Troubleshooting

Troubleshoot recording, network & transcription issues

Fix Vowise recording, network, upload, and transcription problems without losing a retained recording. Learn what is saved locally, what still needs connectivity, and how to retry safely.

Audience: Users who hit a blocked workflow and need a calm recovery path.

Time: 16 minutes

Level: All levels

Scenario

Start with the workflow this guide is designed for before moving into the steps.

As a user preparing a real workflow, I want a checklist for common failures so I can fix the problem without losing the recording or exposing private data.

You can identify the likely failure layer and collect safe details for support.

Steps

Classify the failure

Decide whether the problem is capture, upload, transcription, template output, API auth, or review.

You can name the first failing step.

You avoid changing unrelated settings at random.

Collect safe support details

Capture screenshots or logs that show the error, but redact keys, private names, and raw sensitive transcripts.

No credentials appear in the support details.

The timestamp and failing action are visible.

Retry the smallest safe step

Use a small sample recording or placeholder API request instead of repeating the full high-stakes workflow.

The retry is small enough to repeat.

The result tells you whether the failure moved or stayed.

Why does my transcript look wrong?

Start by finding the first layer that became wrong. If the audio is missing or unclear, fix capture or input first. If the raw transcript repeatedly mishears the same name, acronym, or product term, compare it with the source and test a focused Dictionary entry if your current transcription path supports it. If the raw transcript is acceptable but the final note changes structure, emphasis, or wording, inspect the template or AI-output layer instead. Change one layer at a time, test on a small real sample, and verify names, numbers, dates, quotes, decisions, and owners against the source before reuse.

Symptom → first layer to check

What you see	First layer to check	What to do first	Useful next guide
Audio is missing, extremely quiet, or clearly the wrong input	Capture / input	Check the microphone, source, permission, and the saved source before changing AI settings	Make your first recording
Raw transcript gets ordinary speech wrong across much of the clip	Audio / transcription	Compare the raw transcript with the source; test a shorter, clearer sample with the current input and language settings	Make your first recording
Raw transcript mainly misses recurring names, brands, acronyms, or technical terms	Vocabulary	Keep the source; on a supported path, test only repeated, high-value terms in Dictionary	Create a custom dictionary
Raw transcript is mostly right, but headings, action items, summary shape, or formatting are wrong	Template / AI output	Compare raw transcript with shaped output; change output instructions rather than rewriting uncertain source words	Use prompt templates
Output sounds fluent, but an important quote, date, number, owner, or decision may be wrong	Verification	Return to the original audio/transcript and label uncertainty instead of filling gaps by inference	Make your first recording / Chat with records
Processing is stuck, fails, or the expected output never appears	Processing / workflow	First check whether the source or saved record is available on the current supported surface; then retry the smallest safe step	Troubleshoot common issues

Change one thing, then test again

Keep the source and compare each result before retrying a longer workflow.

1. Keep the original source. Do not replace the only copy of an important recording because the first output looks wrong.
2. Find the first wrong layer: audio, raw transcript, recurring vocabulary, shaped output, or later interpretation.
3. Change one variable: the input, one recurring term, or one output instruction.
4. Use a short representative sample with the same microphone setup, language pattern, and important terms.
5. Compare with the source. Check whether the specific failure improved before retrying a long or consequential recording.
6. Before reuse, verify names, exact quotes, numbers, dates, decisions, obligations, owners, and deadlines.

Examples: vocabulary, output, and capture

These are illustrative diagnoses, not measured product results.

If PostHog is repeatedly misheard while ordinary speech is readable, test its correct spelling as one focused Dictionary entry on a supported path, then compare a new small sample.
If the raw transcript never assigned a task to Alex but a polished note says Alex owns it, inspect output instructions. Remove unsupported completion; do not change the raw transcript to match the note.
If the recording is blank or very quiet, check capture/input and the saved source first. Dictionary or Template changes cannot repair speech that was never captured clearly.
For a mixed-language sentence, check the current platform/build language and input behavior, then compare with the audio. Do not assume every platform exposes the same controls.

What these checks can and cannot fix

Vowise guidance separates capture, transcript review, vocabulary, and output shaping. Exact availability follows your current platform and build.

Dictionary can help with recurring vocabulary; it cannot reconstruct unclear or missing speech.
Templates shape output; they should not invent or silently correct uncertain source facts.
AI cleanup may change emphasis or wording. Compare with the raw transcript where available, or the original audio.
Troubleshooting reduces uncertainty; it does not guarantee a perfect transcript or AI output.

