I like my mids … or why sound personalization isn’t that personal

Sound Personalization App Screenshot

I recently bought the new Beyerdynamic Lagoon ANC headphones. They come with MOSAYC sound personalization. This feature is advertised as follows:

Our sense of hearing is unique and changes over the course of our lives. Like a mosaic, some pieces of the sound picture get lost or they fade. The unique MOSAYC sound personalization by Mimi Defined™ compensates exactly this development and adjusts itself precisely to your hearing ability. With individual sound personalization, the missing tiles of the mosaic are restored. And you can enjoy the full sound picture in all its colourful splendour.

https://north-america.beyerdynamic.com/mosayc

Although I really like speakers and headphones with absolutely no sound “enhancements” turned on (DELL and HP laptops, anyone?), I was intrigued by the idea of getting my very own personal sound, so I gave this feature a try and this post is about what I have found out.

Please bear in mind that this post is NEITHER about the quality of the headphones themselves nor its Active Noise Cancellation (ANC) feature. I personally think they are both great. This post is only about the sound personalization feature.

Beyerdynamic Lagoon ANC

After installing the Beyerdynamic MIY app, you can enter your date of birth and take a hearing test that tests how well you hear certain frequencies: 250 Hz, 500 Hz, 1 kHz, 2 kHz, 4 kHz and 8 kHz. It does that by plaing beeps through the headphones and measuring how soon and how long you are able to hear them. You are supposed to push a button labelled “I HEAR THE BEEP!” as long as you can hear them. I think I’ve put up a good fight.

Sound Personalization Feature
I’m taking a Hearing test…

After taking the test I’ve enabled the feature via pulling the slider all the way up from “No adjustment” to “Full adjustment”.

Turning it up to eleven…

How did it sound?

At the first moment, I was impressed but then I realized what happened to me. That same old trick again: The music was louder than before! Significantly louder!

Why was I impressed then?

There are these things called equal-loudness contours (or Fletcher–Munson curves as they are used to be called). These curves show how humans are less sensitive to low and high frequencies in contrast to mid frequencies. A low frequency (let’s say 50 Hz) needs to be significantly louder in order to be perceived as loud as a higher frequency (1 kHz).

https://en.wikipedia.org/wiki/File:Lindos1.svg

These curves also show something else:

As the volume increases, this perceived loudness difference between the frequencies diminishes. […] [A]s we raise the levels, the low and high frequencies become more prominent in the overall mix, leading to increasing power (lows) and clarity (highs). This makes the music appear more appealing to our ears and thus we often feel that ‘louder is better’. 

https://www.surfacedstudio.com/blog/music-production/louder-is-better-the-fletcher-munson-curves

Turning music louder increases your perception of lows and highs in contrast to mid frequencies. With louder music, the low frequencies don’t need to be so much louder than the mids anymore, but the playing device has made them a lot louder nonetheless.

This makes louder music sound more intense, more vibrant, and more “clearer and fuller”, as mastering engineer Ian Shepherd puts it. Increasing the volume is like turning on an imaginary filter in your ear. On the other hand, when reducing the volume, the music is perceived as duller. Let me make this even more clear with this super-scientific figure:

Louder is misperceived as better

I think that a lot of us (including myself) are (at least initially) tricked by this feature just because the music has become louder. Even audio professionals can be tricked by a level mismatch when comparing two signals, that is why level matching is absolutely crucial in the professional audio world.

I’ve then put a microphone in front of the headphones, played some white noise and recorded what came out of it to visualize this difference in volume:

Don’t let making things louder fool you.

But isn’t there anything else going on with the sound?

Yes, there is. We can measure it. For that, I will be using white noise. White noise is a signal having equal intensity at all frequencies, giving it a constant power spectral density. So it is ideal to be used for measuring frequency responses.

Note that the measurements I have taken are far from professional. They do not give any information on the perceived frequency response or even general sound quality of the headphones. But they give a clue about what is going on with the signal before and after the sound personalization adjustments. After level matching the signal with no adjustment and the signal with full adjustments, here are the two measured spectrums:

The two signals after level-matching
Blue graph: Signal with no adjustments
Reddish graph: Signal with full adjustments

Here’s the picture in original size.

When comparing the two signals after being level-matched, we see that the sound personalization feature attenuates the mids from approx. 300 Hz to 1 kHz and emphasizes the highs from about 10 kHz to 20 kHz. In colloquial German audio slang, we would call this a typical “Badewannen-Filter” (bathtub filter), because the frequency response of this filter brings the form of a bathtub to mind.

These measurements are consistent with my hearing impression. After turning sound personalization on and turning the volume down, the music sounds just a bit brighter. Too bright in my opinion. As I said, I like a frequency response which is as neutral as possible. For this sound to achieve, I could have used the conventional equalizer of my phone. So I will keep this feature turned off and enjoy music on these very good headphones without sound personalization.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.