Cut Speech Logic ( All cut operations have to be in the [ default cut ] range )

1. Main voice in a group conversation is clear, others are low and fuzzy

2. Cut them out and transcribe the clear part

3. In the beginning / end of audio

4. Unclear word / part

4.1. In the middle of the audio

4.1.1. (Speech 1 + Unclear word + speech 2) - cut and transcribe speech 1 OR speech 2

5. Overlapping speech

5.1. Entire audio with overlapping speech

5.2. Part of the audio is overlapping speech

6. Transcribe the main voice

7. Music, Melodies, Songs, Animals and Natural sounds

7.1. Entire audio without human speech

7.1.1. Discard

7.1.1.1. Cut the song and transcribe clear speech part ( Ignore song if does not affect speech)

7.1.2. Do without thinking about the completeness of the sentence, while cutting the audio

7.2. Song with lyrics

8. Discard

9. Discard

10. Cut the overlapping part and transcribe the clear part

11. If affects speech - (Speech 1 + noise + speech 2) - cut and transcribe speech 1 OR speech 2

12. Noise

12.1. Entire audio full of noise

12.2. Part of the audio is noise

12.2.1. Hard to cut

12.2.1.1. Discard

12.3. In the beginning / end of audio

13. If does not affect speech - (Speech 1 + noise + speech 2) = ignore noise and transcribe speech 1 AND speech 2

13.1. Cut to keep only two modal words

13.2. Able to count the number

13.2.1. Transcribe strictly according to the number of times they get repeated

14. Modal words

14.1. Clear human speech with Background Music/Sound that does not effect clarity of the speech

14.1.1. Transcribe

14.2. If the noise does not affect the clarity of speech

14.2.1. Transcribe

14.3. In the middle of the audio

14.3.1. Not able to count the number

14.3.1.1. (Speech 1 + laughing + speech 2) - cut and transcribe speech 1 OR 2

15. Overlapping but speaking the same word/sentence and clear

15.1. Transcribe