About Equal Loudness Contours

Perhaps the most critically important issue about playback levels has to do with the equal loudness contours. Take a look at these, below:

	Equal loudness contours from Robinson and Dadson. These curves describe perceived equal loudness for sine waves, compared to 1 kHz. They aren’t entirely accurate for music, but they come pretty close, and they will give you a good idea about what is happening if you take the time to scope ‘em out.

These curves have been determined by some fairly exhaustive research into the way humans hear over the past 70 years. The curves themselves are essentially upside-down frequency response plots. If you take any given curve and invert it, keeping the 1 kHz. point at the same given sound pressure level, you have the frequency response of an average human’s hearing at that particular sound level. Note that the lows are rolled off seriously, the extreme highs cave in as well, and there is a response peak in the 3-4 kHz. range. These inverse curves are shown below for the sound levels of 70, 80, 90 and 100 dB SPL.

	Frequency response curves (by octave) of human hearing for sound levels at 1 kHz. of 70, 80, 90 and 100 dB SPL.

Now if you set up a response curve with an equalizer, and change its level with an attenuator, the response curve remains the same while the level changes, so that at different levels all the contours would be parallel. Us humans don’t do that with our hearing. The frequency response of our hearing changes as the level changes, in seriously audible ways! Take a careful look at Figure 4. At 4 kHz., for example, at 100 dB SPL our hearing is up 11 dB, while at 90 dB SPL it is only up 7 dB. Meanwhile, at 63 Hz., we’re 7 dB down at 100 dB SPL and 14 dB down at 70 dB SPL.

If you were to listen to these response curves, using a graphic EQ, you’d notice serious changes in musical quality. Figure 5, below, shows the difference between 100 vs. 90 dB SPL, and, more importantly, 70 vs. 90 dB SPL. If you mix at 90 dB SPL and listen at 70 dB SPL, that 90 vs. 70 curve is the equalization curve you’ll have applied to the recording: down 7 dB at 30 Hz., 4 dB at 63 Hz., plus a little wiggle at 4 and 8 kHz. Very, very audible!

	Two response curves that each show the approximate difference between equal loudness contour hearing response curves at various levels: 100 vs. 90 dB SPL and 70 vs. 90 dB SPL. These curves show enough variance that they would represent grossly unacceptable amplifier performance and considerably less than excellent loudspeaker or microphone performance.

There are two reasons I belabor this. First, when you are mixing, if you change the monitor level, you are changing the equalization applied to the monitor system (which includes your ears)! My recommendation is that you do all of your active mixing and decision making at a single, predetermined level, to reduce confusion, fatigue and the possibility of extra errors in your work. Second, your listeners will almost definitely play back your recordings at different levels than you mixed them at. You must check the mix at those levels to make sure that it sounds OK at the different equalizations those levels bring with them.

With all that said, we can now talk about how to manage mixing levels. First off, you need to maintain a proper working level to tape or hard drive. That is established with the master fader. Assuming VU meters, probably your pop, jazz or rock mix should cruise between -6 and 0 VU, with the loudest points getting up to +2 VU.[1] There’s much more to talk about here, and we’ll tackle that next month. Meanwhile you’ve got to translate that mix level into a relevant and meaningful playback level.