Evidence, Limits
& Design
Phantas is built on real phenomena with honestly mixed evidence. This page separates what we use, what the research actually supports (and where it does not), what we do not claim, and how those limits shaped the product. Where findings are mixed, we say so.
What Phantas uses
Frequency-following response, with binaural and isochronic tones.
The guided structure of the session itself, not just the signal.
Photic stimulation (SSVEP), offered as opt-in depth.
A described-need prompt that suggests a session.
Nothing to picture — the session never asks you to visualize.
What the evidence suggests
Better supported
Photic driving / SSVEP. Rhythmic light produces a steady-state visually evoked potential — visual-cortex EEG that follows the stimulation frequency. In one foundational study, ten participants showed resonance responses around 10, 20, 40, and 80 Hz. (Herrmann, 2001; small sample.) This supports that light entrains electrical activity — it is not evidence of a specific cognitive outcome, and correlation is not induction.
Visual imagery varies between people. Imagery ability varies substantially and can be measured through both self-report and perceptual effects: people with little or no voluntary imagery are far less likely to experience vivid flicker-induced pseudo-hallucinations. (Königsmark, Bergmann & Reeder, 2021.) Prevalence estimates for aphantasia vary depending on how strictly it is defined — from well under 1% for a complete absence of imagery to several percent under broader thresholds. (Zeman et al., 2015; Dance et al., 2021.) This supports our "nothing to picture" design premise, not an efficacy claim.
Alpha and theta oscillations track attention and memory. A robust, highly cited correlation. (Klimesch, 1999.) This is a correlation of the brain's own rhythms with cognitive states — not proof that driving those rhythms from outside reproduces the state.
Promising but mixed
Auditory beat stimulation. Reviews and meta-analyses have reported modest positive effects from auditory beat stimulation across several outcomes, including cognition, anxiety, and pain. (Huang & Charyton, 2008, a positive review of 20 studies; Garcia-Argibay et al., 2019, a meta-analysis of 22 studies with a medium pooled effect, g ≈ 0.45.) But the studies are heterogeneous — differing by design, outcome, frequency, duration, and population — study quality is uneven, and the findings do not establish that Phantas's specific protocols will produce the same effects. Protocol-to-outcome mappings remain uncertain, which is why our audio sessions are described by intended use, not guaranteed effect.
Binaural beats as a perceptual phenomenon. The effect is real and perceptible, with an understood auditory mechanism. (Oster, 1973 — a foundational Scientific American article, not an efficacy trial; Moore, 2012.) That a beat is perceived is not evidence that it changes cognition.
Experimental
40 Hz gamma stimulation. Early 40 Hz research showed biological effects from visual flicker in Alzheimer's-model mice — reduced amyloid load and changes in microglia. (Iaccarino et al., 2016.) Later studies have explored additional modalities and human applications, but that work is separate, remains early and condition-specific, and some findings have been difficult to replicate. Animal amyloid results are not evidence of a cognitive benefit in healthy users; we cite this line of work as mechanism background only.
Gamma and conscious "binding." An influential but still-debated theory linking gamma-band synchrony to awareness. (Engel & Singer, 2001.) It does not establish that 40 Hz stimulation produces insight.
Flow on demand. Flow is a well-described experience (Csikszentmihalyi, 1990), and transient hypofrontality is a leading theoretical model of altered states (Dietrich, 2003, a hypothesis). Reliably triggering flow with entrainment is unproven.
Specific frequency-to-task mappings. Pairings like 10 Hz for steady work or 40 Hz for synthesis are design heuristics drawn from the oscillation literature — not validated induction-to-outcome results. A protocol frequency describes how a session is designed; it is not a measurement of your brain.
Reading the evidence honestly
- —Correlation is not induction. A rhythm that accompanies a state is not proof that driving that rhythm creates the state.
- —Entrainment is not a cognitive outcome. Measurable EEG following a stimulus is not the same as thinking, feeling, or performing differently.
- —A protocol frequency is not a measurement. The Hz label describes how a session is designed, not what your brain is doing.
- —General-technique evidence is not evidence for Phantas. Pooled effects across many studies do not validate our specific sessions.
- —Animal findings are not human efficacy. Results in mouse models do not transfer to healthy people.
- —Subjective and objective effects are different. Feeling more settled is not the same as a measured performance gain, and we do not treat them as interchangeable.
What Phantas does not claim
Phantas is a performance tool, not a medical device. It does not diagnose or treat any condition. It does not guarantee focus, flow, relaxation, insight, or a measured change in your brainwave activity. The Hz and brainwave labels describe how a session is designed — not a confirmation of what your brain is doing. If you are managing a neurological or psychological condition, talk to your clinician before using stimulation-based tools.
How the evidence shapes the product
- —Audio is the complete baseline. Because audio-only evidence is mixed, every session is complete with sound alone.
- —Light is optional and screened. It is the stronger-evidence channel but carries real photosensitivity risk, so it is opt-in and behind a safety screening.
- —Sessions are described by intended use. You choose by what you need to do, not by a promised neurological result.
- —Technical labels stay visible but modest. Hz and brainwave labels are shown for the technically curious, never presented as proof of an induced state.
- —“Nothing to picture” is a design principle. It reflects the imagery-variation evidence and our no-visualization design, not a clinical claim.
- —We surface limits. Mixed findings are shown as mixed; we do not cherry-pick positive studies.