Dictation Machines: What They Do and Who Still Uses Them

A dictation machine records your speech so that someone else, or speech recognition software, can turn it into text later. The recording and the typing are separate steps. Today's machines are digital handheld recorders, desk microphones with record and rewind controls, foot pedals that control playback, and workflow software that moves files from the person speaking to the person typing. Physicians and lawyers are the best-known users, because they need a document produced from speech while their hands and attention are on something else. For personal writing, a phone or laptop microphone with live speech recognition now covers most of the same work.

A dictation machine records speech so someone or something can type it later

The defining feature is the split between recording and writing. The speaker records a note, and a person or a program produces the text afterward. A dictation machine therefore has a microphone, storage and playback controls, but nothing that types on a screen. The broader practice is covered in what is dictation.

The idea is older than the electronics. Thomas Edison expected business dictation to be the phonograph's main job, more than entertainment ¹. For most of the twentieth century the workflow was the same: business people spoke into a machine, and a secretary typed the document from the recording ². That two-person workflow is the model that modern transcription software and foot pedals still follow.

Three things separate a dictation machine from an ordinary recorder:

1. A close microphone that favors the speaker's voice over the room.
2. Controls built for stopping, rewinding and marking files without looking at a screen.
3. An output meant to be transcribed, which means files that can be sent, queued and played back by someone else.

A phone can record audio and can do the first two, but the file handling around the recording is what a dedicated device adds. The terms are explained in the dictaphone and digital dictation glossary entries.

Digital handheld recorders replaced tape and still serve doctors and lawyers

A digital handheld recorder saves each dictation as a file. The file can be copied to a computer, labeled, sent to a typist and played back at a chosen speed. Tape cassettes had to be physically handled and mailed or passed along, so the change from tape was a change of workflow as much as of technology.

Manufacturers often save dictation in a compressed format designed for the purpose. Philips says its DSS format was designed for digital voice recorders and gives a high compression ratio with little loss of quality ³.

The table describes typical designs, not every product. Some law firms have moved this workflow onto phones, and the legal dictation software guide covers the tools used for that work. A 2014 BizTech Magazine article described a Pennsylvania firm where ⁴. The report describes one firm, so it shows that the workflow exists, not how common it is.

Doctors still dictate too, though the tools have changed. The evidence on that is in the section on live recognition below.

Desk dictation microphones put record, play and rewind under your thumb

A desk dictation microphone puts the record and playback controls on the microphone itself. Philips sells the SpeechMike line, and its SpeechExec guide for a SpeechMike with a slide switch describes the controls. On a SpeechMike with a slide switch, you start recording by moving the switch to REC ⁵.

On that model, the steps for a first recording are:

1. Make sure the docking station of the SpeechMike is connected to your computer, as Philips' SpeechExec guide instructs before you start.
2. Move the slide switch to REC. The recorder window appears, and a light glows red or green while recording, depending on the mode.
3. Move the switch to STOP to end the recording. For a short pause, use STOP and then move the switch back to REC to continue.

Button models use a different layout, and the rewind and fast-forward controls vary by model. Check the manual for your exact device before following a guide. The SpeechMike question explains the device in more detail, and the dictation microphone guide covers what matters when you choose one.

The foot pedal lets a typist control playback with both hands on the keys

The foot pedal is the piece that makes transcription practical. A typist types while the pedal controls the audio, so the hands stay on the keyboard. Philips describes its pedal as one that starts, pauses, rewinds and fast-forwards audio with a slight movement of the foot tip ⁶.

The AS-2700 kit from OM System paired its software with a foot switch of three pedals ⁷. The AS-2700 foot switch is HID keyboard compatible and can be customized with a free Windows configuration tool ⁸. Because a keyboard-style device sends ordinary key presses, software that accepts keyboard input can use it for playback. The pedal controls playback only. The recording itself is stored on the recorder or the computer.

Check availability before you plan around a kit. The AS-2700 Transcription Kit is no longer available as a bundled product ⁹, so check which parts you can still buy one by one before you plan a setup. The foot pedal question explains the device, and the dictation foot pedal glossary entry defines the term.

Workflow software routes recordings from author to typist and back

A dictation machine on its own only stores recordings. The workflow that makes it useful is software that moves each file from the author to a typist, tracks its status and returns the finished document. The sequence is the same whether the typist works in the same office or at a third-party company.

Each handoff in that chain is a place where a recording can be delayed or lost. A clinic or firm that adopts a dictation machine should check four things before it buys: whether the recorder's file format is supported by its transcription software, whether files reach the typist without a manual step, who sets the priority and marks a recording as finished, and how the files are protected in transit and storage. The last point is covered at the end of this guide.

