- Stereo width is an illusion that we either capture in a recording or create within a mix
- To make wide sounding mixes, capture width when you record or create width inside your mix
What is stereo width?
Stereo width is an illusion of the left-to-right dimensions of the sound field (i.e. sound stage or panorama) in a recording, perceived by a listener.
Imagine yourself standing on a sidewalk listening to busy downtown traffic. You are hearing cars constantly rushing back and forth, left to right and right to left. This “field of sound” you are hearing is quite wide. Because we hear binaurally (another future topic), you’re hearing a three-dimensional world of sounds from left to right, front to back, and even up and down. But let’s simplify for a moment and focus on the fact that cars are coming from one direction or the other and we largely are hearing that back and forth movement across the sound field because we have two ears and we can sense where each of the sounds are coming from.
Imagine that you capture a few minutes of what you are hearing on the sidewalk with a stereo audio recorder, go back to a quiet place to listen to what you recorded. Setting aside mic quality, technique and other factors for the moment, you will hear a rough representation of what you experienced on the sidewalk because you will have captured the scene in stereo. Your recording will reveal how the traffic moves from left to right and right to left across the sound field captured by the recorder. Show that recording to someone else, who wasn’t there when you recorded it, and they will be able to imagine to a degree what it was like standing on the sidewalk because the recording has enough information in it to recreate a panorama where they can hear the traffic moving back and forth, similar to how you experienced it.
Now imagine if you use a mono (i.e. monophonic or monaural) voice recorder to capture what you heard on the sidewalk, instead. You will hear a much different representation, because your recording is missing a massive amount of information. A mono voice recorder only has one microphone (i.e. one ear) so it cannot capture stereo information. Your recording will reveal a mush of engine noise, maybe the odd vehicle horn, but everything is crammed together and it sounds like varying levels of noise. Show that recording to someone else, who wasn’t there when you recorded it, and they will find it challenging to imagine being on that sidewalk. The resulting panorama from the mono recording simply does not contain enough information to do that.
Stereo width therefore depends upon stereo information being captured and presented back to the listener. When you think about that traffic scene, what makes the scene stereo? It is all about differences between what you hear from one side of the scene to the other. Stereo width is that simple. You create stereo width by creating or enhancing the difference between the left and the right sides of the sound field that you are presenting back to your listeners.
One more example before we move on. Picture yourself sitting in front of a stage with two musicians standing side by side, a few feet apart. Forgetting the room acoustics and other factors for a moment, you focus on the two musicians, and from your vantage point, they are relatively close together. You therefore perceive them both coming largely from the center of the sound field. Now imagine that each musician walks to the opposite end of the stage. Now you perceive them as being distinctly separate where you hear one largely off to the left, and the other way off to the right. The musicians have just “widened the panorama” presented to you simply by separating themselves further, relative to you. Now imagine replacing yourself with a stereo array of microphones. See how microphone placement relative to the subject(s) can make a big difference in this situation? You can drastically alter what you present to a listener – which impacts their perception of width – simply by changing the placement of the subject(s) relative to the microphones.
This is how stereo width within a mix works. When you create a mix, you are building a sound field with layers of sound. Stereo width is an illusion of the left-to-right dimensions of the sound field. The more you can create differences between what you present on the left versus what you present on the right, the wider your mix will seem to a listener.
How to Create wide Stereo mixes
- You can only enhance stereo width if it exists to some degree in your mix
- Make your mix sound wider using differences in left vs right gain, time, pitch, tone or polarity
Mono sound is single-channel sound, and stereo (stereophonic) uses two channels. If you reproduce monophonic sound over two loudspeakers, you will hear the sound coming from a narrow area between the two speakers (in the center). This is called a phantom image because it can be quite palpable to a listener. It can sound almost as real and distinct as a sound that comes from just the left or the right speaker on its own. The phantom image sounds focused (narrow) because each loudspeaker is presenting the same information, and there is nothing that our brain can use to establish a difference between sounds coming from the left versus the right loudspeaker.
Does your mix sound too narrow? Width is a key component of modern mixes because most music is mixed in stereo. Many listeners listen in stereo (e.g. with earbuds). We perceive a mix as being too “narrow” when we sense that the music is coming from the center when it seems that it should not. That is to say, a center-centric mix can be perfectly fine if it sounds natural, but you know you have a problem when you or your listeners sense that things are too narrow, where the mix sounds too confined and the panorama is not expansive or convincing enough to draw them in.
