TERRES — all the prompts.

The brand board is generated in one go, with four already-produced visuals as reference: the model redraws nothing, it composes. Logo, palette, type and imagery come out on a single page — it's the piece that turns a series of images into a brand.

Design a single BRAND IDENTITY BOARD for the fictional brand TERRES, presented as one flat designed page seen perfectly straight on, filling the whole frame. This is a graphic design layout, NOT a photograph of a physical board — no desk, no hands, no perspective.

USE THE FOUR ATTACHED PHOTOGRAPHS AS THE ACTUAL IMAGE TILES OF THE BOARD. Place them unmodified inside clean rectangular frames: image 1 as the large hero tile, images 2, 3 and 4 as supporting tiles. Do not redraw them and do not replace them with other pictures — crop them only if the layout needs it.

LAYOUT — a strict editorial grid with generous margins on a flat neutral background from the palette:
· top left, the wordmark TERRES very large, set in a confident modern serif for the wordmark and a small letter-spaced sans for captions
· under it, the descriptor TORRÉFACTION ARTISANALE in small letter-spaced caps
· the four photographs as a clean asymmetric grid, one large and three smaller
· a horizontal PALETTE row of five solid colour chips, each with its name and hex code in small caps beneath: ESPRESSO #2B1B12, KRAFT #C9A77C, CREAM #F0E7D8, COPPER #A9663B, ASH #6E635A
· a TYPOGRAPHY block showing A B C D E F G and 0 1 2 3 4 5 in the brand typeface, with the type names labelled
· small section labels in tiny letter-spaced caps: LOGOTYPE, PALETTE, TYPOGRAPHY, IMAGERY

MOOD — artisanal, warm, single-origin. Every colour comes from the palette above and from the photographs. Spell every word correctly. Crisp vector-clean typography, high resolution, professional brand guidelines aesthetic. No watermark, no lorem ipsum, no extra logos.

Design a single BRAND IDENTITY BOARD for the fictional brand [BRAND NAME], presented as one flat designed page seen perfectly straight on, filling the whole frame. This is a graphic design layout, NOT a photograph of a physical board — no desk, no hands, no perspective.

USE THE FOUR ATTACHED PHOTOGRAPHS AS THE ACTUAL IMAGE TILES OF THE BOARD. Place them unmodified inside clean rectangular frames: image 1 as the large hero tile, images 2, 3 and 4 as supporting tiles. Do not redraw them and do not replace them with other pictures — crop them only if the layout needs it.

LAYOUT — a strict editorial grid with generous margins on a flat neutral background from the palette:
· top left, the wordmark [BRAND NAME] very large, set in [TYPEFACE FOR THE WORDMARK] for the wordmark and [TYPEFACE FOR THE CAPTIONS] for captions
· under it, the descriptor [PRODUCT DESCRIPTOR] in small letter-spaced caps
· the four photographs as a clean asymmetric grid, one large and three smaller
· a horizontal PALETTE row of five solid colour chips, each with its name and hex code in small caps beneath: [COLOUR 1 #HEX], [COLOUR 2 #HEX], [COLOUR 3 #HEX], [COLOUR 4 #HEX], [COLOUR 5 #HEX]
· a TYPOGRAPHY block showing A B C D E F G and 0 1 2 3 4 5 in the brand typeface, with the type names labelled
· small section labels in tiny letter-spaced caps: LOGOTYPE, PALETTE, TYPOGRAPHY, IMAGERY

MOOD — [THREE ADJECTIVES THAT DEFINE THE BRAND]. Every colour comes from the palette above and from the photographs. Spell every word correctly. Crisp vector-clean typography, high resolution, professional brand guidelines aesthetic. No watermark, no lorem ipsum, no extra logos.

The starting point of the whole campaign: the product, generated alone on a neutral background. It's the one you then feed in as reference in every visual — the set changes, the bag doesn't move. Here everything hangs on the text: three exact lines, a hairline rule between them. You write every single word in the prompt, otherwise the model invents a tagline and the brand stops existing from one image to the next.

Ultra high-end commercial packshot of a 250g stand-up coffee bag, upright and centred, on a seamless deep espresso-brown backdrop with a soft graduated falloff. The bag is uncoated natural kraft paper with a matte finish, a folded tin-tie top and a small brass degassing valve. Across the front, a rectangular off-white label reads TERRÈS in a confident modern serif with ORIGINE ÉTHIOPIE letter-spaced small beneath, then a fine rule and the words TORRÉFACTION ARTISANALE. Loose coffee beans are scattered at the base. Studio lighting: one large softbox from the upper left revealing the kraft fibre texture, a warm rim light on the right edge, a soft shadow to the lower right. Photorealistic, 8K, macro-level detail, premium advertising aesthetic, immaculate reflections. No other text, no watermark, no hands.

