I am trying to design a way to record line-in audio from a push to talk radio. The recording would be incoming and outgoing audio. Whilst at the same time parse a nmea gprmc string and save the audio files name as the gps date, time (hhmmss), and lat-long coordinate. I have the gps part all sorted out. I may have to save coords another way, could be too long.
Example would be 27Jul09234521.30.446944_86.561389.wav
The problems is that I have no idea where I need to go next.... Seeing alot of message traffic on waveshield and the waverp library.
-I do not know how to write a audio file to a sd card.
-If one can write audio file to sd card what is the max character length?
-Would there be a hugh delay between file save and new file created?
-Can a Arduino with ? shield be programmed to record line-in audio?
-Would this method be too slow to create write and save file that I would miss radio traffic?
You first have to do the math and exploring experiments to understand what can be done:
Q: What is the sound quality you want?
Q: how fast can you write to SD card?
Q: how long does it takes to read GPS coordinates
Note that Arduino is not multitasking, so I think you need a separate audio board.
Afaik the Arduino libraries do not support long filenames for SDcards, so you need to look for another naming scheme. An old trick is to use subdirectories for part of the name e.g
\30\446944\86\561389\20101207.wav (YYYYMMDD sorted alphabetically is sorted in time!)
-I do not know how to write a audio file to a sd card.
Audio is just a stream of bytes. The SDcard does not know what they mean.
-If one can write audio file to sd card what is the max character length?
There are no characters in an audiofile, only bytes. Please rephrase your question
-Would there be a hugh delay between file save and new file created?
Do the experiment to see if it fit your expectations.
-Can a Arduino with ? shield be programmed to record line-in audio?
I did google that, that's how I found this site. I also purchased the wave shield and an Arduino UNO for the project. I also have about 5 other arduinos to choose from to spread the load if I need to multitask. I have also been reading about realtime audio processing, waverp, and freqcounter1 libraries. Some of what I have been reading is like a "Pig looking at a wrist watch" I don't understand, but I will get throught it.
An old trick is to use subdirectories for part of the name e.g
\30\446944\86\561389\20101207.wav (YYYYMMDD sorted alphabetically is sorted in time!)
What about writing two files: (YYYYMMDDSS) 2010120743.wav and 2010120743.txt; one has audio and the other has GPS coordinates.
Would it be posible to write to a SDcard, two files near realtime? and by near realtime I mean write the two files and not miss recording additional radio taffic?
I want the audio to be understood not music quality.
I am currently parsing the GPS data once a second. I have it broken into three parts date, time, and Lat-long.
Yes I have already hacked into a recorder and through signal start and stop recording. I gues at the same time I could capture the GPS date time and coordinates and write that to a file. I would have a recorder, GPS, data logger, and power. I could sync the date and time of recorder and GPS to serve as a common marker.
I don't care too much about the quality of recording. I just want to be able to understand it. I would later save audio and location data for archive. I need a "black box" for the individual.... Where was I when I received radio trans; where was I when I said something on the radio. Lets save the audio in the event we need to dispute or prove what was said....
The audio should be able to play on a XP machine or windows 7 and the location data would later be mapped to google maps or another mapping program.
If I should be using another processor for this project, what should I use?