mawic / ptttl Goto Github PK

The Polyphonic Tone Text Transfer Language (PTTTL) is a way to describe polyphonic melodies, and is a superset of Nokia's RTTTL format, used for monophonic ringtones.

I needed a good way to store simple tones and melodies for some project. RTTTL looked pretty good but only works for monophonic melodies. I needed polyphony.

Valid RTTTL strings are also valid PTTTL strings. A parser that properly handles PTTTL can also handle RTTTL.

A PTTTL string is made up of three colon-seperated sections; name section, default values section, and data section.

Whitespace characters, empty lines, and lines beginning with a "#" character are ignored.

The initial "name" section is intended to contain the name of the ringtone in the original RTTTL format. PTTTL requires this field to be present, to maintain backwards compatibility with RTTTL, but places no constraints on its contents.

The very first statement is the default value section and is identical to the section of the same name from the RTTTL format.

b=123, d=8, o=4

• b - beat, tempo: tempo in BPM (Beats Per Minute)
• d - duration: default duration of a note if none is specified
• o - octave: default octave of a note if none is specified

The PTTTL data section is just like the RTTTL data section, in that a melody consists of multiple comma-seperated notes to be played sequentially. Unlike RTTTL, PTTTL allows multiple melodys to be defined, separated by the vertical pipe character |, all of which will be played in unison.

The format of a note is identical to that described by the RTTTL format. Each note includes, in sequence; a duration specifier, a standard music note, either a, b, c, d, e, f or g (optionally followed by '#' or 'b' for sharps and flats), and an octave specifier. If no duration or octave specifier are present, the default applies.

Valid values for note duration:

• 1 - whole note
• 2 - half note
• 4 - quarter note
• 8 - eighth note
• 16 - sixteenth note
• 32 - thirty-second note

Dotted rhythm patterns can be formed by adding a period "." either after the note letter (e.g. c#., or c#.5), or after the octave number (e.g. c#5.)

Valid values for note pitch (non case-sensitive):

• P - rest or pause
• A
• A# / Bb
• B / Cb
• C
• C# / Db
• D
• D# / Eb
• E / Fb
• F / E#
• F# / Gb
• G
• G# / Ab

Valid values for note octave are between 0 and 8.

Consider the following PTTTL string:

# 123 beats-per-minute, default quarter note, default 4th octave
Test Melody:
b=123, d=4, o=4:

16c, 8p, 16c | 16e, 8p, 16e | 16g5, 8p, 16g5

This would play 3 sixteenth notes simultaneously (C, octave 4; E, octave 4; G, octave 5), followed by an eighth note rest, followed by the same three sixteenth notes again

Note that the above sample is much easier to read if we put each melody on a new line and align the notes in columns. This is the recommended way to write PTTTL:

# Nicely aligned
Test Melody:
b=123, d=4, o=4:

16c,  8p,  16c  |
16e,  8p,  16e  |
16g5, 8p,  16g5

In order to keep things readable for large PTTTL files with multiple concurrent tracks, a semicolon character ; can be used further break up melodies into more practical blocks. Just as the veritcal pipe character | seperates concurrent tracks within a single melody, the semicolon character seperates multiple sequential melodies within a single data section. Melodies seperated by semicolons will be stitched together, one after the other, in the final output.

The semicolon does not affect any of the timings or pitch of the generated tones; it just makes the PTTTL source a bit more readable. Have a look at this larger PTTTL file, with 4 simultaneous melodies, for a good example of why the semicolon is useful:

All Star but it's a Bach chorale:
d=4,o=5,b=40:

#some   bo  -   dy      once    told    me      the     world   was     go -

16g3.,  32g3.,  32g3.,  16g3.,  32f#.,  32g.,   32g.,   16a3.,  32g#.,  32g#. |
16g.,   32g.,   32g.,   16d.,   32a3.,  32b3.,  32b3.,  16a.,   32b.,   32b.  |
16g4.,  32d.6,  32b.,   16g.,   32a.,   32d.,   32d.,   16e.,   32b3.,  32b3. |
16g4.,  32b.4,  32d.,   16b.,   32d.,   32d.,   32d.,   16c6.,  32d.,   32d.  ;



#-na    roll    me,     I       aint    the     sharp - est     tool    in

32a.,   32a.,   16g.,   32g.,   32g.,   32g.,   32g.,   32f#.,  32d#4., 32g.  |
32c4.,  32c4.,  16e.,   32c.,   32d6.,  32b.,   32d4.,  32d4.,  32f#.,  32e4. |
32e.,   32e.,   16e4.,  32c4.,  32b3.,  32c4.,  32b.,   32c.,   32a.,   32b.4 |
32e.,   32e.,   16c.,   32e.,   32d.,   32d.,   32d.,   32a.,   32c.,   32e.  ;



#the    she  -  ed,             she     was     loo  -  king    kind    of

32g.,   16c4.,  32a3.,  8p.,    32g.,   32g.,   32g3.,  32g3.,  32g3.,  32f#. |
32d4.,  16e.,   32d.,   8p.,    32g4.,  32g4.,  32g.,   32g.,   32g.,   32a3. |
32b.4,  16c.,   32f#4., 8p.,    32b3.,  32a3.,  32b4.,  32b.,   32b.,   32a.  |
32e.,   16g4.,  32a4.,  8p.,    32d.,   32d.,   32d6.,  32d.,   32d.,   32d.  ;



#dumb   with    her     fing  - er      and     her     thumb   in      the

32g.,   32g.,   32g.,   16a4.,  32b.,   32b.,   32a.,   32a3.,  32g.,   32g.  |
32a.,   32b3.,  32e4.,  16e.,   32g#4., 32g#4., 32c.,   32c.,   32b4.,  32c.  |
32b3.,  32d.,   32b4.,  16c6.,  32e.,   32e.,   32e.,   32a.,   32c4.,  32c4. |
32d.,   32d.,   32e.,   16c6.,  32d.,   32d.,   32a4.,  32e.,   32e.,   32e.  ;



#shape  of      an      L       on      her     for  -  head

16g.,   32g.,   32g.,   16a.,   32g.,   32g.,   16a.,   32a4. |
16d.,   32b.,   32d4.,  16f#.,  32e4.,  32e4.,  16a4.,  32d.  |
16d6.,  32c4.,  32b.,   16c.,   32b4.,  32b4.,  16f#.,  32e.  |
16b3.,  32d.,   32d.,   16d#4., 32e.,   32e.,   16c.,   32e.

A sample implementation of a PTTTL parser and .wav encoder is provided in ptttl_parser.py and ptttl_audio_encoder.py.

These sample scripts will work on Linux & Windows (untested on OSX).

Try the sanmple scripts out with some of the included melodies in the rtttl_examples directory.

git clone https://github.com/eriknyquist/ptttl
cd ptttl
pip install -r requirements.txt

Converting a PTTTL file into audible tones in a .wav file:

python ptttl_audio_encoder.py polyphonic_example.txt monty.wav