I'm not entirely sure if StackOverflow is the right site for this question, but I looked at all the other sites in the StackExchange network and none of those seemed more suitable, so apologies if this is not a StackOverflowy enough question.
Anyway, I'm working on making a pretty bare-bones WAV file writer. In addition to the usual metadata required for playback, I need to be able to add cues at arbitrary points in the file, but I ran into trouble with files over 4GB. I understand there are multiple approaches to getting larger WAV files, but the approach I would like to take is to use the RF64 WAV file format. I ran into trouble with the cue structs because these of course use 32-bit sample position offsets which limit cue markers to the first ~4 billion samples but not after.
Does anyone know how I can achieve this?
EDIT 1: I confused sample offset with data. The cue chunk uses sample offsets to position cue points, which may allow this to be used without modification in files over 4GB (e.g. there are multiple channels in the file). The problem remains for files containing more than 4 billion samples. I understand this is a high threshold, but I also don't want to impose a limit on audio length if I don't have to.
EDIT 2: After some experimentation with large .wav files in Adobe Audition 6, I figured out that Audition side-steps this limitation by generating a .xmp file with XML-like data when saving a sufficiently large .wav file. However, I would still like to implement cues in as standardized a way a possible within a single .wav file if possible.