How to Set Up Push to Talk for Dictation

Push to talk for dictation is four steps, and none of them happen on the hardware. Our pads, like the Voice Pad ($79), are a microphone and some keys that send ordinary keystrokes. The speech becomes text inside software you choose. Setting this up is mostly a matter of getting those two halves pointed at each other.

What do you need before buying anything?

Dictation software with a hold style hotkey. This is the step people skip, and it is the one that decides whether any of this is possible. Some dictation tools only offer a toggle: press once to start, press again to stop. A toggle cannot be turned into a hold by a keypad, because the keypad is only sending the same keystroke. Find the hotkey settings in whatever you plan to use, look for wording about holding a key, and confirm it before you spend money. We have not tested every dictation tool and cannot tell you which yours is.

Step one: choose the right microphone as your input

Plug in the microphone that came with the pad, then go into your sound settings and actually select it as the input device. This sounds obvious and it is the most common failure we would expect. Nothing tells you when it is wrong. You press the key, you speak quietly into the device in your hand, and the laptop microphone across the desk is the one recording, so the result is worse than not buying anything.

Step two: find the hotkey and write it down

Open your dictation software, find the hold to talk hotkey, and choose something your editor and terminal are not already using. Function keys and unusual modifier combinations are safer than anything a code editor might want. Test it on the keyboard first, before hardware is involved, so that you know the software half works on its own.

Step three: map that hotkey onto the pad

Now tell one key on the pad to send exactly that combination, using whatever remapping software your pad supports. This mapping is between you and the pad; the dictation software never learns that a pad exists, it just receives the keystroke it was waiting for. If your pad has spare keys, a good use for one is Enter, so that submitting what you dictated is also a press rather than a reach.

Step four: test somewhere harmless

Open a scratch text file, not a terminal with a live agent in it. Dictation types into whatever has focus, and a stray transcription landing in a shell is a command you did not mean to run. Hold the key, count to one, then start talking. If your first word keeps disappearing, that pause is the fix: the software needs a moment between the key going down and the audio it keeps.

What if your hands are busy?

Then hold the gate with your foot instead. The Pedal Three ($69) is three ribbed pedals on a curved steel base with a captive cable, and the Pedal Four ($85) is a wider flat plate with four. They send keystrokes exactly like the pads do, so step three is identical; only the body part changes.

If you are still deciding whether a held key is the right shape at all, the trade-offs are laid out in a push to talk key versus an always on microphone.