How to Smoothly Edit Interview Audio in DaVinci Resolve: Mid Word Cutting Technique

In this final part of the series, I’ll share a powerful method to make your interview cuts sound as smooth and natural as possible. This technique relies on cutting at precise points in the audio, avoiding jumpy transitions caused by pitch differences. First, a special thanks to Luciana Pandolfelli, a talented Italian film editor, for introducing me to this method over a decade ago.

As always, you can watch the tutorial from the link below. Let’s dive in and see how it works!

Understanding Audio Rhythms and Pitch

When editing an interview, it might seem logical to cut between words, as we do in written text. However, spoken language has a rhythm where words blend together, and pauses rarely align perfectly with word boundaries. For example:

Original sentence:
"We've got mathematical models of our proteome, of our genome, of our connectome, all the connections and informational bioinformatic connections in the brain."

When listening closely, the words flow together, and cutting at word boundaries results in awkward transitions. Examining the waveform reveals that pauses don’t align exactly with word endings, making the cuts sound jumpy.

Why Repeated Words Can Be Tricky

People often repeat themselves during interviews, but even repeated words don’t sound identical due to pitch variation. Consider this example:
"The works that we do works like the Philip K. Dick Android."

Although “works” is repeated, the two instances differ in pitch, making a direct cut sound unnatural. On paper, you might write:
"The works like the Philip K. Dick Android."
But in the audio, this would create an uneven cut:
"Otherwise, the works like the Philip K. Dick Android."

To create a smoother cut, we need a different approach.

Cutting on Voiceless Letters

The key to natural-sounding cuts lies in voiceless letters—sounds that don’t require the larynx to vibrate, such as P, T, K, S, and F. These letters lack pitch, so they sound consistent and provide clean cutting points.

1. Identify Voiceless Sounds in the Waveform:
   Voiceless letters are often preceded by small pauses, making them easy to spot in the waveform. For instance, in “of our connectome,” the “K” in “connectome” provides a clean cutting point.

2. Use Voiceless Sounds to Cut Mid-Word:
   If a sentence repeats a word, cut during a voiceless sound to combine the best parts. For example:
   Original:
   "We need AIs that are meditation assistants, that are teachers, that are elder care workers, that are therapists, that are scientists."
   Cut:
   "We need AIs that are meditation assistants, that are teachers. We need AIs doing loving, positive, compassionate things."

   In this example:

   • Use “teach” from “teachers” in the first part.

   • Cut during the “S” in “scientists” to connect the second part.

3. Blend Repeated Phrases with Voiceless Sounds:
   In "The works that we do works like the Philip K. Dick Android," cut mid-word between “WOR” and “CS” for a seamless transition.

Matching Repeated Sounds

Sometimes, you can use repeated words or phrases to smooth cuts by aligning their voiceless components. For instance:
"We've got mathematical models of our proteome, of our genome, of our connectome, all the connections and informational bioinformatic connections in the brain."

Here, “connections” is repeated. Cut mid-word using the second “K” sound in “connections”:
"We've got mathematical models of our genome, of our connectome, all the connections in the brain."

The result is a more natural-sounding sentence.

Handling Exceptions

In certain cases, cutting on voiceless sounds isn’t ideal:

1. Strong Vowels and Reverb:
   If a loud vowel like “E” follows a voiceless sound, the reverb can introduce noticeable pitch differences. In this case, cut before the strong vowel.

2. Offsetting Video and Audio Cuts:
   When mid-word cuts are noticeable, offset the audio and video cuts by a few frames:

   • Place the video cut slightly before or after the audio cut.

   • Use a full word from one version of the repeated phrase in the video to mask the cut.

For example:
"Otherwise, the works like the Philip K. Dick Android, our curiosities."
By offsetting the video and audio, the cut becomes invisible.

Final Thoughts

This method of cutting on voiceless letters ensures smooth, natural-sounding transitions in interview edits. It eliminates jumpy pitch changes and maintains the rhythm of speech. In previous videos, I covered organizing files, creating multicam clips, and selecting audio bites for interviews. If you missed them, check out the other tutorials on my channel to complete your editing workflow. Happy editing!