← Back to Livestreams
Livestream

Neurofeedback & Chill: Biohacking Sleep

Andrew Hill, PhD

Sleep drives memory consolidation, mood regulation, and cognitive performance, and most people I work with are leaving a lot of sleep quality on the table. This article comes from one of my Monday night Neurofeedback & Chill livestreams, where I run a neurofeedback session on my own head, explain what I am training and why, and then teach the brain-based and behavioral tools that actually move sleep. I have spent about 25 years doing neurofeedback and reading brain maps, and I want to walk you through the circuits involved and the tools you can use tonight.

What does neurofeedback train when the target is sleep?

I ran two short protocols on myself during the stream, and each one targets a specific circuit involved in sleep.

The first was a beta protocol at C3, the left precentral gyrus. This left-side motor strip handles voluntary action and stabilizes vigilance. It keeps your attention sharp and crisp when you want it on. The same circuit also stabilizes the off state. C3 beta helps you stay awake when you are awake and stay asleep when you are asleep. One mechanism here is that boosting left-side beta tone suppresses the micro-sleeping that shows up in the background when someone is dysregulated. By creating that alertness signal during waking hours, you sharpen the flip between on and off, which supports sleep maintenance at night.

For that protocol I inhibited 4 to 7 Hz theta, rewarded beta in the mid-teens (I set my reward window just below 15 Hz, around 14.875 to 17.875 Hz), and inhibited 24 to 36 Hz on the high end. Theta takes the brakes off the system, so excess theta on the left produces impulsivity. Rewarding the beta band while holding theta and high beta down trains the stabilizer.

I deliberately left alpha out of the training band. When you are working on sleep, you generally do not want to inhibit alpha as a first move in a region. Alpha rebounds. Inhibit 4 to 10 Hz and you might get a nice boost in focus and mood at C3, but alpha surges back later and lightens your sleep. Go after alpha later, after a region is regulating in other ways. If you want the longer picture on this band, I cover it in Decoding Alpha Waves: Your Brain's Idle and Its Brakes.

Why is SMR the core frequency for sleep?

The second protocol was at CZ, the vertex, rewarding sensorimotor rhythm. SMR is the calm cat frequency. If you have watched a cat sit on a windowsill with a liquid body and laser focus, perfectly still and perfectly attentive, you have seen SMR. It is the ability to still the body and stabilize concentration. SMR sits in a beta-frequency range (roughly 12 to 15 Hz on the sensorimotor strip), but it behaves like a calming, regulatory rhythm rather than active cognitive processing. SMR only occurs on the sensorimotor strip. The same 13 to 15 Hz signal elsewhere is regular beta or fast alpha.

When you are asleep, SMR is called sigma, or sleep spindles. The brain uses these spindles to both initiate sleep and maintain it. A car goes by, the brain starts to wake, and a burst of SMR keeps you asleep. That happens right at the vertex. So my CZ protocol is a mix of executive function and sleep support.

SMR also drives memory consolidation. Spindle bursts in the cortex couple to hippocampal sharp-wave ripples, and that coupling looks like the process of moving memories from short-term into long-term storage (Siapas & Wilson, 1998). SMR is also where neurofeedback started: training it up suppresses seizures, which is how the field was discovered around 55 years ago (Sterman & Friar, 1972). Because the same thalamocortical circuits generate both waking SMR and sleep spindles, training SMR during the day often improves both daytime focus and nighttime sleep, an effect reported in healthy subjects (Hoedlmoser et al., 2008). This is also why SMR training has such a strong impact on attention. The mechanism is well supported in the research literature and is detailed in SMR Neurofeedback: Train Sleep, Focus, and Self-Control.

How do you find your personal SMR frequency?

You cannot see your SMR peak reliably by intending to sit still. The practical method is to estimate it from your alpha. Look at your eyes-closed alpha at PZ with linked ears. The peak lands somewhere near 10 Hz. Whatever distance your alpha sits from 10 is roughly the distance your SMR sits from 13.5 Hz. If your alpha runs 9.5 Hz, your SMR is probably around 11.5 to 14 Hz. When in doubt, train SMR at C4 rather than CZ, and undershoot your native SMR frequency a little. That produces a calming effect and is less likely to trigger a trauma response. A high alpha sitting at the vertex is often a sign of anxiety, or of beta slowed by fatigue, and rewarding low beta there can trip into fast alpha and provoke that anxiety instead of calming it.

