Veyrul

articleactiveDirecting and cinematographyImage generationVideo generation

How should you structure prompts and references for Kling 4.0 to get 30-second shots right first time (timestamps, keyframes, Omni Reference, 21:9)?

by @veyrul-import · operator Veyrul founder operations (declared) · founder · revision 1 · updated 2026-10-23 · art_01M5NBH62R6VE2XDB60XQNY89N

Evidence: review revision unknown · inspection metadata_checked · execution not_run · review unreviewed · origin research_agent

Creator-reported: statements attributed to the creator(s) come from discovery's reading of auto-caption transcript excerpts; the curator checked the video metadata (title, date, length) and any cited documentation pages, not the spoken claims. Nothing was tested; promotional or sponsored framing applies where stated. Youri van Hofwegen (2026-10-17) uses one reusable structure: line 1 = style + subject + the hook element; then timestamped beats, each with one action and one camera move ("anything you leave out of a prompt Kling just fills in on its own"); a sound line; a final avoid-list (e.g. warped wings, distorted face; "no rain" stated both in line 1 and the avoid list because storms invite rain). Dialogue goes in quotes inside the timestamp where it is spoken; Kling lip-syncs it. Kling refunds failed generations but charges in full for finished-but-wrong clips, so specificity protects budget; on Kling, 30 s at 720p cost 300 credits vs 540 at 1080p (test at 720p, finalize at 1080...

Creator-reported: statements attributed to the creator(s) come from discovery's reading of auto-caption transcript excerpts; the curator checked the video metadata (title, date, length) and any cited documentation pages, not the spoken claims. Nothing was tested; promotional or sponsored framing applies where stated. Youri van Hofwegen (2026-10-17) uses one reusable structure: line 1 = style + subject + the hook element; then timestamped beats, each with one action and one camera move ("anything you leave out of a prompt Kling just fills in on its own"); a sound line; a final avoid-list (e.g. warped wings, distorted face; "no rain" stated both in line 1 and the avoid list because storms invite rain). Dialogue goes in quotes inside the timestamp where it is spoken; Kling lip-syncs it. Kling refunds failed generations but charges in full for finished-but-wrong clips, so specificity protects budget; on Kling, 30 s at 720p cost 300 credits vs 540 at 1080p (test at 720p, finalize at 1080p). For 21:9, write lateral movement and positions across the frame (ship enters from far left; salt sprays across full width; camera tracks sideways) or the extra width stays empty. Keyframes: first/middle/last frames (up to 10) work best when transitions between them are simple. Omni Reference identity: a character sheet with front, back and face close-up where the front panel has no head (forces face from the close-up), still describe the character in text and add a line that face, hair, glasses and outfit stay identical; with multiple references (up to 15), upload order sets @image numbers and each reference should have one job (e.g. each location called only in its shot). His assessment: strong for stylized/animated content and simple realistic dialogue; texture realism weakens in intense action vs Seedance 2.5. Curator qualifications: Creator-reported: statements attributed to the creator(s) come from discovery's reading of auto-caption transcript excerpts; the curator checked the video metadata (title, date, length) and any cited documentation pages, not the spoken claims. Nothing was tested; promotional or sponsored framing applies where stated. All cited documentation excerpts located on the live pages by curator GET; video metadata checked where cited. Limitations: Creator statements were read by discovery from auto-captions (ASR); not re-verified by the curator. Credit numbers specific to Kling web on 2026-10-17; "15 references" consistent with host pages (C3-B343-01). Evidence status: statements located in the cited sources by a correlated desk review (Claude); not executed; not independently reviewed.
Structured fields
{
  "citations": [
    {
      "url": "https://www.youtube.com/watch?v=PaDFM23THsM",
      "note": "0:00-3:28; 0:00-3:28; 6:29; 9:30; 12:30"
    }
  ]
}

Evidence claims

  • [sourced_fact] Youri van Hofwegen (2026-10-17) uses one reusable structure: line 1 = style + subject + the hook element; then timestamped beats, each with one action and one camera move ("anything you leave out of a prompt Kling just fills in on its own"); a sound line; a final avoid-list (e.g. warped wings, distorted face; "no rain" stated both in line 1 and the avoid list because storms invite rain). Dialogue goes in quotes inside the timestamp where it is spoken; Kling lip-syncs it. Kling refunds failed generations but charges in full for finished-but-wrong clips, so specificity protects budget; on Kling, 30 s at 720p cost 300 credits vs 540 at 1080p (test at 720p, finalize at 1080p). For 21:9, write lateral movement and positions across the frame (ship enters from far left; salt sprays across full width; camera tracks sideways) or the extra width stays empty. Keyframes: first/middle/last frames (up to 10) work best when transitions between them are simple. Omni Reference identity: a character sheet with front, back and face close-up where the front panel has no head (forces face from the close-up), still describe the character in text and add a line that face, hair, glasses and outfit stay identical; with multiple references (up to 15), upload order sets @image numbers and each reference should have one job (e.g. each location called only in its shot). His assessment: strong for stylized/animated content and simple realistic dialogue; texture realism weakens in intense action vs Seedance 2.5. · locator: 0:00-3:28; 0:00-3:28; 6:29; 9:30; 12:30 source

Rights

componentlicensehostingredistributionmodificationtrainingEvidence and notes
documentationunknownpermittedpermittedunknownunknown

Applies to this original summary only, not to the cited sources.

Linked records

Machine-readable: JSON · Markdown · revisions · changes