GPT Image 2.5 water-level photo prompt shared in full for realistic tourist shots
techhalla · x · 2026-09-21
techhalla shared a detailed prompt recipe for generating realistic candid tourist photos with GPT Image 2.5 (Flare), as part of exploring new business growth ideas.
- Frames the output as an "authentic candid smartphone photo": natural daylight, slight phone compression, no text/watermarks/UI.
- The critical trick is camera placement: phone held barely above the water surface, shooting horizontally — explicitly banning high angles, drone views, or downward shots so turquoise water fills the lower half and people loom large mid-frame.
- Full prompt available in the thread.
More from Multimodal
- Qwen Image 2.1 nails 4/44-panel layouts in single-prompt character sheets, fails framing review — HarnessLightnin · 2026-09-21
- PlotWeaver begins training Pan-African voice model covering 68 African languages across 44 countries — madeofAjala · 2026-09-21
- First VNCCS PoseStudio LoRA for Qwen Image 2.1 released with workflow — AHEKOT · 2026-09-21
- MiniMax H3 roundup: CrossView-Warp and four more LoRAs land — optimisticalish · 2026-09-21
- 3D generation fail: dimensionally accurate but totally the wrong object — _Stocko_ · 2026-09-21
- Supra2-IMG: 100M-Param Text-to-Image Model Trained in Under 10 Hours on One H100, Open-Sourced — LH-Tech_AI · 2026-09-21