You can make your mixes sound wider by maximizing the difference between what you present in the left versus the right channel. The “difference” can be in time (arrangement of notes or elements, and includes phase as well), gain, pitch (arrangement of notes and/or tuning), tone, or polarity.
Let’s look at gain first. Panning is used to adjust the “gain difference” between left and right. Pan hard left, and you have full gain in the left, and no gain in the right. This is the easiest way to maximize left-to-right difference. Take two different sounds in your mix, pan one hard left, the other hard right, and you have maximum separation, maximum difference between left and right. The more you “pull them in” by panning closer to center, the more you reduce difference information, bringing them closer and closer to perceived center.
You can use timing to create left-to-right difference by delaying one channel relative to the other. You can take a mono string section track in the left channel, add a 25ms delayed version to the right channel, and with some gain adjustments you have created a very wide, very stereo sounding string track. If you experiment with this, a 15ms delay is a good starting point. As you pull it towards 0ms, you will hear the delay getting harder to discern and other things start to happen as the sounds fuse together, which can be good or bad. I recommend listening in stereo for the effect you want but then listen in mono to make sure you haven’t created a mono-compatibility issue in the process. If you hear the tone change drastically where it is clear in stereo but muffled in mono, that is an issue. Also listen to the attack portions of notes. Sharper attacks (guitars, vocals etc.) will only tolerate shorter delays because if the delay is too long you’ll get a flam effect which can be distracting to a listener. Sounds with softer or slower attacks will often work well with over 20ms of delay.
You can create tonal differences between left and right by equalizing each channel differently. A common technique with mono piano tracks is to EQ the left to have more bass and lower mids and the right to have more upper mids and treble so that there is a sense of movement from left to right as the player moves up and down the scale.
You can use tuning differences between left and right to create width. This is partly how chorus effects work. Take a mono track in the left, and then add a detuned version in the right channel to create width.
A mix engineer uses all of these methods different ways to either create or enhance stereo width. The ideas are almost endless. You can keep it simple or get really creative depending upon the situation. For example, putting reverb on one side of the field but not the other is often used on guitar tracks. Another option is putting vibrato on one channel and not the other, or mono chorus on one channel or the other. Hammond’s popular “Leslie” effect, captured in stereo, works effectively because it uses the Doppler effect which essentially creates timing and tuning changes between what’s captured in the left and right microphone.
MAKE WIDER MIXES BY MAXIMIZING CONTRAST BETWEEN MONO AND STEREO
This is absolutely key. To perceive width, we need a reference point. Making every sound in your mix super-wide will not necessarily lead to an engaging or musical mix. Certain anchor points in a mix that are kept central, such as a lead vocal, a bass guitar, a kick drum etc. are essential to creating a wide mix because we will judge the width of the panorama relative to those central elements! Never underestimate the power that mono, central elements have in enhancing the perceived width of your mix by how they create contrast.
WIDENING YOUR MIX WITH SPATIAL ENHANCERS AND STEREO WIDENERS
Remember: Width must exist within your mix before you can enhance it in a meaningful way.
Please note that I consider these next tips to be shortcuts. If your goal is to truly understand stereo width to create musically wide-sounding mixes, these shortcuts will mainly help you by allowing you to easily experiment. They are a time saver that can be useful in some mix situations after you have learned the concepts above.
Stereo wideners, ambience retrieval, or other forms of spatial enhancers have become very popular. They employ different combinations of the techniques above to broaden and deepen the sound stage. They are a quick fix. I caution against using them across a whole mix because that can greatly distort the sound field you have spent so much effort to build if you are not careful. I have lumped ambience retrieval in with stereo widening but they are often different processes, so research any processor first to understand what it will do to the sound. Use these products sparingly, and consider only using them on certain elements within your mix that would benefit from widening or ambience retrieval effects.
“FreeHaas” and “FreeOutsider” are free plugins offered by VescoFX. They are well worth experimenting with. I recommend using them sparingly, on one or two elements in your mix at the most. FreeHaas adds a Haas Delay (see Haas effect) which you can adjust to your liking. FreeOutsider is a much more obvious beyond-the-speakers effect. You can combine them with tone and gain adjustments to maximize the difference between left and right, thereby maximizing the width of certain elements of your mix.
