Billionmail/BillionMail · error

lip sync failed: %s

Error message

lip sync failed: %s

What it means

The lip-sync provider explicitly reported the job as 'failed'. pollLipSync wraps the provider-supplied error message (resp.Error) in this error and surfaces it to RunPipeline. The real cause is inside the %s payload.

Source

Thrown at core/internal/service/video_gen/orchestrator.go:419

	deadline := time.Now().Add(lipSyncTimeout)
	for {
		if time.Now().After(deadline) {
			return "", fmt.Errorf("lip sync timed out after %v", lipSyncTimeout)
		}

		resp, err := CheckLipSyncStatus(ctx, cfg, jobID)
		if err != nil {
			return "", fmt.Errorf("check lip sync status: %w", err)
		}

		switch resp.Status {
		case "completed":
			if resp.VideoURL == "" {
				return "", fmt.Errorf("lip sync completed but no video URL")
			}
			return resp.VideoURL, nil
		case "failed":
			return "", fmt.Errorf("lip sync failed: %s", resp.Error)
		}

		time.Sleep(lipSyncPollDelay)
	}
}

// withRetry retries a function up to maxRetries times with exponential backoff.
func withRetry[T any](fn func() (T, error)) (T, error) {
	var result T
	var err error
	delay := retryBaseDelay

	for attempt := 0; attempt <= maxRetries; attempt++ {
		result, err = fn()
		if err == nil {
			return result, nil
		}
		if attempt < maxRetries {

View on GitHub (pinned to fc36c76c05)

Solutions

1. Inspect the %s message for the provider's specific failure reason
2. Re-encode/validate the input media (duration, codec, resolution) before submitting
3. Re-submit the job — some failures are transient provider errors
4. Test with a known-good small input to isolate whether the input or the provider is at fault

Example fix

// before
jobID, err := SubmitLipSync(ctx, cfg, videoURL, voiceID)
// after
if err := validateLipSyncInput(videoURL, voiceID); err != nil {
	return fmt.Errorf("invalid lip sync input: %w", err)
}
jobID, err := SubmitLipSync(ctx, cfg, videoURL, voiceID)

if videoDurationSec < minSupportedSec || videoDurationSec > maxSupportedSec {
	return fmt.Errorf("video duration %ds outside supported range", videoDurationSec)
}
if voiceID == "" {
	return fmt.Errorf("voice ID required before lip sync")
}

func isRetryableLipSyncFailure(providerErrMsg string) bool {
	msg := strings.ToLower(providerErrMsg)
	return strings.Contains(msg, "capacity") || strings.Contains(msg, "internal") || strings.Contains(msg, "timeout")
}

url, err := pollLipSync(ctx, cfg, jobID)
var lipErr *LipSyncFailedError
if errors.As(err, &lipErr) && isRetryableLipSyncFailure(lipErr.ProviderMessage) {
	return resubmitLipSync(ctx, cfg)
}

When it happens

Trigger: CheckLipSyncStatus returns Status=="failed"; the provider rejected or crashed the job (bad audio/video input, unsupported format, content policy rejection, provider-side error).

Common situations: Input video/audio too short or too long for the model; corrupted screenshot/footage file; voices whose sample URL expired; provider capacity errors reported as job failure.

Related errors

• voice clone: %w
• lip sync timed out after %v
• lip sync completed but no video URL

AI-assisted analysis of Billionmail/BillionMail@fc36c76c05 (2026-09-05). Data as JSON: /api/errors/86ec6116e735a5b3. Report an issue: GitHub.