How to Generate Automatic Transcripts, Chapter Marks, and Show Notes with Auphonic

Podcasters can generate automatic speech transcripts, timestamped chapter marks, and comprehensive show notes in Auphonic by pairing its audio leveling algorithms with its integrated Automatic Speech Recognition (ASR) and AI structuring tools. Instead of managing separate tools for audio mastering, transcription, and copy generation, podcasters configure these features directly in the Auphonic Production form.

Auphonic processes the audio signal, runs speech-to-text transcription via Whisper or connected third-party engines, extracts thematic sections using language models, and bakes the resulting metadata directly into exported audio files, web players, and distribution feeds.

1. How Auphonic Combines ASR with Content Structuring

Auphonic handles metadata generation during its standard post-production pipeline rather than as a disjointed afterthought:

1. Signal Classification: Before transcription begins, Auphonic analyzes audio signals to distinguish speech from background noise, music, and silence. This prevents non-speech elements from corrupting the speech recognition engine.
2. Speech Recognition (ASR): The segmented speech passes to an ASR engine. Auphonic provides its own self-hosted Auphonic Whisper ASR (based on OpenAI's Whisper model), supporting roughly 100 languages. Because Whisper runs on Auphonic's own infrastructure, audio data stays on their servers, processes rapidly, and incurs no external subscription fees for paying users. External engines such as Amazon Transcribe, Google Cloud Speech API, Wit.ai, and Speechmatics are also supported.
3. AI Content Analysis: When speech recognition completes, an integrated language model analyzes the resulting transcript. It detects thematic shifts, calculates start times for distinct topics, generates chapter headlines, and produces both a long summary and a brief summary, along with relevant episode tags and keywords.
4. Multitrack Diarization: When podcasters upload separate tracks for each host and guest via a Multitrack Production, Auphonic processes each track independently. This eliminates vocal bleed and crosstalk, resulting in higher transcription precision and automatic speaker identification throughout the transcript.

2. Setting Up Automatic Transcripts, Chapters, and Show Notes

Configuring metadata automation in Auphonic takes five core steps within a single production or reusable preset.

Step 1: Upload Source Audio

Create a new Production (for single-file recordings) or Multitrack Production (for multi-mic setups). Upload your raw audio in uncompressed WAV or high-bitrate lossy formats.

Step 2: Configure Speech Recognition

Scroll to the Speech Recognition section in the production form:

• Select Service: Choose Auphonic Whisper ASR. This requires no API keys or third-party accounts.
• Select Language: Pick the primary language spoken in the recording, or rely on Whisper's automatic language detection.
• Speaker Diarization: For single-track recordings with multiple participants, enable diarization to let Auphonic identify voice changes. For multitrack recordings, speaker labels are mapped automatically from individual track names.

Step 3: Enable Automatic Shownotes and Chapters

Tick the checkbox for Automatic Shownotes and Chapters. Once checked, additional customization options appear:

• Shownotes Summary Example: Provide a short sample summary from a previous episode. The AI uses this sample as a stylistic prompt to mirror your customary tone, structure, and formatting (such as bullet points, Markdown headings, or specific callouts).

Step 4: Configure Output Formats and Metadata Mapping

In the Output Files section, select which audio and document formats you need:

• Audio Output: Choose MP3, AAC/M4A, Opus, or FLAC. Auphonic automatically maps chapter titles, start times, summaries, and tags into ID3 tags (MP3), MPEG-4 metadata (M4A/MP4), or Vorbis Comments (FLAC/Opus).
• HTML Transcript: Generates a standalone web document containing the full text, timestamps, speaker tags, and the interactive Transcript Editor.
• Subtitles / Captions (WebVTT or SRT): Generates timestamped subtitle files compatible with web players and video platforms.
• Speech Data (JSON or XML): Exports raw word-level timing and confidence scores for custom automations.
• Chapter Files: Option to export MP4chaps text files, Podlove Simple Chapters (JSON/XML), or Podcasting 2.0 Chapter JSON.
• Burn-In Subtitles (Optional): If generating audiograms or video outputs, check Burn in subtitles to hardcode synchronized captions with selectable fonts, karaoke highlights, and custom layouts directly onto the video.

Step 5: Start Production

Click Start Production. Auphonic executes loudness normalization, filtering, noise reduction, transcription, and AI metadata extraction in a single cloud run.

3. Reviewing and Polishing in the Transcript Editor

Once processing finishes, open the Auphonic Transcript Editor from the production status page by clicking Correct and edit speech recognition results (or by opening the downloaded HTML transcript).

+-----------------------------------------------------------------------+
|  AUPHONIC TRANSCRIPT EDITOR                                  [ Save ] |
+-----------------------------------------------------------------------+
|  AUTOMATIC SHOWNOTES                                                  |
|  Summary:   [ AI-generated episode summary with editable text...   ]  |
|  Keywords:  [ podcasting, auphonic, post-production, metadata     ]  |
+-----------------------------------------------------------------------+
|  CHAPTER MARKS                                                        |
|  00:00:00.000  - Introduction and Welcome                             |
|  00:04:15.500  - Comparing Speech Recognition Engines                 |
|  00:14:30.000  - Managing Metadata and ID3 Tags                       |
+-----------------------------------------------------------------------+
|  TRANSCRIPT                                                           |
|  [00:00:00] Host A: Welcome back to the audio production podcast...   |
|  [00:00:15] Host B: Today we are looking at automated chapters...    |
+-----------------------------------------------------------------------+

The editor provides purpose-built tools for rapid post-production cleanup:

• Confidence Highlighting: Words transcribed with lower confidence appear in red tints. Darker red indicates lower statistical certainty, allowing creators to spot and correct misheard names or technical jargon without reading every line.
• Play on Click: Clicking any word immediately cues audio playback to that exact syllable, making verification instantaneous.
• Paragraph vs. Subtitle/Line Mode: Switch between fluent paragraph reading and precise line-by-line subtitle timing. Press Enter to split a subtitle cue or Backspace at the beginning of a line to merge cues.
• Speaker Management: Click any speaker label to rename the participant across the entire document. In multitrack sessions, saving updates the underlying track identifiers.
• Chapter Adjustments: Refine chapter headlines and summary text inside the Automatic Shownotes panel. To adjust chapter start timestamps, change them directly within the transcript flow to ensure sample-accurate alignment with the spoken transitions.

4. Saving Updates Without Consuming Extra Credits

A common concern in automated workflows is paying twice when fixing transcription typos or adjusting chapter boundaries. Auphonic handles this gracefully:

1. After making corrections in the Transcript Editor, click Save. The production status switches to Production Changed.
2. Click Start Production to re-render the final assets.
3. Because the input audio has not changed, Auphonic applies the updated metadata, chapter marks, and transcript text to all output files without deducting additional processing credits from your account.

[!TIP] When re-running a production solely to refresh edited metadata or chapter titles, disable external paid speech services if any were selected, ensuring no external API costs are incurred.

5. Metadata Distribution and Player Compatibility

Auphonic ensures that the chapters and show notes created in post-production actually reach listeners across major podcast ecosystems:

By consolidating speech recognition, LLM-driven structuring, and metadata packaging into its audio pipeline, Auphonic turns an otherwise tedious post-production chore into a standardized, one-click routine.

References

• Automatic Speech Recognition, Shownotes and Chapters — Auphonic Help 2026 documentation
• Production — Auphonic Help 2026 documentation
• Auphonic Speech Recognition Engine using Whisper by OpenAI (Beta) | Auphonic