Talking Weather is a scalable AVAS/ATIS/VOLMET system, creating multiple audio outputs from incoming data. It gives you a warranty of compliance to ICAO Annex 3, including the Amendment 73.
According to the source and the issue time of the reports, Talking Weather automatically synthesizes and plays the output to independent audio channels. The standard number of audio channels is four, but additional channels can be included. For each channel, the list of associated observing stations can be specified.
The composed message is transmitted in a text form to the destination where it is synthesized and converted into the voice. This allows for the delivery of the voice messages to virtually any distance. Each audio channel requires a bandwidth of approximately 24 kb/s to ensure the transfer of three channels over one ISDN leased line. It is also possible to use any other transport technology which is able to carry TCP sessions, for example local networks.
The voice output does not contain any pre-recorded samples. It is completely based on a state-of-art text-to-speech engine, which makes the system easily upgradeable and configurable. Users can define different pronunciations for different observing stations and channels. This feature enables to reconfigure the product when the METAR format changes, or when there is a need for a different pronunciation of some part of a report.
The list of observing stations associated with each channel can also be configured by user.
In the beginning there is a METAR report:
EDDF 280650 32012KT 280V350 9999 FEW025 SCT100 13/07 Q1022 NOSIG=
You can customize its conversion to the text, or use a predefined ICAO Annex 3 compliant form:
“Frankfurt VOLMET 1 Broadcast. Frankfurt VOLMET 1 Broadcast. FRANKFURT, 0 6 5 0, wind 3 2 0 degrees 1 2 knots, variable between 2 8 0 and 3 5 0 degrees, visibility 1 0 kilometers, few 2 thousand 5 hundred feet, scattered 1 0 thousand feet, temperature 1 3, dewpoint 7, Q N H 1 0 2 2, nosig.”
Finally the text is translated into a digital voice using the state-of-art text-to-speech synthesis. You may download the resulting sample (MP3, 96 KB) to hear for yourself.
• for Talking Weather Server: Solaris®, Linux, Unix
• for Talking Weather Broadcast Client: micro distribution of Linux
For Talking Weather Server:
• Petium 3 1 GHz
• 256 MB RAM
• 10 GB harddisk
• dual-system cluster is recommended for high availability solutions
For Talking Weather Broadcast Client (optional for remote-broadcasting):
• Intel 386 compatible processor with at least 200MHz
• 32 MB RAM
• 64 MB harddisk (including Compact Flash(TM) or Microdrives(TM)
• Sound Card (supported by Linux)
• 10/100BaseT Ethernet plug or ISDN modem or any other connection providing TCP/IP (supported by Linux)
AM/FM voice radio broadcasting equipment connectable to PC audio jack
• ICAO Annex 3 (incl. Amendment 73)
• WMO Manual No. 306 (Manual on Codes)
• all kinds of TCP/IP, asynchronous, file-exchange based and FTP channels
• automatic reception of text coded reports (TAC) as well as BUFR reports (TDCF)
• automatic configurable translation of coded messages into human readable text form of both TAC and TDCF
• automatic computer-based synthesis of text into voice form
• AM/FM radio transmission of voice
• any text-coded reports or other text advisories (METAR, TAF, SPECI, SIGMET, GAMET, warnings, etc.)
• BUFR/CREX coded reports (e.g. METAR)Output formats• Airport radio broadcast
• WAV or MP3