Static on the line.
This was by far the hardest and most technical challenge.
There were several problems for me here. First, I have a hearing disorder that allows me to hear fine but makes it hard to interpret meaning from what I hear. Second, the audio was in a language I do not understand, and one of the hardest to translate from phonetic sound to spoken word: Welsh.
The first thing I had to do was manipulate the audio to separate the background noise from the foreground. I tried Audacity, but eventually had to use an online AI audio splitter instead.
I listened to the background track continuously and still couldn’t quite make it out, so I searched online for the Welsh names of train stations and worked out which lines ran close to Ogmore.
My hearing problem kept getting in the way. Feeling stuck, I tried to think of how I could confirm whether I was hearing and understanding the pronunciation correctly. That led me to wonder: if sounds produce wavelengths, and those wavelengths correlate to shapes, could there be a way to match wavelength shapes to sounds? It turns out there is.
By overlapping the sound shapes with the waveform in the recording, I discovered I was hearing a C sound as an L sound.
With a renewed direction, and a little push from a friend who had already completed the challenge, I was able to identify the station: Cardiff Central, known in Welsh as Caerdydd Canolog.
Fourth flag found.