage, irrelevant
|
|
# parameters for the current operation are randomly varied.
|
|
#
|
|
|
|
printf("Running Loudness effect tests.\n");
|
|
printf("This requires the octave-forge-signal package to be installed.\n");
|
|
|
|
pkg load signal;
|
|
|
|
EXPORT_TEST_SIGNALS = true;
|
|
TEST_LUFS_HELPER = true;
|
|
# LUFS need a higher epsilon because they are a logarithmic unit.
|
|
LUFS_epsilon = 0.02;
|
|
|
|
# A straightforward and simple LUFS implementation which can
|
|
# be easily compared with the specification ITU-R BS.1770-4.
|
|
function [gated_lufs] = calc_LUFS(x, fs)
|
|
# HSF
|
|
f0 = 38.13547087602444;
|
|
Q = 0.5003270373238773;
|
|
K = tan(pi * f0 / fs);
|
|
|
|
rb0 = 1.0;
|
|
rb1 = -2.0;
|
|
rb2 = 1.0;
|
|
ra0 = 1.0;
|
|
ra1 = 2.0 * (K * K - 1.0) / (1.0 + K / Q + K * K);
|
|
ra2 = (1.0 - K / Q + K * K) / (1.0 + K / Q + K * K);
|
|
|
|
rb = [rb0 rb1 rb2];
|
|
ra = [ra0 ra1 ra2];
|
|
|
|
# HPF
|
|
db = 3.999843853973347;
|
|
f0 = 1681.974450955533;
|
|
Q = 0.7071752369554196;
|
|
K = tan(pi * f0 / fs);
|
|
Vh = power(10.0, db / 20.0);
|
|
Vb = power(Vh, 0.4996667741545416);
|
|
|
|
pa0 = 1.0;
|
|
a0 = 1.0 + K / Q + K * K;
|
|
pb0 = (Vh + Vb * K / Q + K * K) / a0;
|
|
pb1 = 2.0 * (K * K - Vh) / a0;
|
|
pb2 = (Vh - Vb * K / Q + K * K) / a0;
|
|
pa1 = 2.0 * (K * K - 1.0) / a0;
|
|
pa2 = (1.0 - K / Q + K * K) / a0;
|
|
|
|
pb = [pb0 pb1 pb2];
|
|
pa = [pa0 pa1 pa2];
|
|
|
|
# Apply k-weighting
|
|
x = filter(rb, ra, x, [], 1);
|
|
x = filter(pb, pa, x, [], 1);
|
|
|
|
# - gating blocks (every 100 ms over 400 ms)
|
|
block_size = 0.4*fs;
|
|
block_overlap = 0.3*fs;
|
|
block_count = floor((size(x)(1)-block_size)/(block_size-block_overlap))+1+1;
|
|
|
|
x_blocked = zeros(block_size, block_count, size(x)(2));
|
|
for i=1:1:size(x)(2)
|
|
x_blocked(:,:,i) = buffer(x(:,i), block_size, 0.3*fs, 'nodelay');
|
|
end
|
|
|
|
lufs_blocked = 1/(block_size)*sum(x_blocked.^2, 1);
|
|
lufs_blocked = sum(lufs_blocked, 3);
|
|
|
|
# Apply absolute threshold
|
|
GAMMA_A = -70;
|
|
lufs_blocked = -0.691 + 10*log10(lufs_blocked);
|
|
valid_blocks = length(lufs_blocked);
|
|
valid_blocks = valid_blocks - length(lufs_blocked(lufs_blocked < GAMMA_A));
|
|
lufs_blocked(lufs_blocked < GAMMA_A) = -100;
|
|
lufs_blocked = 10.^((lufs_blocked+0.691)/10);
|
|
|
|
# Apply relative threshold
|
|
