How Skin Resonance reads a study
Every evidence section on this site, from the science page to the free tools, is written to the same standard. This page sets it out, so you can see what a claim here rests on and hold us to it.
Skin Resonance sells a €12 app, and the person writing its evidence pages is the person who built it. The only sound way to run that is to say exactly what each study was and let you weigh it, which is what the rules below are for. They apply to the science page, the journal, the concern pages, the free tools and the FAQs, and they have since September 2026, when every article on the site was rewritten to them.
What we record about every study we cite
A study is named on this site with its design, its size, what it was compared against, what it measured, the size of what it found and its main limits, and a link that opens the paper or its abstract. So the June 2026 trial on the science page appears as a randomised controlled trial of 66 students with poor sleep, four weeks of thirty-minute listening before bed, a 3 Hz binaural beat against pink noise and against nothing, with self-rated stress and sleep lower in the beat group, no difference in morning cortisol, and the participants not blinded, and those details are the study. A sentence that says only that research shows binaural beats reduce stress is not a citation, and we do not write it.
Where a study has a conflict of interest we know of, it is named in the same breath: the 2022 Schumann device trial is described with the fact that an author holds the patent and that the groups did not differ significantly, and three 40 Hz studies were kept off the site in October 2026 because their authors owned the company or the patent, or the funder made the product being tested. Sources are opened on the day of writing and the date is recorded at the foot of the page.
How much each kind of study is allowed to say
A systematic review or meta-analysis of randomised trials is the strongest evidence we have, and when one exists it leads the section and its own caveats are repeated, as with the 2024 review that found consistent cortisol and alpha-amylase effects and still called the overall evidence promising but equivocal. A single randomised trial is described with its size and its blinding, and a trial of thirty people is called a trial of thirty people. A pilot, a crossover in nine people or a three-night study is called early, and nothing is built on it alone. A study in cells or in mice is described as what it is, with the sound pressure or dose, and is never allowed to stand in for a human result; the 2025 Kyoto cell study is on the site because it is interesting, with every page that mentions it saying that its sound levels were far above anything from headphones. An observation, a survey or a company’s own measurement is reported as that, or left out.
The chain this product rests on has three links, and they are not equally strong. That stress and cortisol affect skin is established dermatology and is described as such. That deep sleep is when growth hormone is released and collagen is made is well documented and is described as such. That binaural beats lower the stress response is supported by small trials and reviews, and it is described as early and encouraging. The skin outcome at the end of the chain has not yet been measured in a trial of any sound practice, and a page says so once, where the chain is drawn, then moves on to what the practice does deliver.
The words we use, and the ones we avoid
Early research that points the right way is described as early and encouraging, and its limit is named once, in plain words, and then left alone rather than repeated in every paragraph. We do not write that a frequency heals, repairs, boosts or rebuilds anything, because none has been shown to, and we do not dismiss other apps, practices or products, because the people who use them are the people reading this. Where a popular claim is wrong, the 7.83 Hz heartbeat of the planet or a collagen frequency, the page says what the real finding was and where the claim went past it, and ends on what the practice does deliver.
What we never publish
No testimonial, review, quote or buyer story appears on this site unless a named real customer gave it, and rating schema waits until there are real ratings to put in it; as of this review there are none. No scene, anecdote or statistic is invented to make a point. No number that should have been measured is estimated. Routine lengths, prices, what the app does and the number of sounds in it are taken from the app’s own code at the time of writing, and are corrected on the site when the app changes.
How a page is checked, and how to tell us it is wrong
Before a new page is published, a second reader who did not write it goes through it twice, once as a sceptical visitor deciding whether to pay twelve euros and once as a fact-checker with every source open, and what they find is fixed first. Pages carry the date they were last reviewed. If you find a claim that a source does not support, a number that has drifted from the paper or a study that has since been retracted, write to [email protected] with the page and the line, and it will be corrected and the correction noted on the page.
Where to see the method applied
The science page carries the studies behind each link in the chain, with a line on what each measured. Can sound actually improve your skin? is the long answer to the question the whole site turns on. The four free sound tools each end with an evidence section written to the rules above.
Skin Resonance is a wellness web app, not a medical device, and does not diagnose, treat or cure any condition. For a diagnosed skin condition, please see a dermatologist.