Most people know what touch tones are. However, fewer know that touch tones are a subset of a more general class of signals known as DTMF tones. Since the application recognizes not just touch tones, but most DTMF tones, it is named DTMF Decoder.
DTMF Decoder allows you to analyze a sound to determine whether DTMF tones are present, and if so, which ones. DTMF stands for Dual Tone MultiFrequency. DTMF tones are mixtures of two equal amplitude sinusoidal waves of differing frequency. Most of the telephone signalling we are exposed to is performed using DTMF tones. The program recognizes four types: touch tones, the dial tone, rings, and busy signals. Each is described later in this file.
IMPORTANT: This copy is partially disabled- it will not acknowledge 3 of the 16 touch tones. To obtain a full-featured copy, please register by sending $25 to me at this address:
Carl Huben
6929 Lee Crest Drive
West Bloomfield, MI
48322
Registered users will receive the latest (full-featured) copy of DTMF decoder, its documentation, and notification (either by mail or email, please specify) of future updates. I'll send updates to interested users by email (for free), or through the postal service if you send me a SASE. The Think C 5.0 source code is free upon request to all registered users.
Some features to be included in the next version:
• Touch tone decoding "on the fly" and in the background.
• Applescript awareness.
• Support for sound saving.
• Support for sound cutting and pasting.
• Recognition of user-specified tones
I will get to these features soon. However, when I get to it depends in large measure on the response I get from the Mac community. If a lot of people express interest (and pay the shareware fee -the applescript developer's kit is pretty expensive), I will get around to these upgrades sooner.
Thanks to everyone who's tested the program and made suggestions, but especially Rustle Laidman, for his endless testing and great suggestions.
I don't think the program requires much documentation. However, I should mention a few points:
*** Representation of Sounds / Making Selections
• DTMF Decoder uses a graph of the current recording's average amplitude versus time in the small rectangle immediately below the buttons. The graph's function is two-fold: it allows the user to 'see' where the tones are within the whole sample, and it also indicates what portion of the sound is selected for playback or analysis.
• To select a portion of the sound, position the I-beam cursor within the rectangle at one end of the selection, click the mouse button and, while keeping it down, move to the other end of the desired selection, then let go of the button. Your selection will be highlighted (the system highlight color is used- you can change this in the "Colors" control panel).
- NOTE: Double-click inside the rectangle to highlight the entire sample.
- NOTE: When a sound is imported or recorded, the whole sample is
automatically highlighted.
*** Recording Sounds
• Click the button with the red circle. Once you do this, the old recording (if there is one) is erased, and recording begins from the default sound input device. Recording continues until either the stop button (it has a square in the middle) is clicked, or the sound buffer is filled.
• You can choose to have the program warn you that it's going to overwrite the old sound by checking the "Record Protect" box in the preferences menu.
• You can choose to play the sound you're recording through the Mac's speaker by checking the "Play Through" box in the preferences menu.
*** Playing Sounds
• The portion of the sample you wish to play must be highlighted (the system highlight color is used- to change it, choose the "Colors" control panel). To play the current selection, click the button with the green triangle. Playing continues until either the end of the selection is reached, or the stop button is pressed.
*** Analyzing Sounds for Touch Tones
• To analyze the current selection, click the fourth button from the left (it has a wave followed by an arrow pointing to "123," and will be subsequently referred to as the transform button). The results are displayed in the text window (the large, intially black rectangle below the small window). If no DTMF tones are detected, "NO TONES DETECTED" will appear in the window. Analysis continues until either the end of the selection is reached, or the stop button is pressed.
• You may enable or disable the recognition of certain classes of DTMF tones by checking/unchecking the appropriate boxes, located to the right of the transform button.
• You can set the sensitivity of the tone detection algorithm in the preferences menu. It is better to record louder tones than to increase the sensitivity, since false positives can be a problem at the high sensitivity setting. To avoid false positives at the higher sensitivities, try turning off some of the tone switches. For example, if you know the sound only contains touch tones, uncheck the ring, busy, and dial tone boxes if spurious recognition occurs.
*** Changing the Buffer Size
• The size of the sound buffer can be changed in the preferences menu.
*** Changing the Volume
• The volume used when the application plays and records (if play-through is enabled) can be set to either an internal value or the system default from the preferences menu.
*** Miscellaneous
• The Command-Period key combination is equivalent to clicking stop button.
