Closed-loop vs open-loop audio

If you read only one page on this site before evaluating anyone's claims — including ours — make it this one.

Two completely different technologies

Open-loop. The audio plays a specified rhythm. It does not know anything about you. A 10 Hz session is 10 Hz whether you are calm, agitated, asleep or absent. Every consumer audio product works this way, ours included.

Closed-loop. Something measures your brain in real time and the stimulus adapts to it — arriving at a particular phase of your own ongoing rhythm, or changing when your state changes. This requires EEG hardware and a feedback algorithm.

These get discussed as if they were the same field. They are not remotely.

Why it matters so much

Because the most impressive results in this literature are closed-loop, and they are routinely cited to sell open-loop products.

Two examples worth knowing.

Pink noise and memory. In 2017 Papalambros, Zee and colleagues enhanced slow-wave activity in older adults and improved their word recall the next morning. The system read each participant's EEG continuously and delivered sound at a precise phase of their own slow oscillations. The timing to the individual was the mechanism. Play the same pink noise on a schedule and you have removed the part that worked.

Sleep onset. Alpha phase-locked stimulation has been reported to shorten sleep onset by around ten minutes in chronic insomnia — again, locked to the individual's own rhythm. Meanwhile, stimulation delivered on a fixed schedule while participants were still awake delayed sleep onset. Same sound. Opposite result. The difference was whether it was listening.

The honest position for an open-loop product

We are open-loop. So what can we legitimately claim?

The stimulus is real and well-characterised. A tone modulated at a steady rate produces a measurable, sustained auditory response. That does not require any feedback loop and is not in dispute.

The behavior is real. A session with a defined length gets people to sit down, start, and stop. That is not a neurological claim and it is probably the largest practical effect.

The state-matching is plausible but unproven. That a specified rhythm nudges you toward a corresponding state is the contested claim, and open-loop is the weaker version of it.

What we cannot claim is any result that depended on the loop being closed. When you see a product citing the pink-noise memory study or phase-locked sleep work, check whether their product does phase-locking. Almost none do.

Would closed-loop be better?

Probably, for some applications — and it is where the field is going. But it is worth being clear about the costs.

It requires wearing EEG hardware, which for sleep means wearing it all night. Dry consumer sensors are workable for spectral measures and considerably noisier than clinical systems. And brain data is unusually personal, so where it is processed and who keeps it becomes a real question rather than a theoretical one.

We deliberately do not review or recommend specific headsets. Specifications and prices change constantly, we cannot independently verify manufacturers' accuracy claims, and a sound company ranking neurotech hardware is a conflict of interest wearing a lab coat. If you go looking, the questions worth asking are: how many channels and where, is raw data processed on the device or uploaded, and can you export your own data.

How to read any claim in this field

Four questions, in order:

1. Was it closed-loop? If yes, it does not support an open-loop product.
2. Was it audio alone? The 40 Hz gamma work is largely audiovisual. Light plus sound is not sound.
3. Was it humans? A great deal of the mechanistic work is in mice.
4. What was actually measured? "Increased slow-wave activity" and "felt more rested" are different claims with different evidence.

Applied honestly, those four questions dissolve most of the marketing in this category — including, where it applies, ours.