Scentisound is a synesthesia-inducing speaker.

Scentisound is a normal speaker in every way, with the exception of an integrated scent-spray in addition to low, mid, and high frequency speaker-cones. The speaker has a trigger frequency switch in the back, and once the trigger frequency is selected, the speaker starts following the beat, and sprays the scent with the beat.

Intended Usage

I envision Scentisound being used in advertising perfumes in environments where music can be used to market the product as well.

Secondly, it can be used to recreate the atmosphere of live venues at home by inducing synesthesia using custom made scents for venues. ie: a jazz bar atmosphere at home.

Initial Prototype

For the first prototype, I preferred not to hack the circuit together, but rather placed a piezo microphone under a low frequency speaker, and had the speaker spray the scent each time it picks up a threshold signal from the piezo. I’ve hacked the spray canister, and created an initial prototype to showcase the proof of concept. It’s important to keep in mind that this was just to see if the idea is viable, and it was.

The first prototype could follow the sound with a ~50ms. delay, which wasn’t perceivable as a delay, and approximate delay between the trigger and fire was ~500ms. Meaning, it can catch up with beats that are as fast as  ~120BPM, (~2 sprays/second)


Second Prototype

After running into a few problems with the first prototype, I’ve realized it takes time to spray the odor using it’s original canister, and because I’ve hacked it together for the first prototype it’s the main reason why the spray time is slow. For the second prototype I thought it would be better to use my own mechanical trigger to spray the canister using a solenoid, which I think is going to be able to catch up with beats that are 180BPM. (~3 sprays/second) Realistically, this is extreme, and I don’t think it’s a good idea to spray even twice a second. However it was important for the purposes of experimentation, that I see how far I could push the limits.


It took longer to listen to the speaker using my own listening circuit with the piezo, and although it’s a really low latency time, it could be dramatically improved up to 10ms. by making my own amp drivers for the speaker – and the effects of the latency would be impossible to perceive. So I’ve tried my best to create a small trigger circuit and tried implementing a custom driver for the speakers to do faster FFT(fast-fourier-transform) processing.

Currently the box has an unbalanced set of speakers in it, which creates a bass-heavy listening experience. Although this could be beneficial for some genres, it’s not ideal for many. For an actual/final product, this would be different.


Thanks to Aisling Kelliher for her amazing guidance.

Thanks to Mirna Pibernik for the music. Lights (Ellie Goulding Cover).