Ultra high-end commercial packshot of a [PRODUCT TYPE AND FORMAT], upright and centred, on a seamless [BACKGROUND COLOUR] backdrop with a soft graduated falloff. The [PACKAGING] is [MATERIAL 1 + ITS FINISH], [THE CLOSURE] and [THE FUNCTIONAL DETAIL]. Across the front, a [SHAPE + COLOUR] label reads [BRAND NAME] in [TYPEFACE STYLE] with [SECOND LINE] letter-spaced small beneath, then a fine rule and the words [THIRD LINE]. [WHAT IS SCATTERED AT THE BASE]. Studio lighting: one large softbox from the [DIRECTION] revealing the [MAIN MATERIAL] texture, a warm rim light on the [EDGE THAT MUST READ], a soft shadow to the lower right. Photorealistic, 8K, macro-level detail, premium advertising aesthetic, immaculate reflections. No other text, no watermark, no hands.

The shot that sets the brand's world. The drum and the cooling tray tell you where the coffee comes from — that's the entire sales argument of a specialty coffee. The bag itself stays sharp in the foreground: the set works, it doesn't take over.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — a small artisan roastery: the bag standing on a steel bench beside a drum roaster, beans cooling in the tray behind, warm tungsten light and haze. Documentary, warm, 50mm.

Capture using a Phase One XF IQ4 150MP medium-format camera with a Schneider Kreuznach 120mm LS Macro Blue Ring lens, ISO 50, f/11, focus stacking, warm HDR studio lighting with a large scrim and one hard kicker, 16-bit colour workflow, artisanal food-editorial retouching, ultra-high-resolution 4K product photography with razor-sharp kraft fibre and bean oil detail.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — a [THE WORKSHOP WHERE THE PRODUCT IS MADE]: the [PRODUCT] standing on a [SURFACE MATERIAL] bench beside [THE MACHINE THAT MAKES IT], [WHAT IS HAPPENING BEHIND] behind, [THE LIGHT AND THE ATMOSPHERE]. Documentary, warm, 50mm.

Capture using a [CAMERA BODY] with a [LENS], ISO [ISO], f/[APERTURE], focus stacking, [LIGHTING SETUP], 16-bit colour workflow, [RETOUCHING STYLE] retouching, ultra-high-resolution 4K product photography with razor-sharp [YOUR MATERIAL 1] and [YOUR MATERIAL 2] detail.

The shot that sells the quality of the roast. The film of oil on each bean reads in backlight and nowhere else: in fine grocery, it's the material that justifies the price, not the packaging.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — an extreme macro of roasted beans spilling from the open bag, the oil sheen on each bean, the kraft fibre out of focus behind. 100mm macro, backlit.

Capture using a Phase One XF IQ4 150MP medium-format camera with a Schneider Kreuznach 120mm LS Macro Blue Ring lens, ISO 50, f/11, focus stacking, warm HDR studio lighting with a large scrim and one hard kicker, 16-bit colour workflow, artisanal food-editorial retouching, ultra-high-resolution 4K product photography with razor-sharp kraft fibre and bean oil detail.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — an extreme macro of [WHAT IS INSIDE THE PACK] spilling from the open [PRODUCT], [THE SURFACE DETAIL THAT PROVES QUALITY] on each one, the [PACKAGING MATERIAL] out of focus behind. 100mm macro, backlit.

Capture using a [CAMERA BODY] with a [LENS], ISO [ISO], f/[APERTURE], focus stacking, [LIGHTING SETUP], 16-bit colour workflow, [RETOUCHING STYLE] retouching, ultra-high-resolution 4K product photography with razor-sharp [YOUR MATERIAL 1] and [YOUR MATERIAL 2] detail.

The shot that shows what it's for. A full cup, a portafilter, and the bag set beside it: you're selling the result in the cup, not the pack. Two objects are enough to make a counter.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — the bag on a zinc counter beside a filled espresso cup and a portafilter, morning light through a window. Warm, lived-in, caf documentary.

Capture using a Phase One XF IQ4 150MP medium-format camera with a Schneider Kreuznach 120mm LS Macro Blue Ring lens, ISO 50, f/11, focus stacking, warm HDR studio lighting with a large scrim and one hard kicker, 16-bit colour workflow, artisanal food-editorial retouching, ultra-high-resolution 4K product photography with razor-sharp kraft fibre and bean oil detail.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — the [PRODUCT] on a [COUNTER MATERIAL] counter beside [THE FINISHED RESULT] and [THE TOOL THAT MADE IT], morning light through a window. Warm, lived-in, documentary.

Capture using a [CAMERA BODY] with a [LENS], ISO [ISO], f/[APERTURE], focus stacking, [LIGHTING SETUP], 16-bit colour workflow, [RETOUCHING STYLE] retouching, ultra-high-resolution 4K product photography with razor-sharp [YOUR MATERIAL 1] and [YOUR MATERIAL 2] detail.

Worn hands, an apron, nothing else. This shot shows neither the whole product nor the face: it sells the fact that somebody made it by hand. It's the shot industrial brands can't produce.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — a close crop of weathered hands folding the tin-tie top of the bag, apron and steel bench behind. Soft side light, craft, tactile.

