how to (and how not to) add scent to VR and immersive experiences
Over the last five years of running Scentient, we’ve spoken to a lot of people who had already experimented with adding scent to immersive experiences, using all sorts of different technologies, but found that the result did not quite work as expected. That led us to ask a broader question: when scent does not work well in an experience, is it actually a hardware problem, or is it an experience design problem? Spoiler alert: the answer to this question led us to launching Scentient Studio.
The hardware for adding scent already exists across a huge range of scales. At one end, you have highly personal systems such as Scentient Escents, designed to deliver scent close to an individual user. At the other, there are large room-scale and HVAC-based systems capable of scenting entire venues. In between, there are standalone scent machines, directional systems, embedded emitters and completely manual effects such as scented objects, sprays or incense.
So producing a smell is not usually the difficult part. The challenge is that scent does not behave like the media we are used to designing around. With audio, you can stop driving a speaker; with visuals, you can turn the pixels off. Once you release scent, however, you have physically introduced molecules into an environment. They move with the air, disperse, mix with other smells and take time to clear.
That makes scent a different medium that follows a different set of design rules, and once you start designing around those rules rather than treating scent as another audiovisual output, it becomes much easier to use effectively.
Start with the experience, not the scent machine
This is the approach we take with Scentient Studio. When we work on an immersive experience, we don’t start by asking how we can fit our hardware into it, or which scent machine should be used. We start with the experience itself and work backwards towards the technology.
Sometimes a wearable device is the best solution. Sometimes a room-scale emitter makes more sense. In other cases, we might design completely custom hardware, use an existing third-party system or incorporate a simple practical effect. The aim is to choose the technology that fits the experience.
Before making that decision, there are a few questions we always work through.
What role does scent actually play?
The first question is not how to deliver the scent, but what it is actually supposed to do. Is it there to establish an atmosphere? Is it a specific narrative cue? Is it attached to an object the user interacts with? Should the participant consciously notice it, or should it simply make the environment feel more convincing?
This matters because scent is usually more effective as a highlight than as a constant stream of information. If every object, every interaction and every scene has its own smell, scent quickly stops adding meaning. It becomes the sensory equivalent of highlighting every sentence on a page: once everything is highlighted, nothing really stands out.
A few carefully placed scent cues can often contribute much more than trying to add smell to everything simply because the technology allows it.
Does the scent need to change?
A single persistent environment is relatively straightforward. If you are creating a projection-mapped forest and want the room to smell like a forest for the duration of the experience, you have a forgiving problem to solve.
Now imagine that the same experience moves from a forest, to starting a fire, to brewing coffee over that fire, all within five minutes. Suddenly the challenge is not just scent delivery, but also transition and clearance. What happens to the forest scent when the fire appears? How much smoke is still present by the time the coffee arrives? Do those smells work together, or does one interfere with the next?
This is where scent behaves very differently from a visual scene change. You cannot simply cut from one smell to another, and in many cases the best way to create a clean transition is to stop the previous scent earlier than you might expect. We call this scent storytelling.
Once a smell has established an environment, it does not necessarily need to keep being released throughout the entire scene. We can gradually reduce or stop it while the participant still associates that scent with what they are seeing. That gives the physical scent more time to dissipate before the next cue arrives. But how much time exactly…? That leads us nicely to the next question.
How precise does the timing need to be?
The amount of timing precision you need depends heavily on the experience. In an exhibition where visitors spend several minutes within the same environment, a scent arriving a few seconds later may make very little difference. In VR, the requirements can be much tighter. If you open a virtual door onto a burning room and only release the smell a few seconds later, the delay itself can become noticeable and undermine the connection between what you are seeing and smelling. But if you are in the virtual forest and are slowly starting a fire, a room-scale system might be a fitting solution.
It is also important to remember that scent timing is not simply the timestamp at which a device is triggered. There is a physical journey between releasing the scent and somebody perceiving it. Distance, airflow, device output, user position and even head movement can all affect that journey.
Synchronising scent with the audiovisual and tactile sides is therefore an experience-design problem in its own right, and something we will cover in more detail in a future post.
Who needs to smell it?
It sounds like an obvious question, it is the person going through the experience, of course; but it quickly becomes more complicated. Does one person need to smell it, or everyone in the room? Should several people receive the same scent at the same time? Could different participants need different smells? Are they sitting still, walking around or moving between physical interaction zones?
A solution designed for one seated VR user looks very different from a system designed for a room of 100 people. The same applies to localisation. In some installations, the goal is simply for everybody to experience the same atmosphere. In others, a scent might need to exist only around a particular object or in a very specific part of the room.
Where the user is matters too. A seated VR experience, a walk-through installation, an outdoor activation and a large indoor venue all create different constraints around airflow, positioning and how closely you can control where scent travels. Once those requirements are understood, the range of sensible delivery methods tends to narrow quite quickly.
