Annotation file vs. searchable text?

What are the advantages and disadvantages of choosing an annotation file vs. searchable text option when transcribing speech?

Also, Searchable text is added to the media file (like a video recording) directly and accessed via the “alternate view” button, is my understanding. However, I can’t delete or edit the transcribed text later; removing it reappears. The transcribed text doesn’t match the speech in the video, resulting in gibberish.

Searchable text is of course only searchable and remains invisible usually. Annotations can be edited and might include clickable timestamps.

Which settings did you actually use and did you choose the right language? Especially in case of a poor audio track or a wrong language Whisper might hallucinate.