SAM: Software Automatic Mouth

Software Automatic Mouth — commonly known as SAM, or sometimes misspelled as "System Automatic Mouth" — is one of the most important programs in the history of personal computing. Released in 1982, SAM (Software Automatic Mouth) gave home computers the ability to speak for the very first time, turning typed text into synthesized speech on machines with just 64 kilobytes of memory.

Mark Barton and Don't Ask Software (1982)

SAM was created by Mark Barton and published by Don't Ask Software (later renamed SoftVoice, Inc.) in 1982. At a time when speech synthesis required expensive dedicated hardware, Barton achieved something remarkable: a complete text-to-speech system that ran entirely in software on the Commodore 64. The program occupied roughly 10 kilobytes of machine code — a testament to the elegant efficiency of its design.

Don't Ask Software, based in Los Angeles, marketed SAM as an affordable alternative to hardware speech synthesizers that cost hundreds of dollars. The software retailed for around $59.95 and came with a companion program called "Reciter" that converted standard English text into the phonetic notation SAM required. The company later transitioned to SoftVoice, Inc. and continued developing speech technology for various platforms throughout the 1980s and 1990s.

How SAM Works: Formant Synthesis

SAM uses a technique called formant synthesis to generate speech. Unlike modern concatenative or neural TTS systems that stitch together recordings of human speech, formant synthesis builds sounds from scratch using mathematical models of the human vocal tract.

The process works in two stages. First, the Reciter module converts English text into a sequence of phonemes — the basic sound units of speech (like "TH", "AE", "S"). Then, SAM's synthesis engine takes these phonemes and generates audio by simulating the resonant frequencies (formants) of the human mouth and throat. This is why SAM exposes four parameters: Pitch controls the fundamental frequency of the voice, Speed controls how fast phonemes are produced, and Mouth and Throat adjust the shape of the simulated vocal tract, changing the character of the sound.

The result is SAM's characteristic robotic voice — clearly artificial, but impressively intelligible for a program running on 1980s hardware. The formant synthesis approach is what gives the Commodore 64 SAM voice its distinctive mechanical quality that has made it beloved by retro computing enthusiasts.

From Commodore 64 to the Web

After its initial release on the Commodore 64, SAM was ported to several other popular home computers of the era:

Sebastian Macke's JavaScript port is particularly notable because it doesn't just imitate SAM — it faithfully reproduces the original synthesis algorithm, meaning the output sounds identical to the 1982 Commodore 64 version. This is the engine that powers the SAM voice generator on this page.

SAM vs Microsoft Sam: What's the Difference?

Despite sharing the "SAM" name, Software Automatic Mouth (1982) and Microsoft Sam (2000) are completely different programs with no technical connection:

The original SAM predates Microsoft Sam by nearly two decades. While Microsoft Sam became famous through Windows XP and internet memes, SAM (Software Automatic Mouth) holds the distinction of being one of the first software-only speech synthesizers for personal computers. If you're looking for the Microsoft version, try our Microsoft Sam Text to Speech generator.