* 无转场镜头。无过渡。动态、高能量的杂技式战斗场面。流畅、有力且具有电影感的运动，动态清晰度极佳。角色执行快节奏的武术连击，包括回旋…

```* No cut scenes. No transitions. Dynamic, high-energy acrobatic fight sequence. Fluid, powerful, and cinematic movement with excellent motion clarity. The character performs fast-paced martial arts combos including spinning kicks, flying knees, aerial flips, wall runs, jumps, rolls, and brutal takedowns. Smooth, athletic, and realistic physics with strong momentum and weight behind every strike. Dynamic camera work — tracking shots, circling camera, dramatic angles, occasional slow-motion on impact. High intensity, graceful yet powerful choreography, excellent fluidity and speed. Cinematic lighting, sharp motion details, no stiffness, highly dynamic and exciting fight scene, masterpiece quality, 6-second video...., highly dynamic and exciting fight scene, smooth fluid motion, masterpiece quality, 6-second video. This one for the 24-30 second extension I added to the beginning of the prompts "they traverse the terrain while continuing to fight" because I simply wanted more movement. But otherwise it was literally the same prompt on repeat in the extension with a base photo I'd made previously. The same way I'd done the previous video (quoted post). Same prompt, but now it carries far more details / happenings. It also didn't blow through the part of the prompt that asks for no cut scenes. Less inconsistency of detail as well. If you notice in the original fight scene one of the attackers randomly has hair and then it's gone again. And this photo is a lot to carry clothing hair movement, as well as detailed attackers, etc. I was also able to continue on into more extensions before hitting an almost imaginative tire wall with the choreography. In the previous video I didn't continue to make the video longer because it kind of like stopped knowing what to do next. Like it had used up all its moves. But now Imagine doesn't run out of ideas on the third extension etc. *Note: base photos where the character has a ton of space/room around them works way better than if they start off in a smaller or more confined looking area. *I will probably improve upon the prompt to make sure when on the ground they don't just lay there. lol

* 无转场镜头。无过渡。动态、高能量的杂技式战斗场面。流畅、有力且具有电影感的运动，动态清晰度极佳。角色执行快节奏的武术连击，包括回旋踢、飞膝、空翻、跑墙、跳跃、翻滚以及凶狠的摔技。动作流畅、富有运动感且物理写实，每一击都带有强大的动量和重量感。动态镜头运用——跟踪镜头、环绕镜头、戏剧性角度，偶尔在命中瞬间使用慢动作。高强度、优雅而有力的编排，流畅度与速度俱佳。电影级打光，锐利的运动细节，无僵硬感，极为动态且激动人心的战斗场面，大师级品质，6 秒视频……，极为动态且激动人心的战斗场面，流畅顺滑的运动，大师级品质，6 秒视频。 这个是用于 24-30 秒扩展的，我在提示词的开头加上了"they traverse the terrain while continuing to fight"，因为我只是想要更多的运动。除此之外，扩展中的提示词几乎完全重复，使用的是我之前制作的一张基础照片。和之前那个视频的做法一样（引用的帖子）。 同样的提示词，但现在承载了更多的细节/事件。它也没有略过提示词中关于无转场镜头的那部分要求。细节的不一致性也更少了。如果你注意到的话，在原始的战斗场景中，有一个攻击者莫名其妙地有头发，然后头发又消失了。这张照片要承载服装、头发运动以及细节丰富的攻击者等等，负担很大。我也得以继续做更多的扩展，才在编排上撞到一堵近乎想象力的墙。在之前的视频中，我没有继续把它做得更长，因为它好像就停下来不知道下一步该做什么了。就像它已经把所有招式都用完了。但现在 Imagine 在第三次扩展时仍然不会没有创意，等等。 *注：角色周围有大量空间/余地的基础照片，比角色一开始就处于较小或更局促区域的照片效果好得多。 *我可能会改进提示词，以确保角色倒地后不会只是躺在那里。哈哈