GAMMA_R = -0.691 + 10*log10(sum(lufs_blocked)/valid_blocks) - 10;
|
|
lufs_blocked = -0.691 + 10*log10(lufs_blocked);
|
|
valid_blocks = length(lufs_blocked);
|
|
valid_blocks = valid_blocks - length(lufs_blocked(lufs_blocked < GAMMA_R));
|
|
lufs_blocked(lufs_blocked < GAMMA_R) = -100;
|
|
lufs_blocked = 10.^((lufs_blocked+0.691)/10);
|
|
hold off
|
|
|
|
gated_lufs = -0.691 + 10*log10(sum(lufs_blocked)/valid_blocks);
|
|
end
|
|
|
|
if TEST_LUFS_HELPER
|
|
printf("Running calc_LUFS() selftest.\n");
|
|
printf("Compare the following results with a trusted LUFS calculator.\n");
|
|
|
|
fs = 44100;
|
|
k = 1:1:60*fs;
|
|
x = 0.3*sin(2*pi*1000/fs*k) + 0.2*sin(2*pi*1200/fs*k);
|
|
x = (x .* [1:1:30*fs, 30*fs:-1:1]./60./fs).';
|
|
|
|
audiowrite(cstrcat(pwd(), "/LUFS-selftest1.wav"), x, fs);
|
|
printf("LUFS-selftest1.wav should be %f LUFS\n", calc_LUFS(x, fs));
|
|
|
|
randn("seed", 1);
|
|
x = [0.2*randn(2, 10*fs) zeros(2, 10*fs) 0.1*randn(2, 10*fs)].';
|
|
x(:,1) = x(:,1) * 0.4 + 0.2;
|
|
|
|
audiowrite(cstrcat(pwd(), "/LUFS-selftest2.wav"), x, fs);
|
|
printf("LUFS-selftest2.wav should be %f LUFS\n", calc_LUFS(x, fs));
|
|
|
|
fs = 8000;
|
|
randn("seed", 2);
|
|
x = [0.2*randn(2, 10*fs) zeros(2, 10*fs) 0.1*randn(2, 10*fs)].';
|
|
x(:,1) = x(:,1) * 0.6 - 0.1;
|
|
|
|
# MMM: I'm not sure how trustworthy free loudness meters are
|
|
# in case of non-standard sample rates.
|
|
audiowrite(cstrcat(pwd(), "/LUFS-selftest3.wav"), x, fs);
|
|
printf("LUFS-selftest3.wav should be %f LUFS\n", calc_LUFS(x, fs));
|
|
end
|
|
|
|
## Test Loudness LUFS mode: block to short and all silent
|
|
CURRENT_TEST = "Loudness LUFS mode, short silent block";
|
|
fs= 44100;
|
|
x = zeros(ceil(fs*0.35), 2);
|
|
audiowrite(TMP_FILENAME, x, fs);
|
|
if EXPORT_TEST_SIGNALS
|
|
audiowrite(cstrcat(pwd(), "/Loudness-LUFS-silence-test.wav"), x, fs);
|
|
end
|
|
|
|
remove_all_tracks();
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
select_tracks(0, 100);
|
|
aud_do("LoudnessNormalization: LUFSLevel=-23 DualMono=1 NormalizeTo=0 StereoIndependent=0\n");
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=2\n"));
|
|
system("sync");
|
|
|
|
y = audioread(TMP_FILENAME);
|
|
do_test_equ(y, x, "identity");
|
|
|
|
## Test Loudness LUFS mode: stereo dependent
|
|
CURRENT_TEST = "Loudness LUFS mode, keep DC and stereo balance";
|
|
randn("seed", 1);
