Sound volume
Otokun
I just noticed something interesting.
father
You've been turning the TV volume up and down a bit just now, haven't you? Is that what you're talking about?
Otokun
Yeah. When I'm wearing headphones and gradually increase the volume, the treble doesn't change much, but the bass becomes easier to hear, and conversely, when I turn it down, the bass becomes harder to hear. I wondered if it was a problem with the headphones, so I tried it with the TV, but it was the same.
father
That's an interesting observation. The way you perceived the change in volume is correct.
Otokun
I thought so. It seems that louder sounds make lower frequencies sound louder.
father
Do you remember when we talked about noise levels* before?
Otokun
I kind of get it. Human perception of loudness differs depending on the frequency, and noise levels are calculated by correcting for that, right?
father
That's right. I don't think I explained it that thoroughly last time, so today I'll explain it a bit more carefully. As Oto said, human hearing is designed so that even at the same sound pressure level, different frequencies can be perceived as louder.
Otokun
I believe people are less sensitive to lower-pitched sounds. That means you have to produce a louder sound to hear it as loudly as a high-pitched sound.
father
That's right. Look at this graph. This is called the equal-loudness curve for pure tones, and it's a frequency response created by connecting the sound pressure levels at which sounds of various frequencies perceive themselves as having the same loudness.
Otokun
There are red and blue lines, but I should look at the red one, right?
father
Yes. The standard was changed a few years ago, reflecting the results of extensive experimentation. 1000 Hz is the standard, for example, 40 phon *1 Looking at this line, you can see that the sound pressure level at 1000 Hz is 40 dB, while at 125 Hz it's around 60 dB, and at 63 Hz it's around 73 dB.
Otokun
So, to hear a sound at 1000 Hz with a volume of 40 dB as loud, you would need to hear it at 125 Hz with a sound pressure level 20 dB higher, and at 63 Hz with a sound pressure level about 33 dB higher.
father
That's right. But the point that Oto noticed was that when he changed the volume, it felt like the balance between the high and low notes changed.
Otokun
Yes. So, in this graph, we can compare the 40 phon line with, for example, the 80 phon line.
father
In an 80 phon line, the sound pressure level at 1000 Hz is 80 dB, and at 125 Hz, it's 90 dB, both at the same volume. Since the difference was 20 dB in the 40 phon line, there's now a 10 dB difference.
Otokun
Conversely, sounds with frequencies higher than 1000 Hz aren't affected much by the volume. So that's why the low frequencies seemed to become particularly loud when the volume was turned up.
father
That's right. The A-weighting characteristic of noise levels, a frequency-weighted characteristic, uses a 40dB sound as a representative example of the most likely sound level in everyday noise, and employs an almost inverse characteristic of the equal-loudness characteristics of a 40 phon pure tone to compensate for perceived loudness. This has been used for many years since the 1930s as an evaluation value for noise. *2
Otokun
Has it really been in use for that long?
father
That's right. It's a testament to its excellent performance rating. It's amazing that such a complex sense of sound volume can be expressed with a single numerical value.
Otokun
But since 40 dB is the baseline, the noise levels for sounds quieter or louder than that will not match our perception of how loud the sound is.
father
That's right. Not only that, but the volume of the sound is also related to the masking effect we talked about a while ago.
Otokun
That's right. I'd like to hear more about that sometime.
*Noise level: A-weighted sound pressure level
*1 For a steady-state sound, the loudness level (phon) is defined as the sound pressure level of a 1 kHz pure tone that a person with normal hearing would perceive as being the same loudness as the sound.
*2 The frequency response of A-weighting was defined in the 1930s by Fletcher and Munson of Bell Laboratories in the United States, based on the equal loudness curves of a 40-phon pure tone, and has remained unchanged ever since.
