| tones(1) | Linux System Manual | tones(1) |
tones - a sequential tone generator program
tones [options] [waveform] T freq(s)|notes(s)|command_file(s)
tones generates one or more tones of various types (waveforms) and duration (T millisecs) of the specified frequencies or notes, or mixtures of frequencies or notes. See tones -h for a list of possible waveforms. The waveforms should include sine, cosine (90 degrees out of phase to sine), square (50% mark/space ratio), sawtooth (a ramp waveform), triangle and noise. Sine is the default. Besides the inbuilt waveforms, waveforms can be loaded from suitable WAV files - see below LOADABLE WAVEFORMS.
T is the default number of millisecs that each tone is to be played. Frequencies (freq(s)) are specified in Hertz as integers. A frequency of 0 causes T millisecs of silence to be played. Notes are specified as the musical note letter with an optional '#' to sharpen the note, then an octave number. Octaves run from C to C. Middle C is C3, the immediately preceding note is B2! The first Octave is from C0 to C1.
Several frequencies/notes can be played at once, by specifying the frequencies/notes required joined by a ',' character (but no spaces!).
e.g. 1000,1500,2000 specifies that the three frequencies are played together, all at the same relative level. See AMPLITUDES sections below for a discussion of how to set absolute amplitude levels or differing relative amplitudes for notes played together or serially.
Each freq specification can optionally contain a duration, by appending ':T', where 'T' is the duration in millisecs. This duration overrides the default. Also the default duration can be changed by using a the ':T' format on it's own - not appended to a freq spec.
e.g. 1200,600:1000 play the two freqs for 1 sec
e.g. c3,e3,g3 play the C major chord
e.g. :250 set the default tone duration to 250ms
Waveforms can be specified/altered at anytime. A single waveform name specifies that waveform to be used for all channels. Alternatively a comma (',') separated list of waveforms can be given to specify or alter the waveform to use for a given channel. Omitting a waveform in a list, means that the previous waveform is left unchanged.
e.g. square,,triangle specifies using square waves for chan 1, chan 2 is left unchanged, and triangular waves are used for channel 3.
The digital samples (either 8 or 16 bits) are played by default to the Linux /dev/dsp device at a samplerate of 22050 samples per second, in mono mode. (see CONFIGURATION FILES section below)
Fractional Hertz frequencies are not supported. Of course, only frequencies less than half the samplerate (number of samples/sec) can be accurately generated; but the program doesn't check this.
Instead of playing the output to /dev/dsp the samples can be written to a file as raw samples (-o file) or written in WAV format (-w wavfile). These data files can then be played back quickly with a raw data or WAV file player (e.g. wavplay) without the overhead of actually generating the samples.
There are some special 'commands' that can be specified, that may be useful in input files.
Further, if the word is not one of the above, then tones checks to see if a file of that name exists, and if it does then the file is assumed to be a file of tones commands which are executed.
e.g. tones -v :100 tune1 tune2 will interpret and play the tones commands in files tune1 and tune2. This file processing is recursive. Files of commands can execute other files of commands etc. As usual, '-' can be used to specify stdin.
tones by default works in a 'relative' amplitude mode, where the output level and sample range are maximally maintained. This ensures the best signal accuracy.
When specifying multiple frequencies/notes to be played together, then the relative amplitudes can be specified in deciBells by appending "@db" to the note.
e.g. 440,880@-12,1760@-30 specifies a mixture with 880Hz -12dB down, and 1760Hz -30dB down relative to the level of 440Hz. The mixed signal samples will span the full 16 or 8 bit range permitted for maximal signal accuracy.