FAQ: Should every wrong word go into my Dictionary?

No. Prefer recurring, high-value proper nouns, product names, acronyms, and domain terms. A one-off acoustic mistake or missing audio is not automatically a vocabulary problem.

FAQ: Can a Prompt Template fix a misheard name?

A template can shape output, but it should not be relied on to repair an uncertain source fact. Verify the raw transcript or audio first; use Dictionary for recurring vocabulary only when the current transcription path supports it.

FAQ: Why can the final note differ from the raw transcript?

A post-transcription template or AI-output layer can reorganize wording, headings, summaries, or extracted actions. Compare available raw text with shaped output separately before deciding that speech recognition failed.

FAQ: Should I re-record immediately after a failure?

Not automatically. First preserve and inspect the source or saved record available on your current platform. Identify whether capture, processing, transcription, or output failed. Re-record when the source is unusable or the current recovery path requires it.

FAQ: What is the fastest safe test?

Use a short representative recording with the same microphone setup, language pattern, and important terms. Change one variable, compare with the source, and only then retry a longer workflow.

FAQ: What should I verify before sharing or reusing a transcript?

Verify names, exact quotes, numbers, dates, decisions, obligations, owners, deadlines, and other consequential facts against the original source or best available primary record.

Can I use Vowise offline?

Vowise should not be treated as a fully offline transcription app based on current product evidence. On Desktop, when you stop a recording that contains captured audio, Vowise saves the recording to local Records before network transcription or later AI processing finishes. If connectivity or later processing fails, check the retained record first and retry from there instead of recording the same content again.

Desktop retains valid captured audio locally before network transcription or AI optimization finishes.
Recent Mobile release notes describe offline retry improvements, but historical release notes do not guarantee identical offline behavior across all mobile builds or workflows.

FAQ: If the connection drops while I am recording, is my audio lost?

On Desktop, a stopped recording with captured audio is retained in local Records before network transcription or AI processing finishes. That makes the recording recoverable after a later processing failure. It does not prove that every Vowise feature or every platform works fully offline.

FAQ: Will transcription finish without an internet connection?

Do not assume it will. Current Desktop documentation explicitly separates the local saved record from network transcription and later AI processing. If processing cannot finish, keep the retained record and retry the processing step when connectivity is available.

FAQ: What should I do when the connection comes back?

1. Open Records and locate the recording created when you stopped capture. 2. Replay or inspect the retained audio before making a second recording. 3. Retry the smallest failed step rather than repeating the whole workflow. 4. If the app still reports a network problem while other online actions work, note the platform, app version/build, timestamp, and failing action for support. 5. Do not send API keys, webhook URLs, passwords, or unnecessary private transcript text in a support report.

FAQ: What about offline use on iPhone or Android?

Recent Mobile release notes mention more robust offline retries, but that is not enough evidence for a blanket promise that all current iOS and Android recording, transcription, sync, and AI workflows work offline. Check the current build and workflow before relying on a specific offline behavior.

FAQ: Does a locally retained recording mean all processing stays on my device?

No. A recording being retained locally before later processing is not the same as a local-only processing or privacy guarantee. Treat storage, transcription, AI processing, sync, and privacy as separate questions. Use the current Privacy Policy and product documentation for applicable data-handling facts.

Failure map

Most issues become easier once you separate where the failure happened.

Capture: microphone, permissions, file format.
Processing: upload, transcription, long-audio behavior.
Output: template quality, dictionary terms, review state.
Automation: auth header, endpoint, secret handling.

Recover a stopped Desktop recording

A stopped Desktop capture with actual audio is retained in local Records before later processing decisions. If speech detection, the network, transcription, or AI processing fails, check Records before making another recording.

Open Records and locate the item created when you stopped recording.
Replay or inspect the retained audio before retrying transcription.
If no local item exists, record the time and capture state for support without exposing private audio.

Confirm whether AI optimization ran

Desktop can start one AI optimization while final speech is settling, then reuse that prefetched result when the recording finishes. A fast final response therefore does not mean AI was skipped, and it should not create a second workflow run.

First check whether the saved result is marked optimized, raw fallback, pending, or failed; these states are different.
After a successful Records retry, desktop diagnostics should preserve the actual n8n workflow name, workflow ID, and call ID returned by the server.
For support, include the timestamp and the redacted workflow/call identifiers from diagnostic logs. Never include transcript text, API keys, or webhook URLs.
A prefetch cache hit should reuse the first workflow call and skip a duplicate dispatch.