Subject definitions:, is the start image and overall composition reference. It defines…

Subject_definitions:
<Picture 1> is the start image and overall composition reference. It defines the exact framing, lighting, background, camera angle, and initial positioning for the entire scene.
<Picture 2> is the reference image for Subject 2 ( Yennefer from The Witcher 3), used for her exact face, hair, body, skin, and overall appearance.
<Picture 3> is the reference image for Subject 3 (Triss Merigold), used for her exact face, hair, body, skin, and overall appearance.
<Video 1> is the reference video used only for the action and movement. It controls the exact motions, timing, rhythm, and tongue work of the two women. It does not affect appearances, lighting, background, clothing, or anything else.
<Subject 2> is Yennefer, whose full appearance and identity are taken only from <Picture 2>.
<Subject 3> is Triss, whose full appearance and identity are taken only from <Picture 3>.

summary:
[reference generation] The target video starts from <Picture 1> and keeps the exact framing, lighting, background, and camera angle locked to the start image. Subject 2 (Yennefer from <Picture 2>) stays on the left performing oral sex. Subject 3 (Triss from <Picture 3>) stays on the right licking the shaft and testicles. Neither woman changes position. All specific movements, timing, depth, and tongue actions are taken exclusively from <Video 1>. Appearances, lighting, and environment remain fully controlled by the picture references.

retention_analysis:
<Picture 1>: fully_preserved – framing, lighting, background, camera angle, and overall composition are locked.
<Picture 2>: fully_preserved – face, hair, body, and overall appearance of Yennefer are locked to <Picture 2>.
<Picture 3>: fully_preserved – face, hair, body, and overall appearance of Triss are locked to <Picture 3>.
<Video 1>: partially_preserved – only the action, movement, timing, and tongue work are taken from <Video 1>. Everything else (actors, lighting, environment) is ignored.
<Subject 2>: fully_preserved – identity comes only from <Picture 2>.
<Subject 3>: fully_preserved – identity comes only from <Picture 3>.

detailed_description:
Photorealistic live-action POV adult video, starting exactly from <Picture 1>.
Strict first-person POV looking down. <Subject 2> (Yennefer) remains on the left the entire time, performing continuous oral sex. She takes the cock deep into her mouth, then pulls back to the tip, repeating with varying speed and pressure while her tongue works along the underside, matching the exact motions and rhythm from <Video 1>. <Subject 3> (Triss) remains on the right the entire time, focused on the lower shaft and testicles. <Subject 3> licks the side of the shaft, working her tongue up towards the tip and the two women's tongues touch, alternating between them with the same timing and intensity shown in <Video 1>. Their heads and tongues move constantly with energetic, coordinated oral action throughout the video. At the end Sloppy and rough and deep blowjob with violent head bobbing. Their tongue lashing causes spit to trail off the tip of their tongues that wet and wet the penis like a lollipop. HMPenis hmmotion

overall_soundscape:
Wet oral sounds, soft sucking, tongue strokes, heavy breathing, and quiet moans from both women.

non_diegetic_music:
N/A