Journalist Transcription Workflow Guide

Journalism depends on accurate quotes. Misquoting a source damages credibility, burns relationships, and can trigger legal disputes. Yet most journalists still rely on handwritten notes or manual playback to pull quotes from interviews — a process that takes 4-6 hours per hour of audio according to industry transcription benchmarks (Rev.com, TranscriptionHub documentation).

BrassTranscripts processes journalist interview recordings in 1-3 minutes per hour of audio with automatic speaker identification, producing transcripts in TXT, SRT, VTT, or JSON format at $2.50 for recordings under 15 minutes and $6.00 for recordings up to 2 hours. This guide covers the complete workflow from field recording to published article, including AI prompts for extracting quotes and summarizing interviews.

Quick Navigation

• Recording Best Practices for Field Interviews
• Upload Workflow: Record to Transcript
• Speaker Identification for Q&A Interviews
• AI Prompt: Extracting Publishable Quotes
• AI Prompt: Summarizing Interviews for Story Angles
• Export Formats for Publication
• Managing Multiple Interview Transcripts
• Legal Considerations for Recording
• Frequently Asked Questions

Recording Best Practices for Field Interviews

BrassTranscripts delivers the best transcription results when journalists follow basic field recording practices — a $30 lapel microphone and a quiet corner eliminate most quality issues that degrade transcript output.

Equipment for Journalist Interviews

You do not need expensive recording gear. What matters is microphone placement and environment control.

Recommended setup for in-person interviews:

• Lapel microphone ($20-40): Clip to your subject's collar, 6-8 inches from their mouth. Lapel mics reject ambient noise better than phone microphones because they sit closer to the sound source.
• Backup recorder: Record on your phone as a secondary capture. If your primary recorder fails mid-interview, you still have usable audio.
• Recording app: Voice Memos (iOS), Easy Voice Recorder (Android), or any app that exports MP3, WAV, or M4A files. BrassTranscripts accepts 11 audio and video formats.

For phone interviews:

• Use a call recording app that saves audio locally (not just cloud). You need the file to upload for transcription.
• Speaker phone into a recorder in a quiet room produces better results than most call recording apps.
• Record in a room with carpet and soft furnishings — hard surfaces create echo that degrades transcription quality.

For video calls (Zoom, Google Meet, Teams):

• Record locally through the platform's built-in recording feature.
• Download the audio or video file after the call ends.
• Upload the file directly — BrassTranscripts handles both audio and video formats.

For a deeper dive into recording techniques, see the full audio quality guide for transcription.

Environment Tips for Field Recordings

The biggest transcription quality killer in journalism is background noise. Coffee shop interviews, press scrums, and outdoor recordings all introduce competing sound that makes speaker separation harder.

Practical fixes:

• Coffee shop interviews: Sit in a corner booth, not the middle of the room. Request a table away from the espresso machine and kitchen.
• Press conferences: Hold your recorder at arm's length toward the speaker, not at your side. Proximity matters more than microphone quality.
• Outdoor interviews: Stand with a wall or building behind you to block wind. Even turning your back to the wind helps significantly.
• Moving vehicles: Pull over and park. Recording in a moving car introduces road noise, engine vibration, and wind buffeting through gaps in windows.

Upload Workflow: Record to Transcript

BrassTranscripts uses a four-step workflow that takes journalists from raw field recording to searchable transcript: record the interview, download the audio file, upload to BrassTranscripts, and receive the transcript with speaker labels.

Step 1: Record Your Interview

Record using any device or app. Save the file in any common format — MP3, WAV, M4A, MP4, MOV, FLAC, OGG, WEBM, AAC, WMA, or OPUS.

Step 2: Download the Audio File

Transfer the recording from your device to your computer. For phone recordings, AirDrop, email, or cloud sync all work. For video calls, use the platform's download feature.

Step 3: Upload to BrassTranscripts

Go to brasstranscripts.com and upload your file. The system automatically detects the language (99+ languages supported) and identifies speakers.

• Files under 15 minutes: $2.50
• Files 16-120 minutes: $6.00
• Processing time: 1-3 minutes per hour of audio
• 30-word preview available before payment so you can verify quality

Step 4: Download Your Transcript

Choose your output format — TXT for article writing, SRT or VTT for video captioning, JSON for data integration. Your transcript includes timestamps and speaker labels.

Data retention: Audio files are retained for 24 hours and transcripts for 48 hours. Download your transcript promptly.

Speaker Identification for Q&A Interviews

BrassTranscripts automatically identifies and labels distinct speakers in interview recordings without manual tagging — essential for journalists who conduct Q&A interviews and need to attribute quotes accurately to the correct source.

