Log a hard hearing moment with a measurement of the sound, and your practice sees it as it comes in.
A hearing companion that links patient and practice. When hearing is hard, the patient logs the moment in HearLab Log: the situation, a rating, a note, and a measurement of the last 15 seconds of sound, taken on the phone. The sound itself is thrown away. After explicit consent the moment goes to the patient’s own practice, stored in the EU, and shows up in the practice dashboard. Non-medical: HearLab observes and surfaces, the clinical decisions stay with the professional.
Tap New moment where hearing is hard, pick the situation, rate it and save. The phone measures the last 15 seconds (level, background, how well speech stands out, room, a prominent tone, the spectrum) and then throws the sound away. Linked with a code from your practice and after your explicit consent, the moment goes to your practice. You can withdraw at any time, and your shared moments are then deleted.
A non-medical companion. Designed alongside audiological care, never replacing it.
How this is measured▸ Press Start listening for captions and the level meter; when a moment is hard, log it in HearLab Log and your practice sees it.
Wind and traffic are where hearing aids struggle most: the automatic scene classifier keeps changing its mind, so the device keeps re-balancing. That “program hunting” makes a drive or a windy street exhausting.
This runs a real, transparent classifier on your mic or a synthesised scene, and puts a naïve switch-on-every-frame version next to our hysteresis-smoothed one. Same recognition; far fewer switches.
Non-medical. This demo’s audio never leaves the browser, and for the mic we deliberately turn off the browser’s own noise suppression, so the classifier hears what you hear. Live captions in the companion are a separate feature and do use your browser’s speech recognition.
The patient logs a hearing moment where it happens. The dashboard shows it to the practice as it comes in: rating trend, when and where it gets hard, at what speech-over-background margin, which frequency bands the background takes, and for each moment the full measurement from MixLab, VoiceLab, SignalLab and NoiseLab, with CSV and JSON export for the practice's engineers.
Below is the real dashboard in example mode: the measurements come from real recordings (a phone in a room and test scenes with speech, TV, air handling and music), the patients, ratings and notes are made up. A patient's moments arrive only after that patient gives explicit consent in HearLab Log. No audio is stored, the data stays in the EU, and it is deleted after the practice's retention period.
Test it yourself · both sides, about five minutes
Only the patient side? Open /hearlab on your phone and choose “No code, try it first”: your moments stay on the phone.
In HearLab Log the patient taps New moment where hearing is hard, picks the situation and how it went, and saves. The phone measures the last 15 seconds (level, what is in the background, how well speech stands out, room, a prominent tone) and then throws the sound away: only numbers are kept.
Linked with a code from the practice and after explicit consent, each moment goes to the practice dashboard: the rating, the situation, the note and the measurement, per patient, as it arrives. The consult starts from what actually happened instead of what the patient managed to remember.
Measurements and ratings across many moments could one day help tune hearing devices or the software that drives them. That is not what HearLab does today: the data is used by the practice only, and any research use would need its own, separate consent.
HearLab closes the gap between what the patient experienced and what the professional sees: the patient’s in-context reaction plus a measurement of the sound at that moment, logged when it happens instead of recalled weeks later. Explicitly non-medical: HearLab shows, the professional interprets. No audio is stored, the data stays in the EU, and the patient can withdraw at any time.
Also in HearLab
Patients as pseudonyms, their moments as they arrive: the rating, the situation, the note and the measurement, with export and deletion per patient. Stored in the EU (Frankfurt) under a data processing agreement, deleted after the retention period the practice picks.
Speech to text through your browser’s speech recognition, in the language you choose, so you can follow along without asking anyone to repeat.
See how loud the room actually is while you listen, so "it was too noisy" becomes a number you can point at.
Is HearLab a medical device?
No. HearLab is a non-medical companion. It does not diagnose, treat, or fit hearing aids. It records what you experienced at a moment (your rating, the situation, a note) with a measurement of the sound around you, so your hearing care professional has real context. The professional decides.
Does HearLab record audio?
No sound is stored or sent. When you log a moment, HearLab Log keeps the last 15 seconds in the phone’s memory, measures them (level, background, how well speech stands out, room, a prominent tone, the spectrum) and then throws the sound away. Only those numbers are kept. There is no location.
Where do my moments go?
Without a link, they stay on your device. Once you link with your practice and give explicit consent, each new moment goes to your practice, stored encrypted in the EU (Frankfurt) by AudioLab.tools on the practice’s behalf, and deleted after the practice’s retention period. You can withdraw in the app at any time: sharing stops and your shared moments are deleted at the practice. Live captions are separate and come from your browser’s speech recognition, which may send audio to an online speech service.
Continue the workflow
MixLab’s LUFS metering helps you hit the loudness consistency that benefits assisted-listening setups.
Before publishing voice content, run it through VoiceLab to check the dimensions that matter for hearing-aid users.
SignalLab’s region indexer surfaces the segments that need caption coverage in long-form recordings.
CueLab routing maps make the hearing-aid output bus an explicit destination in your show plan.
SkillLab challenges train listeners on intelligibility under noise, which is the dimension that matters most clinically.
MixLab, VoiceLab, and SignalLab hand audio off to one another without re-uploading. When you finish an analysis in one of those, the → Lab buttons in the demo chrome carry your decoded audio directly into the next lab’s analysis flow. The other labs open fresh.