Mid/Side (a.k.a. sum and difference) processors are widely misunderstood and are often thought of for stereo widening, but I caution that merely adjusting the ratio of Mid to Side will not work that well unless the source has a lot of difference information in it already. This is why many spatial enhancers include a Mid/Side adjustment control to allow you to adjust how much of the widening effect you hear, after the spatial processor has done the initial work of creating more difference information in the first place. If your mix has a lot of difference information, but does not sound wide enough to you, Mid/Side adjustments probably cannot help. If your source is completely mono, Mid/Side will do absolutely nothing and if it is close to mono, then boosting the side doesn’t do much either. I expand upon this much further in Part 1 and Part 2 of my Mid/Side articles.
Widening your mix beyond the loudspeakers
While much of the effort in a mix is in creating a wide field within the loudspeakers, you can create the illusion of going past them as well, to further extremes. There is risk to this. If you overdo it, it will lower the quality of your mix.
Take any sound and add it to two channels in your workstation. Pan one channel hard left, pan the other hard right. Press play and you’ll hear the sound coming from the exact center. Flip the polarity of one of the channels and listen to how the sound changes. It will sound strange – very wide depending upon your listening environment. The reason is because yet again, you have created another difference between left and right – this time it is a difference in polarity! This is a completely unnatural sound and for most people it is fatiguing after a while. But it is also completely incompatible in mono. Collapse your mix bus to mono, and everything disappears because the left and right channels cancel each other out. While absolute polarity differences like this are unnatural, they can also be useful. Experiment with them, particularly on occasional ambient sounds or even reverb effects, and with the other techniques that I mentioned above to help push the sounds a little further outside the speakers.
By combining various techniques, we can recreate more complex combinations of binaural cues that our brain uses to determine where a sound is located, allowing you to present a three-dimensional sound stage to your listeners. This is worthy of a separate article but here is a simple example. When we hear a sound coming from our left, our left ear hears it slightly louder, slightly sooner, and slightly brighter than our right ear does. Our brain uses that combination of differences in loudness, timing and tone to perceive where that sound is located within a three-dimensional sound field. Our brain analyzes that, along with the information about ambient environment (early reflections and the left-vs-right loudness, timing and tonal differences of sounds in the environment) to map out how large a space is and where the sound is located within it. The more location information we can present to our brain, the more we can create the perception of location be it within or beyond the loudspeakers.
It is not easy to recreate these types of cues over loudspeakers because of bleed (crosstalk). If you sit between any two loudspeakers, you will hear both loudspeakers with both ears to varying degrees whereas with headphones, each ear only hears one speaker. It is therefore much easier to recreate these complex cues when you are mixing for playback over headphones because you won’t have to overcome crosstalk. Creating these cues for headphones can be as easy as using a binaural microphone array to capture the source, and then working to preserve those cues throughout the production process. You can also add the cues after the fact with an HRTF (Head Related Transfer Function) processor.
When mixing for playback over loudspeakers, the only way to effectively recreate these cues is to find ways to minimize the perceived crosstalk. There are crosstalk cancellation technologies available such as QSound, Ambiophonics and BACCH that accomplish this to various degrees. I will plan expand more on this in a future article. These types of filters are well suited to situations where your listeners are sitting in a fixed position, in front of two loudspeakers which makes them well suited to gaming applications or small portable devices. BACCH might be the most promising from a pure performance perspective but unfortunately at the time of this writing it is priced outside of the range of most users and productions.
Outside of crosstalk, there are other challenges to reproducing a three-dimensional sound field over two loudspeakers. There is no way to know exactly where your listener will be placed, but we can be certain that most will not stay in the same position. We also cannot know what performance level their listening room and sound system is capable of, whether it is mono, stereo or surround. We can’t even guess at how it will be configured (EQ or tone, speaker placement etc.). Therefore, if you plan to build complex binaural cues into your mixes, carefully consider all of the possible end points (ear buds, loudspeakers, television, movie theater etc.), and how your music will be consumed (while gaming, while driving, while travelling, while housecleaning). In many cases, building these complex positional cues into your mix is only worth the effort if you can do so without reducing the sound quality for other listeners in the process. This is why checking for mono compatibility and loudspeaker vs headphone compatibility is so important.