|
|
# Include some silence in the test signal to test loudness gating
|
|
# and vary the overall loudness over time.
|
|
x = [0.1*randn(15*fs, 2).', zeros(5*fs, 2).', 0.1*randn(15*fs, 2).'].';
|
|
x(:,1) = x(:,1) .* sin(2*pi/fs/35*(1:1:35*fs)).' .* 1.2;
|
|
x(:,2) = x(:,2) .* sin(2*pi/fs/35*(1:1:35*fs)).';
|
|
audiowrite(TMP_FILENAME, x, fs);
|
|
if EXPORT_TEST_SIGNALS
|
|
audiowrite(cstrcat(pwd(), "/Loudness-LUFS-stereo-test.wav"), x, fs);
|
|
end
|
|
|
|
remove_all_tracks();
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
select_tracks(0, 100);
|
|
aud_do("LoudnessNormalization: LUFSLevel=-23 DualMono=1 NormalizeTo=0 StereoIndependent=0\n");
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=2\n"));
|
|
system("sync");
|
|
|
|
y = audioread(TMP_FILENAME);
|
|
do_test_equ(calc_LUFS(y, fs), -23, "loudness", LUFS_epsilon);
|
|
do_test_neq(calc_LUFS(y(:,1), fs), calc_LUFS(y(:,2), fs), "stereo balance", 1);
|
|
|
|
## Test Loudness LUFS mode, stereo independent
|
|
CURRENT_TEST = "Loudness LUFS mode, stereo independence";
|
|
audiowrite(TMP_FILENAME, x, fs);
|
|
remove_all_tracks();
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
select_tracks(0, 100);
|
|
aud_do("LoudnessNormalization: LUFSLevel=-23 DualMono=0 NormalizeTo=0 StereoIndependent=1\n");
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=2\n"));
|
|
system("sync");
|
|
|
|
y = audioread(TMP_FILENAME);
|
|
# Independently processed stereo channels have half the target loudness.
|
|
do_test_equ(calc_LUFS(y(:,1), fs), -26, "channel 1 loudness", LUFS_epsilon);
|
|
do_test_equ(calc_LUFS(y(:,2), fs), -26, "channel 2 loudness", LUFS_epsilon);
|
|
|
|
## Test Loudness LUFS mode: mono as mono
|
|
CURRENT_TEST = "Test Loudness LUFS mode: mono as mono";
|
|
x = x(:,1);
|
|
audiowrite(TMP_FILENAME, x, fs);
|
|
if EXPORT_TEST_SIGNALS
|
|
audiowrite(cstrcat(pwd(), "/Loudness-LUFS-mono-test.wav"), x, fs);
|
|
end
|
|
|
|
remove_all_tracks();
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
select_tracks(0, 100);
|
|
aud_do("LoudnessNormalization: LUFSLevel=-26 DualMono=0 NormalizeTo=0 StereoIndependent=1\n");
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=1\n"));
|
|
system("sync");
|
|
|
|
y = audioread(TMP_FILENAME);
|
|
do_test_equ(calc_LUFS(y, fs), -26, "loudness", LUFS_epsilon);
|
|
|
|
## Test Loudness LUFS mode: mono as dual-mono
|
|
CURRENT_TEST = "Test Loudness LUFS mode: mono as dual-mono";
|
|
audiowrite(TMP_FILENAME, x, fs);
|
|
|
|
remove_all_tracks();
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
select_tracks(0, 100);
|
|
aud_do("LoudnessNormalization: LUFSLevel=-26 DualMono=1 NormalizeTo=0 StereoIndependent=0\n");
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=1\n"));
|
|
system("sync");
|
|
|
|
y = audioread(TMP_FILENAME);
