Link Transcription

TODO (New Screenshot): Link transcription task queue (App 2.0) Include: URL input area, task status columns (recognizing/downloading/preparing/done), and retry controls. Suggested filename: link-transcription-queue-v2-en.png

Scope

Link Transcription is an end-to-end URL workflow:

• Link parsing and platform recognition
• Audio/video format selection
• Download progress and retry handling
• Download-to-transcription handoff

Use Cases

• Public content processing from supported platforms
• Podcast/live replay text extraction
• Batch URL ingestion pipelines

Steps

1. Paste one or multiple URLs in the Home input (one URL per line).
2. Click Start Transcription and wait for recognition.
3. Select desired formats for each task.
4. Start downloads and monitor stage/status transitions.
5. Retry failed tasks and continue to transcription once ready.

Preflight Checklist

• Confirm network/proxy access to target platforms.
• Verify link engine installation and readiness in Settings.
• Run 1–2 URLs as a smoke test before launching large batches.
• Ensure output and cache paths have enough free disk space.

Term Explanations

• Recognition: URL parsing and support validation phase.
• Task status: each stage reflects a different failure domain for debugging.
• Retry: usually re-runs failed stages, not always full restart.

Example Workflow (Podcast Replay Batch)

1. Paste 3–5 replay URLs and select audio-only format.
2. After downloads complete, run batch transcription with Small/Medium first.
3. Use Note to split sections, summarize, and export Markdown/subtitles.

Real Scenario: Weekly Content Operations Review

Many teams run a fixed weekly link-ingestion cycle:

1. Collect replay URLs before the review session.
2. Validate 1–2 sample URLs before launching full batches.
3. Route final transcripts into workspace-specific folders by function.

This prevents large-batch failures discovered too late and improves downstream reuse across teams.

Common Mistakes

• Mistake 1: Launching large batches without smoke tests.
  Fix: validate a small sample first, then scale up.
• Mistake 2: Re-running everything after partial failures.
  Fix: use task status to retry only failed stages.
• Mistake 3: Ignoring storage and naming conventions.
  Fix: predefine directory rules and disk budget before batch runs.

FAQ

Q: Is multi-link mode available for all plans?
A: Multi-link and parallel capacity depend on account entitlements.

Q: Recognition succeeds but download fails. What should I check?
A: Network, proxy, disk space, and format compatibility are the top checks.

Q: Why is a URL reported as unsupported?
A: Engine capability, platform policy, and regional availability all affect support.

Troubleshooting Order

1. Validate source URL availability and access policy first.
2. Validate network/proxy/DNS environment next.
3. Validate chosen format compatibility for the specific source.
4. Validate local capacity (disk/concurrency/background load) last.

Limitations

• Status: Stable (non-Beta), while supported sites/policies may change by version.
• Requires a working link engine installation/version.
• Platform support and accessibility can change over time.
• Large batch jobs can consume significant disk and network resources.
• Platform: Available on both Windows and macOS, with possible differences in proxy and download-toolchain behavior.