Equal-loudness contours
Two tones with different frequencies presented at the same sound pressure level (SPL) may not sound equally loud. Equal-loudness contours represent these variations in perceived loudness across frequencies for an average person. The basis of the equal-loudness contour is the phon, a unit of loudness that represents the dB SPL necessary for a tone to elicit the same loudness as a 1000-Hz reference tone. For example, if a given tone is perceived to be as loud as a 60-dB tone at 1000 Hz, it has a loudness of 60 phon. Equal-loudness contours are defined by standard 226 of the International Organization for Standardization (ISO).
Below is a Python script that provides the frequencies and SPLs associated with a given phon value according to this standard, and plots the contours for phons between 0 and 90.
import numpy as np
from scipy import interpolate
import matplotlib.pyplot as plt
from matplotlib import rcParams as defaults
f = np.array(
[
20,
25,
31.5,
40,
50,
63,
80,
100,
125,
160,
200,
250,
315,
400,
500,
630,
800,
1000,
1250,
1600,
2000,
2500,
3150,
4000,
5000,
6300,
8000,
10000,
12500,
]
)
af = np.array(
[
0.532,
0.506,
0.480,
0.455,
0.432,
0.409,
0.387,
0.367,
0.349,
0.330,
0.315,
0.301,
0.288,
0.276,
0.267,
0.259,
0.253,
0.250,
0.246,
0.244,
0.243,
0.243,
0.243,
0.242,
0.242,
0.245,
0.254,
0.271,
0.301,
]
)
Lu = np.array(
[
-31.6,
-27.2,
-23.0,
-19.1,
-15.9,
-13.0,
-10.3,
-8.1,
-6.2,
-4.5,
-3.1,
-2.0,
-1.1,
-0.4,
0.0,
0.3,
0.5,
0.0,
-2.7,
-4.1,
-1.0,
1.7,
2.5,
1.2,
-2.1,
-7.1,
-11.2,
-10.7,
-3.1,
]
)
Tf = np.array(
[
78.5,
68.7,
59.5,
51.1,
44.0,
37.5,
31.5,
26.5,
22.1,
17.9,
14.4,
11.4,
8.6,
6.2,
4.4,
3.0,
2.2,
2.4,
3.5,
1.7,
-1.3,
-4.2,
-6.0,
-5.4,
-1.5,
6.0,
12.6,
13.9,
12.3,
]
)
def elc(phon, frequencies=None):
"""Returns an equal-loudness contour.
Args:
phon (float): Phon value of the contour.
frequencies (:obj:`np.ndarray`, optional): Frequencies to evaluate. If not
passed, all 29 points of the ISO standard are returned. Any frequencies not
present in the standard are found via spline interpolation.
Returns:
contour (np.ndarray): db SPL values.
"""
assert 0 <= phon <= 90, f"{phon} is not [0, 90]"
Ln = phon
Af = (
4.47e-3 * (10 ** (0.025 * Ln) - 1.15)
+ (0.4 * 10 ** (((Tf + Lu) / 10) - 9)) ** af
)
Lp = ((10.0 / af) * np.log10(Af)) - Lu + 94
if frequencies is not None:
assert frequencies.min() >= f.min(), "Frequencies are too low"
assert frequencies.max() <= f.max(), "Frequencies are too high"
tck = interpolate.splrep(f, Lp, s=0)
Lp = interpolate.splev(frequencies, tck, der=0)
return Lp
def plot_elcs():
"""Makes the equal-loudness-contour plot.
"""
defaults["lines.linewidth"] = 2
defaults["font.size"] = 14
fig, ax = plt.subplots(1, 1, constrained_layout=True)
x = np.logspace(np.log10(f.min()), np.log10(f.max()), 1000)
for p in range(0, 100, 10):
c, l = ("C0", None) if p != 60 else ("C1", "60 phon")
ax.plot(x, elc(p, x), c=c, label=l)
ax.legend(fancybox=False, framealpha=0)
ax.set_xscale("log")
ax.set_xlabel("Frequency (Hz)")
ax.set_ylabel("Sound pressure level (dB)")
plt.savefig(
f"../../assets/images/equal-loudness-contours.svg",
bbox_inches=0,
transparent=True,
)
if __name__ == "__main__":
plot_elcs()
Equal-loudness contours in 10-phon steps in the range 0–100 phon.
It is worth pointing out that numerous factors influence the loudness of a tone besides frequency and level that are not accounted for by ISO 226. For example, longer sounds generally sound louder than shorter ones, and the standard contours are valid only for “side-presented” sounds (i.e., via headphones). Note also that ISO 226 is only valid for a certain range of frequencies and SPLs.
Version history
- Originally posted September 21, 2019.
Related posts
- “Ripple sounds,” Dec 12, 2019.
- “Playing pure tones with PyQt5 on Mac,” Jun 11, 2019.
- “New package “Brian hears” released,” Apr 09, 2019.
- “Spectral splatter,” Mar 26, 2019.
- All posts filed under hearing, python.