Can I dictate in a whisper with EnviousWispr?

Yes. EnviousWispr is built to capture quiet and whispered speech, including soft words at the very start of a sentence that used to get dropped. You can speak in a low voice at night or in a shared office and still get an accurate transcript. It runs on-device on Apple Silicon, so your audio never leaves your Mac.

Why did my first word sometimes get cut off before?

Voice detection trims silence so the transcription engine only hears speech. If your first word was very soft, the old behavior sometimes mistook it for silence and trimmed it away. EnviousWispr now recognizes when a real but quiet opening word is about to be discarded and keeps it, so a soft 'Actually' or 'Overall' stays in your text.

Do I need to change a setting to capture quiet speech?

No. This is the default behavior for everyone. There is no whisper mode to turn on and no sensitivity slider to tune. The app handles soft and far-away speech automatically.

Is whispered dictation private?

Yes. EnviousWispr captures audio, detects speech, and transcribes entirely on your Mac. Nothing is sent to a server, so whispering at midnight stays between you and your keyboard.

dictationmacosprivacyaccessibility

Dictate in a Whisper: Capturing Soft and Quiet Speech

EnviousWispr now captures quiet, whispered, and far-from-the-mic speech, so you can dictate at midnight or in a shared office without raising your voice.

June 7, 2026

You learn a specific kind of quiet when other people are around. The near-whisper next to a sleeping partner. The low murmur at a shared desk when you do not want the person across from you to hear your half-formed draft. The careful voice in a quiet train car.

Dictation is supposed to help in exactly those moments, and for a long time it did the opposite. You would lower your voice, and the words would simply not show up. So you either spoke louder than felt comfortable or gave up and went back to typing.

EnviousWispr now captures that quiet speech. You can whisper, speak softly, or sit back from your mic, and the words still land.

Why quiet speech used to disappear

Before any words reach the transcription engine, the app has to answer one question many times a second: is this sound speech, or is it silence? That step matters. It trims the dead air so the engine only works on the parts where you are actually talking, which keeps things fast and keeps stray room noise out of your text.

The trouble is that a whisper sits much closer to silence than a normal speaking voice does. When the bar for “this is speech” was set for ordinary talking, a soft voice fell underneath it. The app heard you, decided it was probably just quiet room noise, and trimmed it away. You got a short transcript, or nothing at all.

The most frustrating version of this was the clipped first word. You would start a sentence with a soft lead-in, “Actually, we should,” and the opening word arrived a beat before you were fully up to volume. The detector treated that soft start as silence to trim, and your sentence began in the wrong place.

What changed

Two things, both working underneath without anything for you to configure.

First, the app is better at recognizing soft speech as speech. The detection that decides what to keep no longer assumes you are talking at a normal volume, so faint and far-away words make it through instead of being thrown out as noise.

Second, there is a safeguard for that soft opening word. When the app is about to trim what looks like silence from the very start of a recording, it checks whether it is actually about to discard a real, quiet word. If it is, it keeps it. A gentle “Overall” or “Actually” at the top of a thought stays where you said it.

Together, that means the things that used to vanish (a whispered sentence, a soft start, a word spoken while you leaned back in your chair) now stay in your transcript.

Nothing to turn on

This is worth saying plainly: there is no whisper mode and no sensitivity dial to find. We deliberately did not add one. A slider that asks you to guess the right setting for your room is a setting that is usually wrong, and we would rather the app just handle it.

So the answer to “how do I turn on quiet capture” is simple: you do not. Hold your hotkey, speak as softly as the room asks for, and release.

Where this helps

Late at night, with people asleep. Draft the email or the message in a near-whisper without waking anyone.
In a shared office or open plan. Keep your voice down so your neighbor is not pulled into your train of thought, and still get clean text.
On a quiet train or in a waiting room. Speak at the volume the space expects rather than the volume the software used to demand.
When you simply do not feel like projecting. Some days you do not want to talk loudly. Now you do not have to.

Because all of this runs on-device on your Mac, the quiet stays quiet in the other sense too. Your audio is captured, analyzed, and transcribed locally. Whispering at midnight does not send anything to a server.

A note on honesty

We will be candid: this is the kind of improvement that should have been the behavior all along, and for a while it was not. Quiet speech getting dropped was a real gap, and the fix was less about adding a feature and more about the app finally hearing you the way you expected it to. If you tried dictating softly in the past and it let you down, it is worth another try.

What changes for you

Nothing in the steps. Hold the hotkey, speak, release, and the polished text lands in your clipboard or pastes where you are working. What changed is the volume floor. The soft, the distant, and the whispered now make it in.

How EnviousWispr works, end to end. Where your audio goes, and why it never leaves your Mac.
Why Mac dictation keeps stopping, and how to fix it. The other reason dictation cuts out, and what to do about it.
macOS dictation that works offline and stays private. The fully-local setup.

If you have not tried it yet, download EnviousWispr free. Then turn your voice down as far as the room asks, and see what stays.