Transcription is a recognized occupation. Medical transcriptionists typically need postsecondary education that leads to a certificate ¹⁰, according to the U.S. Bureau of Labor Statistics. The median annual wage was $40,410 in May 2025 ¹¹. That is the median for the occupation as a whole, so a typist in a hospital or a third-party company may earn more or less.

The workflow is also changing. The same federal projection says employment is projected to decline 4 percent from 2025 to 2035 ¹². Even so, about 6,200 openings a year are projected over the decade ¹³. The projection does not say why jobs are falling. The decline is consistent with speech recognition replacing some typing work, but that is an interpretation, not a finding of the projection.

A phone or a laptop microphone now covers most personal dictation

For personal writing, you do not need a dedicated device. Speech input on a phone is much faster than typing on its keyboard. In a controlled study of short messages on an iPhone, English text entered by speech reached 153 words per minute ¹⁴, against 52 words per minute on the built-in keyboard ¹⁵. Speech also produced fewer errors during entry than the phone keyboard did, but it left slightly more errors in the final text ¹⁶. That trade-off is why editing after dictation matters, and it is covered in the dictation mistakes guide.

The built-in tools are often enough. A dedicated recorder helps most when the work involves a file that someone else will transcribe, when you need a physical button to control playback during a task, or when the workflow depends on a format the software expects. If you only want to speak a draft and fix it on a screen, a phone or laptop is the simpler choice. The online dictation page works in a browser, and audio to text transcribes an existing recording.

Yap, which we make, is an on-device dictation app for Mac and Windows, and it is one way to get live dictation without buying a device. You can see current options on the download page.

Live speech recognition has replaced the recorder for many, but not all, users

The recorder is not the only way to get a document from speech. Live speech recognition turns the voice into text as the person speaks, and in some settings the document is ready much sooner than it would be after a transcriptionist types it. The clearest evidence comes from radiology, where reports are time-sensitive. At a 150-bed community hospital, speech recognition cut the median report turnaround from 24 hours to 1 hour ¹⁷. The hospital had no radiology trainees, and the authors summarize the change as a 24-fold improvement in median turnaround ¹⁸.

The error picture is less settled. A 2010 MRI audit compared 50 voice-recognition reports with 50 transcriptionist reports for two radiologists, and the voice-recognition reports had more errors ¹⁹. That audit covered two radiologists and 50 reports of each type, a small sample, so it does not give a general error rate. The reasonable reading is that error rates depend on the system, the vocabulary and the review step, and that a measured result from your own workflow matters more than a general claim.

Physicians show a related pattern in their notes. The Epic analysis covered ²⁰, observed from September 2020 to May 2021. Physicians who mainly used transcription and dictation averaged 4.8 minutes per visit on notes ²¹, while those who mainly used copy and paste averaged 13.1 minutes ²². In the adjusted comparison, dictation users spent 1.30 minutes less per visit on notes than SmartTools users ²³. The chart below adds the approximate SmartTools average ²⁴ to the comparison.

These are associations from observational data. The study shows that dictation users spent less time on notes, not that dictation caused the difference, because physicians choose their methods and specialties differ. Note time is also only part of the picture. A 2016 time and motion study observed 57 physicians for 430 hours ²⁵. Physicians spent 27.0% of their office-day time with patients and 49.2% on EHR and desk work ²⁶. The study covers a set of practices, so the figures may not describe every office.

Live recognition and the recorder do not have to be alternatives. A clinic can let speech recognition produce a first draft and keep a transcriptionist to correct it, or a physician can dictate into a recorder on the way to a visit and let a typist finish the note. The right mix depends on how urgent the document is, how accurate it must be, and who is responsible for signing it. The medical dictation software guide covers the software side in more detail.

The table below sets out who still relies on dictation and what the evidence supports.

Documents that contain patient information need protection in transit and at rest. The HIPAA Security Rule lists encryption and decryption of electronic protected health information as an addressable implementation specification ²⁷. Because the rule uses the word addressable, encrypted dictation files should be part of any clinic's review. This is general information, not legal advice.

Where to go next

• The overview is dictation and voice typing, and the broader term is voice typing.
• For how dictation compares with keyboard entry, read dictation vs typing and speaking vs typing speed.
• For the history of the recorder and its successors, read history of dictation.
• For the medical software used in clinical dictation, read medical dictation software.
• For the questions people ask most, see can you use a voice recorder for dictation and is there a device that types what you say.