In this episode, Pilar shares her failed attempts at releasing an episode in Spanish in a dubbed version, and what she’s ended up doing instead.
Hello, it’s been a while, and there’s a reason for that.
I recently interviewed Cristian Curto and Roberto Schlesinger, the hosts of the Spanish-language podcast “La Gran Renuncia” (The Great Resignation), all about remote work. We did the show in Spanish, as they’re more comfortable speaking in that language than in English, and I thought, hey, with the explosion of generative AI, I’ll be able to come up with something to offer to my regular listeners!
(The episode is now out, it’s episode 75.)
I initially used Eleven Labs, but I found the results disappointing due to the strong accents and inconsistencies in the AI-generated voices. I even experimented with dubbing my own voice, which was a surreal experience hearing myself with a completely different accent.
You can listen to some samples of that, at around 07.11mins, and here’s the video I refer to there.
I then thought of translating the transcript, so I looked for a good transcription solution. I explored various options, including Riverside.fm, Apple Podcasts, and Spotify for Podcasters. However, each platform had its limitations, such as the inability to download transcripts or the absence of speaker labels.
In the end, I transcribed the conversation using AssemblyAI, copied the transcript into Google Docs, and then used the built-in translation tool to convert the text from Spanish to English. The resulting translation was quite good, with only minor issues like incorrect name translations. I’ve fixed some of it, but not all, and I have left the US spelling in there.
If you have any experience with dubbing voices or offering different versions of an episode, I’d love to hear from you.