More scent is not necessarily more immersive
One of the easiest mistakes to make is to increase the output whenever a scent is not immediately obvious. But stronger is not always better. More scent is harder to clear, more likely to interfere with whatever comes next and, at some point, starts competing with the rest of the experience and taking attention away from the other senses.
Subtle cues can often work better because they allow the user to remain focused on what they are seeing and hearing. There are absolutely situations where a strong, punchy smell is exactly the effect you want, but intensity should be a deliberate part of the experience design, and not “let’s crank up the intensity so that they definitely smell this”.
In general, we prefer to start with less and increase it where necessary. It is much easier to add more scent than to deal with an environment that has already been oversaturated.
Think about what happens after launch
The requirements also depend on how the experience will actually be used. A one-off prototype, a three-day brand activation and a permanent installation may use similar scent technology, but they have different operational requirements.
Maintenance is one of the areas that is easiest to overlook during the exciting part of designing an experience. If there are six scents and one runs out, can you replace just that one? How do you know when it is running low? Does the system need cleaning or calibration? How much staff involvement is required? If the experience runs 100 times in one day, does it perform the same on the hundredth run as it did on the first?
These are not necessarily the most exciting design questions, but they are often what separates an impressive demonstration from an installation that can reliably operate day after day.
Choosing the delivery method
Only once we understand the experience, we start deciding how the scent should actually be delivered.
Practical effects are still very useful. Scented objects, sprays, incense or manually presented scents can work extremely well for prototypes, performances and one-off experiences. They are difficult to automate and do not scale particularly well, but low-tech does not mean ineffective. Sometimes the simplest way of testing whether a scent cue actually adds something to an experience is also the best one.
Ambient scenting is at the other end of the spectrum. Conventional room diffusers and HVAC systems are excellent when the goal is to create a persistent atmosphere across a larger space, particularly when lots of people should experience the same scent. Their limitation is rapid change: once you have scented a large volume of air, that scent cannot simply disappear in a few seconds.
Localised and directional systems sit somewhere between the two. By controlling airflow, emitter position and where the visitor stands, we can create scent around a particular interaction zone without requiring the user to wear anything. This can work particularly well for physical installations, although it makes the airflow within the space and the movement of visitors an important part of the design.
Personal scent delivery, including systems such as Scentient Escents, becomes particularly useful when individual control, timing and rapid transitions matter. Instead of trying to change the scent concentration throughout an entire room, we only need to create the required concentration close to the participant. That means much less scent needs to be released into the wider environment, which in turn can make transitions considerably easier to manage, we’re talking scent change in a matter of 0.25 seconds, and not 2.5 minutes.
The same principles can also be built into completely custom hardware. An emitter might be hidden inside a prop, integrated into a wall, triggered when somebody picks up an object, or incorporated into an installation so that the participant never knows where the scent is coming from. In many cases, the best scent technology is the technology the user never notices.
Where scent experiences tend to go wrong
Most disappointing scent experiences we see fall into a handful of recurring patterns, and many of them come back to treating smell too much like audio or visuals. A trigger is assumed to mean instant perception, and stopping the emitter is assumed to mean the scent has disappeared. In reality, neither is true.
Using too much scent is another common problem, as is failing to think about how one smell will layer on top of another. The next scent does not arrive into a clean slate; it arrives in whatever olfactory environment the previous scenes have already created. This is why we think about scent experiences as “painting” and layering rather than “editing” and cutting.
It is also easy to choose a piece of hardware first and then force the experience around its capabilities. That can work, but it often leads to compromises around the number of scents, timing, positioning or clearance that would not have been necessary if the delivery method had been selected later.
Finally, there is a temptation to use scent everywhere simply because it is available. If six different interactions happen in one minute, they probably do not all need six different smells. Scent works best when it has a reason to be there. Like a highlighter, its value comes partly from being selective.
Scent follows different rules
The interesting thing is that many of the characteristics that initially look like limitations can become useful creative tools once you start designing around them.
It can remain after something has disappeared visually, creating continuity between scenes. It can precede a visual cue and suggest that something is nearby before the participant sees it or hears it, or operate much more subtly in the background and help someone recognise a place or situation without explicitly telling them what has changed.
Those properties make scent very different from sound and visuals, but that difference is what makes it interesting. What are often described as the limitations of scent are really a different set of design rules, it’s just that we are much less accustomed to working with them.
At Scentient Studio, our process reflects that. We start with what the audience should experience and where scent can contribute something meaningful. From there, we define the timing, audience, space, transitions and operational requirements, prototype the simplest version possible and test what people actually perceive. Only then do we settle on the delivery architecture, integrate it into the wider experience and test the complete system in the environment where it will actually be used.
The technology matters, of course, but good scent experiences rarely start with a scent machine. They start with understanding what you want somebody to experience, and then choosing the right way to make that happen.