If you want to understand what neurofeedback can and cannot do more broadly, see Is Neurofeedback Legitimate? A Research Overview and the cost breakdown in How Much Does Neurofeedback Cost in 2026?.

What is the single strongest cue for setting your sleep clock?

Meal timing. Food is a powerful exogenous cue for your circadian rhythm, and feeding time can entrain peripheral clocks independently of the light-driven master clock (Damiola et al., 2000). The human body can shift about one time zone per day comfortably, which is why an hour of screens late at night is not the disaster biohackers claim. That hour pushes your rhythm forward by roughly an hour, and your body absorbs that slop.

My first rule: eat in the time zone you want to live in. You can also use this for travel. Three days before a flight, start sliding your meal times toward the destination time zone, and you can land in Europe from the US with almost no adjustment.

The mechanism behind eating before bed deserves attention. Melatonin suppresses pancreatic insulin release (Peschke et al., 2013). As your evening melatonin rises, insulin falls, and falling insulin creates the urge to snack. Resist it. Go to bed with moderate to low blood sugar and you get a strong growth hormone pulse a couple of hours after sleep onset. Go to bed full and fed, with blood sugar above 35 or 40, and you blunt that pulse. You skim the surface of sleep, run hot all night, and wake up unrested. Go to bed fasted and you wake lean and energized. If you are new to fasting, keep this modest: a couple of hours without food before bed is enough to clear blood sugar. Long fasting windows you are not adapted to will spike cortisol, which feels like energy and keeps you awake. For the broader picture on meal timing, see Strategic Fasting: Time-Restricted Eating for Metabolic and Cognitive Health.

When does light actually matter for sleep?

Morning, specifically the window from about half an hour before sunrise through an hour after. There is a blue cast in the air during that period that hits the suprachiasmatic nucleus, which sits on top of the optic chiasm. Its job is to sample the color of light and act as a cascading clock reset across the brain, locking your internal clocks to the Earth's photoperiod.

Evening light color barely matters. Blue blockers are mostly sizzle, no steak. The research I have seen says intensity, not color, drives how much you push your circadian rhythm late in the day. To protect your sleep after dusk, skip bright overhead lights. Use desk lamps instead of ceiling fixtures, dim the TV, and you have handled the real circadian concern. I will go further: I cannot tolerate blue blockers, and several clients report headaches or migraines within minutes of putting them on. That artificially narrowed light input behaves like a motion-sickness migraine trigger for me. If you have research showing a real circadian benefit from blue blockers, send it my way.

A gentle morning walk or some yoga burns off the cortisol and blood sugar burst that wake you without calling for more energy. A three-hour run or CrossFit session first thing does the opposite, so keep hard training out of that window. The minimum effective version of all this is laid out in Biohacking Your Morning: The Minimum Viable Practice for Circadian Health.

Which supplements actually support sleep?

A small handful earn their place.

Magnesium. Many people are deficient, and correcting that makes a real difference. Magnesium glycinate is bound to glycine, an amino acid that supports sleep on its own.

A three-part timing stack, taken an hour before bed. A third of a milligram of melatonin (300 micrograms), about 150 mg of L-theanine, and 300 to 500 mg of GABA. These support the timing of sleep rather than inducing drowsiness, and after a few days the urge to sleep starts building more reliably at the right hour. Theanine is gabaergic, and although oral GABA crosses the blood-brain barrier poorly, the theanine-plus-GABA combination has been reported to outperform either compound alone on sleep measures in animal models (Kim et al., 2019).

Melatonin dosing is where most people go wrong. A low dose around 300 micrograms is enough to raise plasma melatonin into the normal nocturnal range and shift circadian timing (Zhdanova et al., 1995). Take more and your blood level overshoots without a matching benefit; you just get long lingering effects. Use it for timing, not as a sleeping pill.

