I have a Sony ICD-PX333 recorder which support the track mark functionality. Track mark allows one to place audio labels whilst recording. Once the audio file is connected to the PC, Sony’s Sound Organizer application will show where the track marks are. This allows one to easily divide the audio file into multiple segments.
I would like to do the same thing using Audacity, however I cannot find clues of where the track mark may be hidden in the audio file. In fact, I have created a silent audio file and then manually added track marks using the Sound Organizer application. After that I looked at the waveform of my silent audio file and I still could not find any clues.
Here is the link to the empty audio file of duration 30 seconds with track marks at 10s and 20s. link to empty.mp3
I know that Audacity does not support the track marks by default. I am wondering what tools are available to find the hidden signal (spectra, etc.) as the waveform shows no difference although track marks have been added. Do you have any idea where the track mark may be stored?
You can add a track mark only to files recorded using the IC recorder.
In other words, it’s the Sony proprietary format to keep you using Sony products. There’s nothing in that sound file but dead silence. My guess is Sony burned a non-standard meta-data format into the file. Sony regards technical standards as suggestions, not goals.
Oh ok, that’s disheartening…I am using the recorder to record my professors’ lectures and was planning on writing an application that automatically divides them into subsections.
I did not record the file using an IC Recorder. I opened up Audacity and generated 30 seconds of silence. I then exported the file as a mp3 using LAME. After that I opened up the file with Sound Organizer and added the track marks. I think that it does not actually matter whether the track marks were created on the IC Recorder or not. Rather, I think that the T-Mark is interfaced with Sound Organizer.
You could search for the marks with a hex editor.
Each mp3 frame has its own header (which isn’t there after decoding, that’s why you’ll don’t see anything in the waveform). one frame is about 1152 samples or so long. There are programs that can display the individual frame information. If you can localize the track marks in the mp3 file, you can search for them via a plug-in and produce a label track. The imported track can’t be searched directly though.
