Timespeak

ESP 8266-based talking clock

created 06 Sep 2017

(c) by Dirk Spiller

This project is in the public domain. It is published under GPL v3

This sketch gets the Time via NTP and speaks it out loud using DFPlayer MP3 module and several voice files as mp3 (e.g. in German language). It utilizes an IR distance sensor TCRT5000 (eg. on Amazon) to recognize if the user holds his hand close to the sensor and then speaks the time again.

It uses:

• ESP8266 WiFi Libs for Arduino IDE
• TimeLib
• Timezone Library for correct handling of DST
• DFPlayer Mini Lib

The mp3 handling is inspired by the mp3 talking clock from LAGSILVA Instead of setting the initial Time manually, it uses NTP as a Sync Provider as in the TimeNTP_ESP8266WiFi example from TimeLib Instead of saying the time every ten or so minutes, it says the time, when you hold your hand close to the IR sensor

You need a lot of mp3 files for the different hours and minutes to be said. Here is what is required, how they needs to be stored and how to easily create them:

MP3 Voice output via DFPlayer module

Attention: On microSD card, all sound files have to be put in a folder called "mp3"

The following files are expected:

• 1.mp3 - 24-mp3 : "zero hours", "one hours", "two hours" to "23 hours"; In German it is required to say "Uhr" meaning "hours" between hours and minutes of a timestamp!
• 25.mp3 - 84.mp3 : "zero" to "fifty-nine" for the minutes
• 85.mp3 : "to"
• 86.mp3 : "past"
• 87.mp3 : "quarter"
• 88.mp3 : "half"
• 85.mp3 - 88.mp3 are meant for colloquial time telling (future expansion, not yet implemented!)

Decent mp3 files can easily be created online using the following strategy:

1. Create a text file containing all required words with breaks, like in German: (included in github repository for your convenience)

   Example file content for German language:

   0 Uhr, 1 Uhr, 2 Uhr, 3 Uhr, 4 Uhr, 5 Uhr, 6 Uhr, 7 Uhr, 8 Uhr, 9 Uhr, 10 Uhr, 11 Uhr, 12 Uhr, 13 Uhr, 14 Uhr, 15 Uhr, 16 Uhr, 17 Uhr, 18 Uhr, 19 Uhr, 20 Uhr, 21 Uhr, 22 Uhr, 23 Uhr, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, Vor, Nach, Viertel, Halb

2. Choose an online Text-To-Speech Service like Naturalreaders

3. Paste the text file contents to the input box

4. Choose language and voice

5. Use Audacity for recording the output directly as explained in the Audacity manual

6. Label recording into separate tracks as described in the Audacity manual

7. Export separate sound files from labels as described in the Audacity manual

8. Now click "ok" 88 times ;-) -> the resulting files are cut and named correctly by default.

Possible changes and additions or future ideas:

• Add a TM1637-based 4 Digit module to also display the time (eg on Amazon)
• Add a DS3231 RTC module to overcome powerloss and network failures (eg. on Amazon)
• Use DCF77 module for most accurate time source
• Add different alarms using several mp3 files
• Clock / Alarm set using serial connection or different "hand-over-events"
• Speak colloquial time like "a quarter to five" in terms of mp3 files this is already prepared
• Speak in dialects or with funny speakers (just replace the mp3 files)