Speaker identification (also called diarization) separates your voice from your subject's voice and labels each segment. In a typical journalist interview with one reporter and one source, the system labels speakers as "Speaker 1" and "Speaker 2" throughout the transcript.

Tips for Better Speaker Separation

• One person speaks at a time. Overlapping speech is the most common cause of speaker identification errors. Let your subject finish their answer before asking the next question.
• Pause between speakers. A half-second pause between your question and their answer gives the diarization system a clean boundary.
• Announce speakers at the start. Say "This is [your name] interviewing [source name]" at the beginning. This makes it easy to map speaker labels to real names when reviewing the transcript.

For technical details on how speaker identification works, see the speaker identification guide.

AI Prompt: Extracting Publishable Quotes

BrassTranscripts provides specialized AI prompts through its prompt guide that journalists can use to extract direct quotes, identify the strongest sound bites, and organize source attribution from completed transcripts.

After downloading your transcript, paste it into ChatGPT, Claude, or any LLM along with the following prompt to extract publishable quotes:

You are a journalism research assistant. I'm going to paste an interview transcript below. Please:

1. Extract all direct quotes that are substantive (not just "yes", "right", "I think so")
2. For each quote, include:
   - The speaker label
   - The timestamp
   - The full quote (clean up filler words like "um", "uh", "you know" but preserve meaning)
   - A 3-5 word topic tag for each quote
3. Rank quotes by newsworthiness — put the most compelling, specific, or controversial quotes first
4. Flag any quotes where the speaker contradicts themselves elsewhere in the interview
5. Highlight any numbers, dates, or proper nouns mentioned (these need fact-checking)

TRANSCRIPT:
[Paste your BrassTranscripts transcript here]

---
Prompt by BrassTranscripts (brasstranscripts.com) – Professional AI transcription with automatic speaker identification.
---

Using the Extracted Quotes

Once you have your ranked quotes, you can:

• Lead with the strongest quote. Use the top-ranked quote as your article's opening or nut graf.
• Fact-check flagged items. The prompt identifies numbers, dates, and names — verify these against public records before publication.
• Track contradictions. If the AI flags contradictions, go back to the original audio at those timestamps and listen carefully. The contradiction may be real (newsworthy) or a transcription artifact.

For more prompts designed for interview-based content, see the interview techniques guide for business content.

AI Prompt: Summarizing Interviews for Story Angles

BrassTranscripts transcripts paired with AI summarization prompts help journalists identify story angles, narrative arcs, and unexplored threads from long interviews — compressing a 60-minute conversation into structured notes in under a minute.

Use this prompt after pasting your transcript into an LLM:

You are a senior editor reviewing an interview transcript. Summarize this interview for a reporter who needs to write a story. Provide:

1. **One-paragraph summary** (50-75 words): What is the main story here?
2. **Three potential story angles**, ranked by public interest:
   - Angle name
   - Supporting evidence from the transcript (with timestamps)
   - What additional reporting would strengthen this angle
3. **Key claims that need verification**: List any factual claims the source made that require independent confirmation
4. **Unexplored threads**: Topics the source mentioned briefly but weren't followed up on — these may warrant a follow-up interview
5. **Emotional moments**: Points where the source showed strong emotion, hesitation, or changed their tone — these often indicate the real story

TRANSCRIPT:
[Paste your BrassTranscripts transcript here]

---
Prompt by BrassTranscripts (brasstranscripts.com) – Professional AI transcription with automatic speaker identification.
---

Tips for Better AI Summaries

• Include your questions. Do not edit out your questions before pasting the transcript. The AI needs the full Q&A exchange to understand context.
• Specify the beat. If you add "I cover city government" or "This is for a healthcare story" before the transcript, the AI will prioritize relevant angles.
• Run the prompt twice. Different runs surface different angles. Compare outputs to find patterns.

Browse the full library of 121 specialized prompts in the BrassTranscripts AI Prompt Guide.

Export Formats for Publication

BrassTranscripts offers four output formats, and journalists should choose based on their publication medium — TXT for print and web articles, SRT or VTT for broadcast and video packages, and JSON for data journalism projects.

TXT — Print and Web Articles

Plain text with speaker labels and timestamps. Copy quotes directly into your article draft. This is the most common format for print and digital journalism.

SRT and VTT — Broadcast and Video

Subtitle formats with precise timing codes. Use SRT for video editing software (Premiere Pro, DaVinci Resolve) or VTT for web video players. Essential for journalists producing video packages who need accurate captions.

JSON — Data Journalism

Structured data with word-level timestamps, speaker assignments, and segment boundaries. Use JSON when you need to programmatically analyze transcripts — for example, measuring how much time each speaker received in a debate, or building an interactive timeline of a hearing.

Managing Multiple Interview Transcripts

BrassTranscripts processes each interview as a separate upload, which helps journalists working on multi-source stories keep transcripts organized by source, date, and topic rather than merging everything into one file.