• Sounds recorded using DTMF Decoder cannot currently be saved. The ability to record sounds is secondary to the purpose of the program, and was included only for convenience. If you want to keep a sound that you wish to analyze with DTMF Decoder, record it using some other application (like the great public domain program, Sample Editor), and import it.
• Edit menu options apply to the text window's contents- you cannot yet copy sounds to or from this application.
• Sounds may be recorded or imported. Imported files must be in the Audio Interchange Format (this entails that they be not compressed). Also, they must have been recorded at 22254.5454 samples/sec. Theoretically, a sampling rate of only 3266 samples/sec or greater is necessary to recognize all common DTMF tones, but the program detects tones more reliably at the higher sample rate.
*** Problems
• If the program fails to recognize DMTF tones in a particular recording, it is probably because they aren't loud enough. Try increasing the tone sensitivity in the preferences menu, or record louder tones.
• If you have other problems, please contact me either by mail or email (addresses given below) with a *full* description of the problem. I'll try to fix it in the next version.
*** Okay...
If you have any bug reports, suggestions, and/or comments, I'd like to hear from you. I can be reached at the address given above, or on:
• Compuserve. My ID is 72157,3506.
• America Online: My user ID is CarlHuben
• The Internet: gsmattes@vela.acs.oakland.edu
***More information about DTMF tones:
Touch Tones:
There are 16 touch tones. Most phones can only produce twelve of these, the familiar 0 through 9, #, and * tones. There are, however, four other touch tones, usually designated A, B, C, and D. Some modems and special purpose phones can produce these, but in general they are seldom encountered. The frequencies used for each touch tone are listed in the table below.
In North America, the minimum touch tone length is 50 milliseconds; the minimum separation between touch tones is also 50 milliseconds. Touch tones must follow these conventions to be processed accurately by the DTMF Decoder application.
When DTMF Decoder encounters a touch tone when analyzing a recording, its corresponding symbol is written to the text window.
1209 1336 1477 1633
_____________________
697 | 1 2 3 A
|
770 | 4 5 6 B
|
852 | 7 8 9 C
|
941 | * 0 # D
Dial Tone:
The standard dial tone is a continuous signal composed of equal amplitude 350 and 440 hz tones.
When DTMF Decoder encounters a dial tone when analyzing a recording, "DT" is written to the text window.
Rings:
There are two common types of rings, normal and PBX. DTMF Decoder recognizes both. The normal ring is a tone containing an equal amplitude mix of 440 and 480 hz tones. It is on for two seconds, and off for four. The PBX ring is spectrally identical to the normal ring, but is on for one second, and off for three.
It may seem surprising that the ring sound is just another DTMF tone, since it appears to chatter on and off during the actual ring ("B-r-r-r-r-r-r-ing"). The reason for this is simple: the difference frequency produced by the mixing of the two nearly equal frequencies that make up the ring is 40 hz, and so the 'volume' of the mixture dips and rises 40 times per second.
When DTMF Decoder encounters a ring when analyzing a recording, "RING" is written to the text window. It does not differentiate between the two rings.
Busy Signals:
Similar to rings, there are three common signals that share the same frequencies but differ in timing. These are the busy, local reorder, and toll congestion tones. They are composed of (again, an equal amplitude mix of) 480 and 620 hz tones. The busy signal is on for half of a second, and off for half of a second. The other two are shorter in duration and their off times are shorter, too.
When DTMF Decoder encounters one of the three 'busy' signals when analyzing a recording, "BUSY" is written to the text window. It does not differentiate between the three.
***History:
Version 1.0 released 10/7/93
Version 1.1 released 12/3/93
• Snazzy finder icon added.
• Spinning cursor added to distract user during pauses.
• DTMF window comes up automatically after launch.
• Docs partially rewritten.
Version 1.2 released 12/16/93
• Support for black & white.
• No longer requires System 7.0 or above. I know for sure it runs on 6.0.7 and above.
• No longer requires a sound input device. This was something I just forgot to fix
earlier.
• Will run even if a printer is not chosen.
• Better color handling- allows user to customize the text and graph window colors.
• Superfluous data removed from pict resources.
• Play-through option added.
• Algorithm sensitivity made user-adjustable.
• Volume may now be set internally, or set to system default.
• No longer clobbers sound manager when it returns to finder.
