LOADING SURVEY…

Voice Quality
Survey

Welcome!
We're glad you are here. Please enter your name below to get started.

About This Test

🎙 Voices in This Survey
Task 1 — Part 1 of 1

—

—
—
⚠ Language-specific notes

    Generate the audio using the TTS engine below, then confirm your result.

    Generate Audio
    Speed iControls the speed of the generated voice. Values below 1.00 slow down, above 1.00 speed up the voice. 1.00
    SlowerFaster
    Stability iMakes the voice more consistent between generations. On longer texts it can sound monotone — lowering this value can help. 50%
    More variableMore stable
    Similarity iHigher values bring the voice closer to the original clone. Lowering it gives it more variation. 75%
    LowHigh
    Style Exaggeration iDefault value is 0. Increasing it results in more dynamic and emphasized generations. Over 50% may lead to instability. 0%
    NoneExaggerated
    Speaker Boost iBoosts the similarity of the generated voice to the original voice clone.
    ⚙ Model v3 — Select Stability Mode
    Final Questions — 1 of 4

    Voice Usability

    Are the voice clones usable as a native voice in the respective language?

    Final Questions — 2 of 4

    Linguistic Difficulties

    Are there any linguistic difficulties, such as incorrect stress or unnatural pronunciation?

    Final Questions — 3 of 4

    Script Optimisation

    Could a better result be achieved if the script were written differently — for example, using capitalised syllables to force stress?

    Final Questions — 4 of 4

    Final Conclusion

    Share any final thoughts — about the voices, the scripts, or this survey tool itself. This field is optional.

    Thank
    You!

    ✓
    Complete

    Your evaluation is complete. Your results have been uploaded to TRO automatically — you're all done!

    Uploading results…
    Submitted by: —
    Project: —
    Tasks completed: —
    Timestamp: —
    STS Task 1

    —

    —
    Input
    Parameters
    Stability iMakes the voice more consistent between generations. 60%
    More variableMore stable
    Similarity iHigher values bring the voice closer to the original clone. 60%
    LowHigh
    Style Exaggeration iIncreasing results in more dynamic and emphasized generations. Over 50% may lead to instability. 0%
    NoneExaggerated
    Speaker Boost iBoosts similarity to the original voice clone. Recommended for most use cases.
    Remove Background Noise iRemoves background noise from the input recording before conversion. Enable if the recording contains ambient noise.
    STS — Voice Refinement

    —

    —
    Input
    Stability 60%
    More variableMore stable
    Similarity 60%
    LowHigh
    Style Exaggeration 0%
    NoneExaggerated
    Speaker Boost iBoosts similarity to the original voice clone. Recommended for most use cases.
    Remove Background Noise iRemoves background noise from the input recording before conversion. Enable if the recording contains ambient noise.
    You can regenerate up to 3 times
    Select best take