Skip to content

coders.md

Latest commit

 

History

History
98 lines (54 loc) · 5.39 KB

coders.md

File metadata and controls

98 lines (54 loc) · 5.39 KB
Raw

Coding Files

QCnext features six specialized coding environments tailored to different media types: Plain Text, PDF documents, Images, CSV/Spreadsheet tables, Webpages, and Audio/Video media.

The left sidebar manages codes and categories. Right-click any code to rename, edit memo, merge into another code, or delete. You can also organize codes in categories and promote/demote them in the right-click menu. Codes can be rearranged via drag and drop. You can also assign custom palette colors to codes. Clicking a code swatch toggles visibility of its highlights in the active coder.

Click Autocode in any coder to open the dialog and code via AI Autocode using natural language or Dictionaries using word list dictionaries to automatically code target terms across documents.

Across all text-based coders (Text, PDF, CSV, Webpage, AV Transcripts), selecting content triggers the floating Selection Toolbar:

Text Coder

Toolbar Actions

  • 🏷️ Code: Immediately codes the selection with the currently active code (selected in the left sidebar code tree). If no code is active, opens the Code Picker.
  • ✨ Code as new code: Creates a brand-new code using the selected text string as its name and codes the passage immediately.
  • 💬 Annotate: Opens a popover to attach a text note directly to the passage without assigning a code.
  • 📥 Send to Crafter: Sends the selected quote directly to a Crafter worksheet as an analytical segment item.

Highlighting & Overlaps

  • Coded passages are highlighted in the color assigned to the code.
  • When multiple codes overlap on the same text, their colors stack.
  • Hovering over a coded segment displays a tooltip with code names and memos; clicking opens the Segment Details Inspector.

Text Coder

The Text Coder is the primary workspace for interview transcripts, field notes, survey text, and imported documents (.txt, .md, .docx).

  • Document Reading & Highlights: Rendered text displays code highlights, annotation underlines (dashed), and outgoing segment links (wavy underline).

  • Live Document Editing: Toggle Edit mode to correct transcription typos directly inside the coder:

    • Text shifts are automatically tracked and debounced.
    • Upon saving (Ctrl/Cmd+S), all character offsets for existing codings and annotations are re-anchored instantly.

  • Code Tree Integration: Single-clicking a code in the left bar sets it as active; double-clicking opens its Inspector details.

PDF Coder

The PDF Coder combines visual page region drawing with extracted plain-text coding in a dual-pane environment.

  • PDF: Shows the PDF view with continuous scrolling or single-page view.
  • Plain text: Display extracted plain text; PDF and Plain text can be shown in split view.
  • Coding: You can code either text or rectangular regions anywhere on a PDF page to code charts, diagrams, formulas, or non-extractable text. Coordinates are stored in vector page relative space.

Image Coder

The Image Coder allows researchers to mark and code graphic material (.png, .jpg, .webp, .svg). Click and drag over photos, diagrams, or scans to mark rectangular regions.

CSV / Table Coder

The CSV / Table Coder provides cell-level text coding for survey data, tabular datasets, and social media exports.

  • Cell-Level Character Coding: Code specific words or phrases inside individual cells rather than marking entire table rows.
  • Plain Text Fallback: Toggle to view the raw tabular document in the standard text coder if preferred.

Webpage Coder

The Webpage Coder handles HTML page captures and web articles imported via URL.

  • Split View Layout: Displays the raw rendered HTML snapshot alongside clean extracted article text, similar to PDFs.
  • Coding: Select text on the rendered webpage to code it directly like in other text modes.

Audio / Video Coder & Transcripts

The Audio / Video Coder integrates media playback, timeline range coding, transcript synchronization, Whisper automated transcription, and speaker detection.

Playback & Timeline Coding

  • Transport Row: Play/pause, playback speed selector (0.5× to 2.0×), and a clickable timeline.

  • Keyboard & Media Keys: Space toggles play/pause, F9 inserts timestamps, and hardware media keys control playback.

  • Timeline Range Coding:

    1. Click Set start at the desired playback timestamp.
    2. Play or seek to the end timestamp.
    3. Click Set end and code to code the millisecond interval (pos0 to pos1).
    4. Coded ranges appear as interactive colored blocks on the media timeline.

Transcribe

  • Click Transcribe to launch background transcription powered by Whisper models.
  • Options: Model size, language selection, translation to English, Voice Activity Detection (VAD), and automatic turn coding.
  • Live Preview: Progress streams live into the transcript pane as \[mm:ss] text while transcription runs in the background queue.

Manual Transcription & Speaker Detection

  • Manual Mode: Turn on Transcribe mode to type transcripts manually. Pressing Enter or F9 auto-inserts timestamp markers \[mm:ss].
  • Speaker Detection: Automatically detect speaker turns from transcript identifiers (e.g., Speaker 1:, \[Alice], {@Bob}) or custom regex patterns. QCnext automatically generates speaker codes and codes every turn across the transcript.