Rescuing Smartphone Audio with Digital Soundstage Widening

It wasn’t too long ago that portable/ mobile audio meant blurry playback on a Walkman, and portable movies were, of course, completely unthinkable. Yet today we consume portable audio in all kinds of environments and for a range of different purposes— for music, games, audiobooks, GPS systems, audio assistants like Siri… the list goes on. And it’s all delivered in the convenient package of a smartphone. For the longest time portable audio also necessitated the use of headphones, and this is still very much the case. The tiny speakers in smartphones, while significantly better than ten years ago, can only do so much. Sure, advanced signal processing has enabled improvements in output level and sound quality over the past few years, but watching a movie on a smartphone without headphones remains a less than massive audio experience. What’s the problem?

Let’s take a look at the current state of things. 

The industry has already made a couple of attempts to manufacture phones intended for audio-minded people. I remember having a Nokia phone with stereo speakers somewhere back around 2005 and, sure enough, it had two speakers. But no one would call it stereo. Today, we see phones with two speakers from time to time, and the speaker units themselves are vastly improved compared to my old Nokia phone, but they are still too closely spaced to give any kind of stereo feeling. The speakers in such phones are rarely identical for space and/or cost reasons, and their mounting may be asymmetrical as well. It is not uncommon to have one speaker aimed along the normal of the screen and a second one aimed perpendicular to it—not an ideal situation from an audio perspective. On the other hand, it seems highly unlikely that someone would want a smartphone large enough to fit a pair of speakers that are spaced far enough apart to achieve good stereo imaging.  

One common proposal for addressing the lack of stereo perspective from closely spaced speakers is a trick called crosstalk cancellation. It was proposed and investigated as early as the 1960s by Schroeder and others. The idea is to exploit how our brains interpret signals from our ears to determine the location of signal sources. A fair amount of research has been done in this field, but the short of it is this: we have two ears with a head in between them. It should be clear that the ratio and timing of what enters a set of ears will help the brain determine the location of the sound source. If we can control how a sound source propagates to each ear individually we can also give the impression that the sound comes from the direction of our choosing.

The problem with all attempts to date is that they assume a set of equidistant speakers that act as point sources. Under this assumption, controlling the ratio between the left and the right ear of a listener in a known location is just a geometrical problem. Unfortunately, reality tends to defy assumptions like these.

Consider the smartphone again. I have yet to see a phone with two identical speakers. It’s rare enough to find one with two speakers that point in the same direction, and they’re not acting as point sources either. The mismatch between assumptions and the real world results in a significantly lacking audio experience. Previous attempts have often been criticized for their “phasy” sound, unstable sweet spot, and lack of a stable center image. 

So, what to do?

We decided to approach this same problem, but from a different direction. 

With Dirac Panorama Sound technology, we exploit the same crosstalk cancellation principle, but we don’t assume anything. Rather, we leverage our experience accurately describing an acoustic system’s properties from measurements, and from this information we formulate an optimal solution for each individual system. That is, we don’t try to make a generic solution that fits all systems with a pair of speakers. Instead, all devices implementing Dirac Panorama Sound have their own unique set of parameters derived from that hardware configuration. And the result is apparent from the performance!

A phone equipped with Dirac Panorama Sound has a very wide sound stage across all frequencies. Voices panned to the left will, very clearly, come from the left, far outside the physical boundaries of the phone itself. It is very hard to describe with words exactly what it sounds like, or what the experience is. It’s almost uncanny. This fact is evident whenever someone picks up my phone equipped with Panorama Sound—it’s been dropped in surprise any number of times! People are simply not prepared to hear sounds coming out of thin air, and that is truly what it feels like.

Just as the camera-equipped cellphone was a novelty of little use some years ago, a smartphone with two speakers won’t do wonders for your audio experience. That is, until today. My firm belief is that Dirac Panorama Sound will make stereo speakers a must-have for smartphones in these next few years. That’s how good it is.

But rather than taking my word for it, I suggest reading what others have to say. That or getting your hands on a phone equipped with Dirac Panorama Sound. They should start hitting the market any time now, starting with the Xiaomi Mi6. ☺

– Jakob Ågren, Product Manager at Dirac Research