Real-Time Deposition Translation —
Follow the Testimony Live

MirrorCaption streams word-by-word translation during depositions — no bot, no install, works alongside your certified interpreter in Zoom or any browser.

MirrorCaption streams real-time legal deposition translation under 500ms — you read each answer as the witness finishes speaking, not 30 seconds later when the consecutive interpreter renders it.

Remote depositions are routine now. What hasn't changed is the interpreter workflow: consecutive translation, thirty-second gaps, rhythm broken every other sentence. You're reading body language on a 720p screen while processing meaning on a delay. MirrorCaption adds a third track to your Zoom deposition: a word-by-word translation stream running in a browser tab while the certified interpreter handles the official record.

Key Takeaways

Why Deposition Translation Is Hard to Get Right

Deposition translation is the rendering of a witness's spoken testimony from one language into another during a legal proceeding. Interpreter requirements vary by court, forum, stipulation, and state law, and the official record still demands precision, accountability, and admissibility.

The problem isn't the interpreter's skill. It's the mechanics of how interpretation works in practice.

Consecutive vs. Simultaneous Interpretation

Most depositions use consecutive interpretation: the witness speaks, stops, then the interpreter renders the complete answer in English. This creates a gap of 30–90 seconds on every exchange. The attorney processes the answer, formulates a follow-up, and then discovers the witness has already hedged or added a qualifier that changes everything — but the moment to probe it has passed.

Simultaneous interpretation eliminates the gap. An interpreter renders the testimony in real time, typically through earphones, while the witness speaks. It's accurate and fast. It also costs 2–3x more than consecutive, often requires two interpreters working in relay for long sessions, and is reserved for high-stakes depositions where the budget justifies it.

Most firms default to consecutive and absorb the rhythm disruption as the cost of doing business.

The Remote Deposition Problem

Remote depositions over Zoom or Teams add another layer. You're already working with compressed audio, network latency, and reduced visual cues. When interpretation is consecutive, the exchange becomes: question — English — witness — foreign language — pause — interpreter renders — English. You are three communicative steps removed from what the witness actually said, and you get there thirty seconds late.

For common deposition languages — Spanish, Mandarin, Vietnamese, Korean, Arabic, Russian — that lag is structurally built into the proceeding. There has been no good technological answer to it. Until browser-based real-time STT caught up.

How MirrorCaption Fits Into a Deposition Workflow

Scenario

Sarah Chen, a litigation associate in Seattle, spent three hours deposing a Mandarin-speaking witness over Zoom. Her certified consecutive interpreter was excellent. But each exchange meant waiting while the interpreter collected the full answer and rendered it. At the two-hour mark, the witness gave a long answer. The interpreter translated: "He says it was his understanding the contract was verbal."

Sarah had a feeling the witness said more than that. She didn't know how to push back. The moment had passed.

The following week, before her next bilingual deposition, Sarah opened MirrorCaption in a second browser window alongside Zoom. As the Mandarin-speaking witness answered, she read the translation streaming word by word. When she saw "...verbal agreement, but also WeChat message" appear on screen while the interpreter was still writing notes, she knew exactly what follow-up to ask — and asked it before the interpreter had finished rendering the official version.

MirrorCaption works as a comprehension layer alongside your certified interpreter — not a replacement for them. Here is the workflow during a remote deposition.

During the Deposition

  1. Open MirrorCaption in a browser tab on your laptop or second screen — no installation, no account setup beyond login.
  2. Select Meet Mode and share your Zoom (or Teams, or Meet) browser tab as the audio source.
  3. Set languages: source language (the witness), target language (your reading language).
  4. Watch translation stream word by word as the witness speaks — under 500ms latency means the caption appears while they are still talking.
  5. The certified interpreter continues uninterrupted — they handle the official record, you handle your comprehension.

MirrorCaption never joins the call as an attendee. Nothing appears in the participant list or waiting room. The audio passes through your browser's local tab via the getDisplayMedia API — no third-party software touches the meeting.

For real-time translation in remote meetings beyond depositions, the same setup works for any multilingual video call.

Before and After the Deposition

Before: use MirrorCaption in client prep sessions. Your client speaks in their native language — Spanish, Mandarin, Vietnamese — and you read the translation live as they walk you through what happened. No fumbling with Google Translate or asking them to slow down for you.

After: export the session transcript to search, skim, and review before the certified translation arrives. MirrorCaption's AI summary gives you a structured overview of the key exchanges while the details are still fresh — useful for deposition digests, pre-trial prep, and next-day follow-up questions.

Try MirrorCaption free — 2 hours/month, no credit card →

Five Legal Workflows Where MirrorCaption Adds Value

📋

Deposing a Foreign-Language Witness

Follow testimony live on Zoom or Teams while your certified consecutive interpreter handles the official record. Catch hedging, qualifiers, and nuance before the interpreter finishes rendering.

🤝

Client Prep Before the Deposition

Prepare your non-English-speaking client in their native language. Read their account in real time as they speak — faster and more natural than typing into Google Translate.

