How to Protect Your Podcast Guest’s Privacy: Harmonize, Voiceover or Analyze

article featured image

Sometimes you have a guest on your podcast whose information and experience are more important than their fame. Particularly if you have a podcast about current events, you might interview a guest whose identity you have to conceal. Privacy concerns are more serious than ever. Fortunately, podcasters are creative, and it’s not difficult to protect your podcast guest. You can use technology, a stand-in, or you can discuss their interview responses with a co-host. Let’s look at the details.

Use a Harmonizer

Most of us have seen a news interview where a person agrees to participate in exchange for anonymity. They appear as a blurry silhouette, with their voice distorted. You can disguise a voice using pitch correction with a harmonizer plug-in.

One example is the Eventide H3000 Factory. This software goes with your existing DAW. It mimics the algorithms of a much more expensive and complex device, Eventide’s H3000 Ultra-Harmonizer® effects processor. It changes the pitch of the person’s voice to sound different enough that it’s unrecognizable.

To hear an example, try listening to Laurie Anderson’s fictitious male alter ego, Fenway Bergamot, from her 2010 album, Homeland. She uses Eventide for much of her experimental audio pieces and multimedia shows.

Garageband has pre-set vocal filters, such as Robot or Helium. If you already use Garageband but haven’t played with these filters yet, why not? Why not let go and have fun once in a while? Seriously. Okay, back to protecting your podcast guest.

Pitch correction isn’t difficult but can be distracting for your podcast audience. The vocal distortion stands out more than the words and content. Plus it may make it harder for the audience to understand (transcription is your friend).

Your Own Virtual Voiceover

In 2015, the BBC launched an AI tool called virtual voiceover translation. When the reporter interviews a non-English-speaking subject, the tool transcribes the audio file automatically using AI. Then, the editor re-records the audio with a synthetic voice, speaking English. In the final broadcast, the audience hears the original speaker’s voice, overlaid by the translation.

You can do something similar with Descript. Import the audio file of your guest to a project, transcribe it, and export the transcription to disk. Then, create a new project using the text that you transcribed, and select one of their synthetic voices. Save this and use this as your podcast episode audio.

This is faster and less expensive than hiring a translator. But, Dmitry Shishkin, digital editor at the BBC World Service, said the initial launch of the tool had “incredible pushback.” Artificially intelligent translation services can be inaccurate. “Google translate is not perfect, but workable. It is a good starting point,” Shishkin said. He added that the BBC editors called it a “robots’ invasion.”

Synthetic voices never sound as convincing or motivating as a real human voice, no matter how much you train it. Again, this can be a distraction for your audience, but it will protect your podcast guest.

Voice Acting

Once you’ve completed the interview, transcribe it, and then hire a voice actor to record your interviewee’s dialogue. Collaborate with someone of similar age, gender, and background. Provide them with the transcript, and work with them to ensure they understand the original speaker’s intention and motivation. Then edit the voice artist’s audio into your recording. Before the interview, tell the audience that the voice in the program is a performer to protect your guest’s identity. Put this in the show notes as well.

The voice performer reads the translation with similar emotion and cadence. The audience has a more visceral connection when the voice performer shares the original’s age and gender. This is less distracting than a synthetic voice or filter and can feel more honest.

The Elephant in the Room is pushing a bunny podcaster and a squirrel podcaster away from the microphone. Sometimes one big thing can distract from what's important.

Analysis and Discussion

The easiest way to protect your podcast guest’s identity is to avoid playing their dialogue at all. If you don’t already have a co-host, call in someone who’s knowledgeable about the topic. Use your notes or transcript of the interview to discuss what your guest said. Discussing it with a co-host provides a little bit of distance and air between you and your guest’s answers. This lets you show sympathy for the guest’s information without bias bogging you down.

A drawback here is that for all anyone in your audience knows, you could be making it up. However, keeping your source safe is better than taking extra time with harmonizers or actors if you’re on a deadline.

Deflect and Protect, Don’t Distract

Ultimately, your goal is to share your guest’s story and message. Whichever method you choose to protect your podcast guest, the words need to be clear. Pitch correction can make the audience think, “Wow, that sounds weird,” instead of focusing on your content. Transcription is extra work and cost but makes the information in your podcast accessible to a wider audience. When you have a guest controversial enough to raise the alarm, it’s worth it to give extra effort and attention to make a memorable episode.