I set up a Motorola base station receiver to grab the audio from the local NWS station (WXM-60) as the handful of streams available were a bit noisy. Outcome:
- It sounds OK, however it is noticeably "different" then a person in front of a decent microphone. VHF bandwidth is much less than commercial FM
- Also ran into a snag trying to figure out the repeat on the audio. It would be nice if NWS put a tone as the start point as without it, automation gets a bit difficult.
On to the TTS project
- Still working on correctly parsing the RSS feed, the issue is my interpretation of the XML
SP
Here is what NWS radio stations should do is have a short version of the weather forecast which is like 2 or 3 min long and longer version version which like 6 to 7 min long. The longer version should only be played like 2 times a hr and the shorter version the rest of the hr. They also need to use a much better sounding voice which does not sound computerized and not monatone.
I miss www.pirateweather.com, sure it was text to speech but all you had to do was enter your city and state or just your zip code and bam!, you an instant weather forecast read by a robo-meteorologist in mp3 format.
I miss pirateweather.com, sure it was text to speech but all you had to do was enter your city and state or just your zip code and bam!, you an instant weather forecast read by a robo-meteorologist in mp3 format.