Assamilator needs to correct the timestamps in the transcripts to use real time based off the file creation date and the telative timestamps in the vtt file.

[sweep-ai]

Summary:
To address the issue, we need to modify the transcribe_audio function in the AudioProcessor class to extract timestamps from the .vtt file and adjust them based on the file creation date. We will also need to modify the save_to_obsidian_vault function to include these adjusted timestamps in the saved transcription.

To checkout this PR branch, run the following command in your terminal:

git checkout assamilator_needs_to_correct_the_timestamps_in_the

[sam1am]

Produces error:
NameError: name 'adjusted_timestamps' is not defined