|
|
# This shall be 3 LU quieter as it is compared to strict spec.
|
|
do_test_equ(calc_LUFS(y, fs), -29, "loudness", LUFS_epsilon);
|
|
|
|
## Test Loudness LUFS mode: multi-rate project
|
|
CURRENT_TEST = "Test Loudness LUFS mode: multi-rate project";
|
|
audiowrite(TMP_FILENAME, x, fs);
|
|
|
|
remove_all_tracks();
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
|
|
randn("seed", 2);
|
|
fs1= 8000;
|
|
x1 = [0.2*randn(2, 10*fs1) zeros(2, 10*fs1) 0.1*randn(2, 10*fs1)].';
|
|
x1(:,1) = x1(:,1) * 0.6;
|
|
audiowrite(TMP_FILENAME, x1, fs1);
|
|
if EXPORT_TEST_SIGNALS
|
|
audiowrite(cstrcat(pwd(), "/Loudness-LUFS-stereo-test-8kHz.wav"), x1, fs1);
|
|
end
|
|
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
select_tracks(0, 100);
|
|
aud_do("LoudnessNormalization: LUFSLevel=-30 DualMono=0 NormalizeTo=0 StereoIndependent=0\n");
|
|
|
|
select_tracks(0, 1);
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=1\n"));
|
|
system("sync");
|
|
y = audioread(TMP_FILENAME);
|
|
|
|
select_tracks(1, 1);
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=2\n"));
|
|
system("sync");
|
|
y1 = audioread(TMP_FILENAME);
|
|
|
|
do_test_equ(calc_LUFS(y, fs), -30, "loudness track 1", LUFS_epsilon);
|
|
# XXX: Audacity does not export at 8kHz through scripting thus this test is expected to fail!
|
|
# To ensure that this works you have to set the project rate to 8 kHz,
|
|
# export the track and check the results manually.
|
|
do_test_equ(calc_LUFS(y1, fs1), -30, "loudness track 2", LUFS_epsilon, true);
|
|
# No stereo balance check for track 1 - it's a mono track.
|
|
do_test_neq(calc_LUFS(y1(:,1), fs), calc_LUFS(y1(:,2), fs), "stereo balance track 2", LUFS_epsilon);
|
|
|
|
## Test Loudness RMS mode: stereo independent
|
|
CURRENT_TEST = "Loudness RMS mode, stereo independent";
|
|
randn("seed", 1);
|
|
fs= 44100;
|
|
x = 0.1*randn(30*fs, 2);
|
|
x(:,1) = x(:,1) * 0.6;
|
|
audiowrite(TMP_FILENAME, x, fs);
|
|
if EXPORT_TEST_SIGNALS
|
|
audiowrite(cstrcat(pwd(), "/Loudness-RMS-test.wav"), x, fs);
|
|
end
|
|
|
|
remove_all_tracks();
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
select_tracks(0, 100);
|
|
aud_do("LoudnessNormalization: RMSLevel=-20 DualMono=0 NormalizeTo=1 StereoIndependent=1\n");
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=2\n"));
|
|
system("sync");
|
|
|
|
y = audioread(TMP_FILENAME);
|
|
do_test_equ(20*log10(sqrt(sum(y(:,1).*y(:,1)/length(y)))), -20, "channel 1 RMS");
|
|
do_test_equ(20*log10(sqrt(sum(y(:,2).*y(:,2)/length(y)))), -20, "channel 2 RMS");
|
|
|
|
## Test Loudness RMS mode: stereo dependent
|
|
CURRENT_TEST = "Loudness RMS mode, stereo dependent";
|
|
audiowrite(TMP_FILENAME, x, fs);
|
|
|
|
remove_all_tracks();
|
|
aud_do(cstrcat("Import2: Filename=\"", TMP_FILENAME, "\"\n"));
|
|
select_tracks(0, 100);
|
|
aud_do("LoudnessNormalization: RMSLevel=-22 DualMono=1 NormalizeTo=1 StereoIndependent=0\n");
|
|
aud_do(cstrcat("Export2: Filename=\"", TMP_FILENAME, "\" NumChannels=2\n"));
|
|
system("sync");
|
|
|
|
y = audioread(TMP_FILENAME);
|
|
# Stereo RMS must be calculated in quadratic domain.
|
|
do_test_equ(20*log10(sqrt(sum(rms(y).^2)/size(y)(2))), -22, "RMS");
|
|
do_test_neq(20*log10(rms(y(:,1))), 20*log10(rms(y(:,2))), "stereo balance", 1);
|
|
|