Translate a Riverside podcast or livestream in real time

Riverside is the recording tool of choice for podcasters who want broadcast-quality audio and video without a studio. Each guest records locally on their own machine; Riverside uploads the local tracks to the cloud after the session ends. The host hears the guest over a live monitoring stream, but the audio that gets published comes from the per-speaker local files — not the compressed monitoring feed.

Riverside also offers livestream-while-recording: the same session that’s being recorded can be pushed to YouTube, X, or a custom RTMP destination in real time. Live listeners hear the monitoring-quality stream; the publisher gets the studio-quality recording for post-production.

For an international audience, both modes — the live broadcast and the published podcast — face the same gap. The live audience that doesn’t share the host’s language has historically had nothing. The published podcast can be subtitled or dubbed after the fact, but that’s a slow workflow and the live audience misses out either way.

What translates in real time vs after the fact

Loquira sits parallel to Riverside, the same way it does for OBS or Zoom. The host’s microphone feeds Riverside for the recording and the livestream; a phone running Loquira reads the same microphone for translation. The translated audio reaches the live audience in real time; the high-quality podcast recording is unaffected and can still be dubbed or subtitled in post if you want a polished version for the published episode.

This means Riverside hosts have a clean separation:

• Live audience — Loquira handles them. Translated audio reaches them within a second of the host speaking.
• Published episode — Riverside’s local-recording quality is preserved. Post-production translation, dubbing, or subtitling continues to work as it always has.

Both audiences are served without compromising either workflow.

Recommended setup

• Microphone. Whatever you already use for Riverside — typically a USB condenser like a Shure MV7 or a dynamic mic like a Procaster through an interface. Riverside is uniquely picky about USB mic stability; nothing about adding Loquira affects that.
• Loquira device. A phone on a small tripod on the desk, positioned near (but not blocking) your microphone. The phone’s mic reads your voice in parallel with the studio mic feeding Riverside.
• For remote guests. Each guest connects to Riverside from their own location with their own microphone. If a guest wants their voice translated for the live audience, they run their own Loquira session locally. The host coordinates by sharing the Loquira “session group” in the show notes.
• For listen-along live audiences. Post the Loquira join link in the destination chat where the livestream is going. For YouTube, paste it into the chat at the start of the broadcast and re-pin it every ten minutes.

Per-guest Loquira sessions: when and why

Riverside’s appeal is interview-format content where a host talks with one or two guests. Translation works cleanly when only the host speaks the source language and the guests speak the same — the host runs Loquira and the conversation flows naturally.

When guests speak different languages — a common pattern for international podcasts — each speaker who wants translation runs their own Loquira session. The host runs an EN-source session that translates into the listener’s chosen language; a guest speaking Spanish runs an ES-source session that translates into the same set of languages. Listeners pick which speaker they want to follow from the Loquira interface and switch when the conversation handoffs.

This sounds heavier than it is. Each guest already has their own setup for Riverside; adding a phone with Loquira beside them is the same complexity as adding earbuds. The host’s job is to coordinate the session-group link ahead of the recording.

Known limitations

• Recording quality vs translation timing. Riverside’s published audio is post-processed for clarity, normalisation, and noise reduction. Loquira’s translation is real-time from the unprocessed source. If you want translated audio in the published podcast, you’ll need to re-translate from the post-processed file in a separate workflow.
• Guest-side microphone variability. Riverside compensates for inconsistent guest microphones through post-processing. Loquira sees the raw signal in real time. A guest with a noisy environment or a poor microphone will produce worse live translation than the published audio suggests. The audio requirements page covers the floor.
• Magic Clips and AI features. Riverside’s AI clip generation and transcript tools operate on the recorded files post-session. They don’t compete with Loquira’s live transcript; they complement it for the published version.
• Studio mode latency. Riverside’s “studio mode” prioritises recording quality over monitoring latency. The host hears the guest with a small delay through the monitoring stream. Loquira reads the host’s voice directly from the microphone, not from Riverside’s monitoring output, so this doesn’t affect translation latency.