The output ran 7.48s versus the 9.08s source, about 18% shorter, with non-uniform compression that required content-based matching rather than timestamps.
What was measured
Automation Level
How much manual effort is required across the pipeline: transcription → translation → dubbing → lip sync → export.
context, not decisivecapability
How much manual setup is required is important operationally, but it is not the main measure of whether the finished video is good. (3 of 3 judges)
What was given, what came back
Test input: Educational airport conversation short (English → Spanish) · video · group: video-translation-voice-clone-lip-sync
Input — what we sent
Input not captured
This run recorded no prompt or input file for the test, so we cannot show you what produced the result below. Capture gaps are tracked, not hidden.
A structured educational/conversation-style English short video used to test translation into Spanish with clear narration, steadier pacing, and easier lip-sync alignment than the other scenarios.
Why this input is hard
- · Structured speech translation accuracy
- · English-to-Spanish narration quality
- · Longer-phrase lip-sync consistency
- · Clear voice rendering for informational content
- · Export/download reliability in a simple speaking scenario
Output — unretouched
Also checked on this input — same tool, 3 other criteria
Lip Sync Accuracy✓ WorkedA content-matched frame comparison confirmed genuine mouth regeneration, with the output mouth shape differing from the source.Output Quality & Export✓ WorkedThe export preserved the source's 360×640 resolution and showed no corruption or color glitches.Translation Accuracy✗ FailedThe banana-ripeness labels stayed in English ('Unripe,' 'Ripe,' 'Overripe,' 'Rotten') instead of being localized to Spanish.
Provenance
- Observation
- 4901b889-9bb1-4d82-9334-e83564b52a4e
- Evidence run
- ec1dd100-6af8-4f49-8f19-b78492702b14
- Study
- Translate Videos with Voice Cloning and Lip Sync Using AI
- Research task
- 86b96dfpf
- Tested at
- not recorded
- Source
- first-party
- Evidence state
- verified
- Proof shown
- output only
- Cost / latency
- not captured
- Repeat run
- not captured
- Tester
- not captured
The last three rows are honest blanks, not placeholders — our capture has no field for them yet.
Query this
get_evidence({
tool: "akool",
scenario: "video-translation-voice-clone-lip-sync"
})MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Same input, same check — 1 other tool
measured on Automation Level
This evidence is published in
From the same study (page rebuilt from a later run)
Real inputs and real outputs, no retouching · every cell queryable via API & MCP · aidemos.com