The dB levels indicate the relative power levels. -3dB being at a relative power level of 0.5, -20dB being at a relative power level of 0.01 . However power levels are proportional to the square of the signal amplitude. So a signal at -6dB (quarter power) will only have its amplitude down by half. To reduce a signal amplitude by 1/10 then specify -20dB, i.e. a power level down by a factor of a hundredth.
dB levels can be specified as decimal values.
tones can work in an absolute amplitude mode, where signal power levels are specified in deciBells (dB) relative to a 0dB level that indicates a peak value of +32767/-32768 for 16 bit signed samples, and 255/0 for 8 bit unsigned values. Hence any signal at a positive dB level will be clipped. Signals at a negative dB level will attentuated. If no level is specified then 0dB is assumed.
e.g. 500@-20,750@-6,1000,-12 gives 500Hz at -20db (amplitude 0.1), 750Hz at -6db (amplitude 0.5), and 1000Hz at -12dB (amplitude 0.25). The final mixed signal will have an amplitude of 0.1 + 0.5 + 0.25 = 0.85 or -1.4dB.
As can be seen, there is no "hands-free" in absolute mode. You have to work out the dB levels yourself and ensure that the resultant mixed signal does not go above 0dB and get clipped. Remember also that a sine wave at -80dB down (amplitude 1/10000th) only has 6 digital levels and is a pretty poor representation of a sine wave, not suitable for post amplification and use!
In absolute mode the base 'zero' level can be altered at any time by use of the @dB command. All subsequent dB levels specified will have this base level added to them.
e.g. @-20 1000,1200@+6,1400@-6 is the same as 1000Hz at -20dB, 1200Hz at -14dB and 1400Hz at -26dB.
Given that the generation method used by tones to generate a waveform of FHz is simply to sequentially select every Fth sample from a buffer containing S samples of one complete waveform at a frequency of 1Hz (treating the buffer as circular, the beginning conceptually joined to the end), where S is the number of samples per second, it is possible to load a customised waveform from a WAV file containing the S samples of a 1Hz waveform. See the -load WavFile and -lw N options below. The name of the waveform is taken as the basename of the WavFile, i.e. with any trailing '.suffix' and leading path removed. Each loaded waveform should hence have this name unique, and different from the inbuilt waveform names.
The samples in WavFile should be 16 bit, mono, of the same number of samples as tones' playing samplerate, e.g. if tones is playing at 32000 samples per sec then the WavFile should contain 32000 16 bit samples. 16 bit samples are needed, because tones works internally with 16 bit samples, even if it is feeding 8 bit samples to the sound card or output file. Ideally the samples should span one complete wavelength, i.e. represent 1 second of a 1Hz signal. However this can be varied if used with some intelligence. If, say, 1 seconds worth of 5Hz of the waveform is used, then the output frequency will be 5 times higher than specified. If you have a mixture of 3Hz and 5Hz samples, then the frequencies generated will be a mixture of 3 and 5 times the frequency specified. I hope that is all understandable!
See the tones.eg directory for some examples of loadable modules and how tones itself can be used to generate the loadable waveforms.
See also the tones.eg directory in the siggen distribution.
Three possible configuration files can be used: a LOCAL config file (usually in current directory), a HOME config file in user's $HOME directory and a GLOBAL config file.
All the siggen suite of programs are compiled with the names of the config files built in. By default the configuration files are:
The config files do not have to exist. If they exist and are readable by the program they are used, otherwise they are simply ignored.
The config files are always searched for configuration values in the order LOCAL, HOME, GLOBAL. This allows a scheme where the sysadmin sets up default config values in the GLOBAL config file, but allows a user to set some or all different values in their own HOME config file, and to set yet more specific values when run from a particular directory.
If no configuration files exist, the program provides builtin default values, and most of these values can be set by appropriate command line switches and flags.
See siggen.conf(5) for details of the configuration files.
tones looks for configuration values CHANNELS, DACFILE, SAMPLERATE, SAMPLESIZE, VERBOSE, LOADABLE_WAVEFORMS.
Copyright 1995-2008 Jim Jackson
The software described by this manual is covered by the GNU General Public License, Version 2, June 1991, issued by :
Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies.
Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one.
Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be included in translation instead of in the original English.
Jim Jackson
Email: jj@franjam.org.uk
| 20 Feb 1998 | Release 2.3 |