🌎

International Arbitration Proceedings

Follow multilingual arbitration hearings live without commissioning a full simultaneous interpreter setup. Useful for proceedings where you need comprehension, not a certified record.

🔐

Immigration and Asylum Interviews

Conduct client intake in the client's native language. Read the translation while your client speaks — no interpreter scheduling, no billing delay, no third-party present in a sensitive conversation.

A fifth use case worth noting: post-deposition transcript review. When a foreign-language deposition transcript arrives, running it as a dictation or read-along through MirrorCaption gives you a translated working copy while you wait for the certified translation. Not admissible. Useful for internal prep.

Privacy and Data — What Attorneys Need to Know

Vague privacy assurances are not enough for legal work. Here is the architecture precisely, so you can make your own assessment.

For detail on how AI meeting tools handle data, see our post on AI meeting privacy. If your firm has specific data governance requirements covering third-party STT vendors, review them with your privacy counsel before use.

Important: MirrorCaption Is Not a Certified Interpreter

Important Limitation

MirrorCaption generates AI translation in real time. It is not a certified court interpreter. Its output is not a court record, is not admissible as evidence, and must not be used as a substitute for a certified interpreter in sworn proceedings.

If your matter, forum, or governing rules require a certified interpreter for the official record, MirrorCaption does not change that. It is a parallel comprehension tool for counsel, not the official interpreter.

What MirrorCaption does: it gives the deposing attorney a live reading stream — a comprehension layer running in parallel with the official interpreter. The court reporter records what the certified interpreter says. You read what you need in order to stay in real time.

Think of it as having a bilingual colleague sitting next to you who whispers a rough translation as the witness speaks. That colleague's whisper is not the record. Your certified interpreter is the record. But the whisper lets you stay in the conversation instead of trailing it by thirty seconds.

Used as a comprehension assist, it's a legitimate litigation tool. Used as a replacement for a certified interpreter, it's not. The distinction matters, and we're stating it plainly so attorneys can make an informed decision.

For a benchmark of how accurate real-time AI translation is in practice, see our analysis of real-time translation accuracy.

What It Costs

Option Typical Cost What You Get
Simultaneous interpreter (4h deposition) $800–$1,600+ Official interpretation of sworn testimony, real-time delivery, admissible record
Consecutive interpreter (4h deposition) $400–$800 Official interpretation, 30–90s lag per answer, admissible record
MirrorCaption Lifetime €49 once Real-time comprehension assist for the attorney, 200 hours included, AI summary, transcript export — not a certified interpretation

MirrorCaption does not replace the certified interpreter you need for the official record. It replaces the thirty-second wait you spend not knowing what was just said.

Start with 2 free hours per month — no credit card required. Test it in a prep session with your client before using it in a live deposition. If you need more than 2 hours/month, the Lifetime plan is €49 once for 200 hours, or €29/year for 100 hours. No per-seat pricing, no subscription that auto-renews.

Frequently Asked Questions

Is AI translation admissible in court?

No. MirrorCaption's AI-generated translation is not a certified translation and is not intended for court records or evidence. If your matter, forum, or governing rules require a certified interpreter for the official record, continue to use one. MirrorCaption is a comprehension tool for the attorney — not a record for the court.

How much does a deposition interpreter typically cost?

Consecutive interpretation typically runs $100–$200/hr for common language pairs (Spanish, Mandarin, Vietnamese). Simultaneous interpretation runs $200–$400/hr and usually requires two interpreters working in relay for sessions over two hours. A full-day deposition with a simultaneous interpreter can cost $1,500–$3,000 or more before travel is factored in.

Can MirrorCaption replace a certified interpreter at a deposition?

No. MirrorCaption is a real-time comprehension assist for the attorney — it can complement an interpreter workflow, but it is not a substitute for any certified interpreter your matter, forum, or governing rules may require. This distinction is explicit and non-negotiable.

What deposition platforms does it work with?

MirrorCaption works with any browser-based audio source. For remote depositions: Zoom, Microsoft Teams, Google Meet, Webex, and any other platform your deposition software runs inside a browser tab. For in-person depositions where the witness is in the room, Talk Mode captures microphone audio directly — no video call required.

Is my client's testimony kept private?

Audio is processed via Soniox's STT engine for real-time transcription and is not retained after processing — no audio is stored on MirrorCaption's servers. Transcript text is saved to your local browser storage (IndexedDB), not to any cloud service. No software joins the call as a participant or appears in the participant list. If your firm has specific data governance requirements covering third-party STT vendors, review those policies before use.

Do I need to install anything?

No. MirrorCaption is a browser-based Progressive Web App. Open it in Chrome, Edge, or Safari — no download, no browser extension, no IT approval required. It runs entirely within your browser session and leaves nothing on the machine when you close the tab.

Follow Every Word. In Real Time.

Start with 2 free hours every month. No credit card. No installation. Works in your next Zoom deposition today.

Try MirrorCaption Free