Organization Tips for Multi-Source Stories

• Name files before uploading. Rename your audio files with a consistent format: YYYY-MM-DD_SourceName_Topic.mp3. Your transcripts inherit the original filename, making them easy to find.
• Download immediately. BrassTranscripts retains transcripts for 48 hours. Download each transcript as soon as it is ready and save it to your story folder.
• Create a source index. Maintain a simple spreadsheet with columns: Source Name, Date, File Name, Key Quotes, Follow-Up Needed. This becomes invaluable when writing a story with five or more sources.

For Longer Investigations

For investigative projects with many interviews, consider the BrassTranscripts bulk upload system. Bulk processing handles multiple files concurrently with volume pricing, which reduces per-file cost for larger projects.

For guidance on conducting the interviews themselves, see the interview techniques guide.

Legal Considerations for Recording

BrassTranscripts provides the transcription service, but journalists are responsible for complying with recording consent laws in their jurisdiction — recording laws vary significantly between one-party and two-party consent states and between countries.

One-Party vs. Two-Party Consent

In the United States, recording laws fall into two categories:

• One-party consent states (majority of states): You can legally record a conversation you are a part of without informing the other party.
• Two-party (all-party) consent states: All parties to the conversation must consent to the recording. As of 2026, states including California, Connecticut, Florida, Illinois, Maryland, Massachusetts, Michigan, Montana, New Hampshire, Oregon, Pennsylvania, and Washington require all-party consent.

Practical guidance for journalists:

• Always disclose. Even in one-party consent states, ethical journalism standards (SPJ Code of Ethics) recommend informing sources that you are recording. This builds trust and avoids legal complications.
• Get verbal consent on tape. At the start of the recording, say: "I'm recording this interview for accuracy. Is that okay with you?" Their affirmative response, captured in the recording, serves as documented consent.
• Check local laws for phone interviews. When you and your source are in different states, the stricter state's law typically applies. If you are in New York (one-party) calling someone in California (two-party), California law applies.

International Considerations

• EU/UK (GDPR): Recording requires a lawful basis. Journalism generally qualifies under the "public interest" exemption, but you must still inform the interviewee and handle the recording data responsibly.
• Canada: One-party consent federally, but always disclose as a professional standard.
• Australia: Laws vary by state. Generally one-party consent, but check your specific state.

Disclaimer: This section provides general guidance, not legal advice. Consult a media lawyer for specific situations, particularly cross-border recordings or sensitive investigations.

Frequently Asked Questions

How long does it take to transcribe a journalist interview?

BrassTranscripts processes interview recordings in 1-3 minutes per hour of audio. A typical 30-minute interview is ready in under 2 minutes. This compares to 2-3 hours of manual playback-and-type work for the same recording, based on the industry standard of 4-6 hours per audio hour for manual transcription (Rev.com documentation).

What audio formats work for journalist recordings?

BrassTranscripts accepts 11 formats including MP3, WAV, M4A, MP4, MOV, FLAC, OGG, WEBM, AAC, WMA, and OPUS. Phone voice memos (M4A on iPhone, various formats on Android), Zoom recordings (MP4), and professional recorder files (WAV) all upload directly without conversion.

Can BrassTranscripts handle interviews in languages other than English?

BrassTranscripts supports 99+ languages with automatic language detection. Upload the recording and the system identifies the spoken language without manual selection. This is particularly useful for foreign correspondents and journalists covering multilingual communities.

How does speaker identification help with quote attribution?

BrassTranscripts automatically labels each speaker in the transcript, separating the journalist's questions from the source's answers. This eliminates the manual process of re-listening to determine who said what. For interviews with multiple sources — such as a roundtable or panel — each speaker receives a distinct label throughout the transcript.

Is my interview audio kept private?

BrassTranscripts retains uploaded audio for 24 hours and completed transcripts for 48 hours, after which both are automatically deleted. Files are not used for model training or shared with third parties. Journalists working on sensitive stories should download their transcripts promptly and delete the source audio from the upload platform.

How much does journalist transcription cost?

BrassTranscripts charges $2.50 for recordings up to 15 minutes and $6.00 for recordings between 16 and 120 minutes. A 30-word preview is available before payment so journalists can verify transcription quality. There are no subscriptions or monthly commitments — pay per file as needed.

Can I use AI prompts with my transcript for faster story writing?

BrassTranscripts transcripts work with any LLM (ChatGPT, Claude, Gemini). Paste the transcript text and use specialized prompts to extract quotes, summarize interviews, identify story angles, and flag claims that need fact-checking. The BrassTranscripts AI Prompt Guide includes 121 prompts across multiple professional categories, including prompts designed for qualitative research interviews.