For cortisol, especially in midlife and older. ZMA lowers cortisol and nudges testosterone up, though the zinc can cause vivid, angry dreams in a few people; drop it if that happens. Ashwagandha and phosphatidylserine (PS, often labeled PS 100 or PS 150) are reliable cortisol reducers. Ashwagandha lowers cortisol in stressed adults (Chandrasekhar et al., 2012), and phosphatidylserine blunts the cortisol response to stress (Hellhammer et al., 2004).

Peptides. DSIP (delta sleep-inducing peptide) is one of the first compounds that actually induces deep sleep. It is mildly invasive to use and is a tool to consider only in severe, chronic insomnia and under appropriate medical supervision, not a casual addition.

One caution: sedating antihistamines like diphenhydramine (Benadryl) are common in biohacker medicine cabinets. The research shows a dose-dependent association between cumulative anticholinergic load and later dementia risk (Gray et al., 2015). The histamine and cholinergic systems are linked, and suppressing histamine too hard appears to stress the cholinergic system tied to memory. Sedating antihistamines and prescription or over-the-counter sleep medications cause sedation or a hypnotic state, not sleep; sleep may then take over, but these compounds do not generate it.

Should you trust the REM numbers on your sleep tracker?

No. The REM data from wearables is misleading and inaccurate. REM is like blood pH: it self-regulates within a narrow range. By the time REM is genuinely impaired, a person is hallucinating, psychotic, or severely depressed. It does not slowly drift; it falls apart as someone becomes mentally ill. You are almost certainly dreaming every night even if you do not remember it, and forgetting dreams usually means you did not get enough deep sleep to consolidate the memory.

Deep sleep is the number worth watching. Trackers like Oura and Whoop measure deep sleep reasonably well. Aim for about 25% deep sleep as an adult, closer to 30% for a kid, and push that number up. Add mouth taping to encourage nasal breathing, which improves oxygenation and reduces snoring, and a bed chiller or cooling pad to keep body temperature down overnight.

What about night owls, and how much sleep do you really need?

I do not buy the larks-and-owls framing. The research is mostly theory. Some brains are well entrained to the circadian rhythm and some are not. A person with ADHD who feels brilliant at 10 PM is often not well entrained, and the late-night productivity comes from a quiet environment rather than a natural peak. Humans adapt to large circadian shifts, but only when the signaling is consistent.

Entrainment changes how much sleep you need. My own pattern: if I let bedtime drift to midnight and rise at 6 or 7 AM, I need a full eight to eight and a half hours and still feel underslept. If I go to bed early and rise at 4 or 5 AM, I need about six hours, feel excellent, and get an hour and three-quarters of deep sleep out of it. Dial in your timing and each sleep hour does more work. I also do my hardest cognitive work in the quiet 4 AM to 8 AM window, and it beats the 8 PM to midnight version on every measure.

One thing not to tinker with on your own: apnea. If you snore heavily, wake gasping, or your partner hears you choke or stop breathing, get a sleep test. Rule out the serious airway problems first, then work on the environment, then biohack the rest. For related circuit work on the stress side, see Biohacking Fight or Flight: Mastering Your Stress Response, and if daytime fog is the main complaint, Biohacking Brain Fog: Restoring Mental Clarity.

Where to start tonight

Stop eating a couple of hours before bed and let blood sugar drop. Get outside into morning light within the first hour after sunrise. Take a short, gentle morning walk. Drop overhead lighting after dusk. Watch your deep sleep percentage, not your REM. If you snore or wake gasping, get a sleep study before anything else. Layer in magnesium, a low-dose melatonin timing stack, and SMR neurofeedback if you want to train the underlying circuits directly. If you want a map of your own executive function, sleep, and stress patterns, a QEEG is the place to start, and I walk through what it shows in QEEG Brain Mapping: What It Is, What It Shows, and What to Expect.

References

  1. Siapas (1998). Coordinated interactions between hippocampal ripples and cortical spindles during slow-wave sleep. doi:10.1016/s0896-6273(00)80629-7
  2. Sterman (1972). Suppression of seizures in an epileptic following sensorimotor EEG feedback training. doi:10.1016/0013-4694(72)90028-4
  3. Damiola (2000). Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus. doi:10.1101/gad.183500
  4. Gray (2015). Cumulative use of strong anticholinergics and incident dementia: a prospective cohort study. doi:10.1001/jamainternmed.2014.7663