What is: DTMF?
DTMF stands for Dual-Tone Multi-Frequency. This is the technical term for the sounds you hear when you press the keys on your telephone. DTMF does more than just let you know that you pressed a key.
Each key has its own sound. Each sound is actually two tones played simultaneously. One of these tones is a low-frequency tone; the other is a high-frequency tone. There are four low-frequency tones which correspond to each row of keys, and three high-frequency tones which correspond to each column. Because each key uses two tones, the sound is almost impossible to reproduce with the human voice, so there is little chance that your voice can be interpreted as DTMF.
Though rarely seen on a phone, there is actually a fourth high-frequency tone that defines a fourth column to the right of the 3-6-9-# column. This column’s keys are labeled A, B, C, and D. These DTMF sounds are used primarily by the military and within telephone networks.
DTMF can be sent over phones lines, over the internet, or through two-way radios. The sound is interpreted on the receiving end and can be used to remotely control or interact with devices such as computers (through an IVR system) or answering machines.
Over regular phone lines, DTMF is sent as audio. VoIP phones are designed to carry voice only, but can send DTMF as a data packet over the internet. This data packet contains a close match to the DTMF but not an exact representation. Sending DTMF over VoIP may create either a duplicated digit or no digit at all on the receiving end. If any issues such as echo or packet loss occur during the transmission, the receiving end may not have enough data to rebuild the keypress combination.
Speech recognition technologies are an alternative to DTMF that allows more flexibility for user input over a telephone. It’s not a perfect substitute, though. Speech recognition must rely on a list of possible responses, including allowances for mood, dialect, pronunciation, background noise, homophones, and speech tics. DTMF, on the other hand, has 16 distinct tones that can be combined to send information reliably, albeit in limited permutations.