From “夯” to “拉” — review of several popular AI voiceovers in Jianying (剪映)
Learn step-by-step how to identify, classify and document AI voice presets in short-video editors (example: Jianying/剪映): locate tier labels, capture grid screenshots including preview and favorite icons, save playback samples, and annotate unreadable/cropped items for reliable comparison and verification.
Image Analysis
Image 1 — Collected visual evidence
Source: not provided (incomplete source)
Observable facts (directly readable or visually verifiable)
SkillPI Creation Analysis
Purpose of the asset:
- The image documents a UI screen showing AI voice presets or voice-character tiles—useful for comparing or recording available AI voice types and their on-screen metadata in a short-video/editor app.
Key evidence from the image (direct observations):
- A left-side tiering column with visible labels such as “顶级”, “人上人”, “NPC”, and a grid of voice tiles labeled with names like “北极熊口音” (Polar Bear accent), “娱乐扒妹” (Entertainment gossip girl), “ASMR女声” (ASMR female voice), “台湾腔甜妹” (Taiwan-accent sweet girl), “刘语熙”, and “曼波讲故事”.
- UI elements include waveform-pattern backgrounds on tiles, heart icons on some items, placeholder tiles, and a larger preview/avatar on the right.
- The screenshot is cropped in places and the source is marked as missing (“Source: not provided”), so the capture is incomplete.
Related skills / reproducible steps for verification and evidence collection (actionable steps that can be followed):
1. Open the target video-editing or short-video application and navigate to the AI voice / voice-pack section (the exact app must be confirmed).
2. Identify tiering or category labels (left column) and the grid/list of voice presets.
3. Systematically capture screenshots that include context: the left tier column, the voice grid, the selected preview on the right, and any heart/favorite indicators.
4. Play each voice sample and save an audio recording or note the playback behavior for offline comparison.
5. Organize screenshots and audio samples by tier/category and mark any items that are cropped or illegible in the original capture.
6. Record app version, account type, and navigation path used to reach the page—these metadata are required for reliable replication.
Prerequisites and cautions for reproduction:
- Access to the specific application or web page; some voice packs or UI elements may be region- or account-restricted.
- Preservation of app version and device/environment details, since UI and available presets may change between versions.
- The provided screenshot has cropped and blurred labels, so not all voice names are directly verifiable from this image alone.
Items requiring verification (not directly established by the image):
- The exact application or service shown (the supplied material includes a reference URL: https://www.bilibili.com/video/BV1bZBHBqEGW/ but the image itself notes “Source: not provided”).
- The definitive meaning of left-column tiers (“顶级”, “人上人”, “NPC”)—whether these are official tiers, editorial labels, or community tags.
- The precise function of the heart icon (favorite vs. like vs. other status).
- Accurate spellings/transcriptions of partially cropped labels (e.g., the small boy avatar label).
Conclusion (facts vs. verification guidance):
- The screenshot reliably documents visible tier labels, several voice preset names, and UI elements like waveform backgrounds and heart icons. However, because the source is incomplete and parts of the image are cropped or unreadable, a full verification requires accessing the original video or application page and following the outlined steps to collect complete screenshots and audio samples.