Capture using a Phase One XF IQ4 150MP medium-format camera with a Schneider Kreuznach 120mm LS Macro Blue Ring lens, ISO 50, f/11, focus stacking, warm HDR studio lighting with a large scrim and one hard kicker, 16-bit colour workflow, artisanal food-editorial retouching, ultra-high-resolution 4K product photography with razor-sharp kraft fibre and bean oil detail.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — a close crop of weathered hands [THE SINGLE MAKING GESTURE ON THE PRODUCT], [THE WORKWEAR AND THE BENCH] behind. Soft side light, craft, tactile.

Capture using a [CAMERA BODY] with a [LENS], ISO [ISO], f/[APERTURE], focus stacking, [LIGHTING SETUP], 16-bit colour workflow, [RETOUCHING STYLE] retouching, ultra-high-resolution 4K product photography with razor-sharp [YOUR MATERIAL 1] and [YOUR MATERIAL 2] detail.

The only shot that shows the product surrounded by its competitors, with one single unit pulled forward. It's the image that reassures a retailer: it proves the packaging holds up from a metre away, not just in macro.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — rows of the bags on a raw timber shelf in a specialty shop, one pulled forward and in focus, price cards handwritten. Retail documentary, warm.

Capture using a Phase One XF IQ4 150MP medium-format camera with a Schneider Kreuznach 120mm LS Macro Blue Ring lens, ISO 50, f/11, focus stacking, warm HDR studio lighting with a large scrim and one hard kicker, 16-bit colour workflow, artisanal food-editorial retouching, ultra-high-resolution 4K product photography with razor-sharp kraft fibre and bean oil detail.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — rows of the [PRODUCT] on a [SHELF MATERIAL] shelf in a [TYPE OF SHOP], one pulled forward and in focus, [THE RETAIL DETAIL THAT MAKES IT REAL]. Retail documentary, warm.

Capture using a [CAMERA BODY] with a [LENS], ISO [ISO], f/[APERTURE], focus stacking, [LIGHTING SETUP], 16-bit colour workflow, [RETOUCHING STYLE] retouching, ultra-high-resolution 4K product photography with razor-sharp [YOUR MATERIAL 1] and [YOUR MATERIAL 2] detail.

High-speed flash, dark background, dust in suspension. This shot no longer shows the bag — it shows the exact instant the product becomes useful. It's the visual that gets the most engagement and the fewest sales: you need both in a series.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — an extreme macro of ground coffee falling into a portafilter basket, a fine cloud of dust suspended, the bag out of focus behind. High-speed flash, dark.

Capture using a Phase One XF IQ4 150MP medium-format camera with a Schneider Kreuznach 120mm LS Macro Blue Ring lens, ISO 50, f/11, focus stacking, warm HDR studio lighting with a large scrim and one hard kicker, 16-bit colour workflow, artisanal food-editorial retouching, ultra-high-resolution 4K product photography with razor-sharp kraft fibre and bean oil detail.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — an extreme macro of [THE PRODUCT'S CONTENT IN MID-TRANSFORMATION] falling into [THE TOOL THAT RECEIVES IT], [THE FINE MATTER SUSPENDED IN THE AIR], the [PRODUCT] out of focus behind. High-speed flash, dark.

Capture using a [CAMERA BODY] with a [LENS], ISO [ISO], f/[APERTURE], focus stacking, [LIGHTING SETUP], 16-bit colour workflow, [RETOUCHING STYLE] retouching, ultra-high-resolution 4K product photography with razor-sharp [YOUR MATERIAL 1] and [YOUR MATERIAL 2] detail.

The calm, lived-in version: two cups, a linen napkin, a torn brioche. The product isn't staged, it's set at the edge of the frame — that's what makes the image believable on a product page.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — the bag at the edge of a breakfast table with a French press, two cups, a linen napkin and a torn brioche, soft morning window light. Appetising, natural.

Capture using a Phase One XF IQ4 150MP medium-format camera with a Schneider Kreuznach 120mm LS Macro Blue Ring lens, ISO 50, f/11, focus stacking, warm HDR studio lighting with a large scrim and one hard kicker, 16-bit colour workflow, artisanal food-editorial retouching, ultra-high-resolution 4K product photography with razor-sharp kraft fibre and bean oil detail.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — the [PRODUCT] at the edge of a [MEAL OR MOMENT] table with [OBJECT 1], [OBJECT 2], [OBJECT 3] and [SOMETHING HALF-EATEN OR HALF-USED], soft morning window light. Appetising, natural.

Capture using a [CAMERA BODY] with a [LENS], ISO [ISO], f/[APERTURE], focus stacking, [LIGHTING SETUP], 16-bit colour workflow, [RETOUCHING STYLE] retouching, ultra-high-resolution 4K product photography with razor-sharp [YOUR MATERIAL 1] and [YOUR MATERIAL 2] detail.

The closing shot: strict symmetry, hard shadow, and the beans arranged by roast level. A flat-lay must not be a pile of objects, it has to tell a ranking — otherwise it's filler.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — a strict overhead flat lay on deep espresso brown: the bag lying flat dead centre, surrounded by a radiating arrangement of beans by roast shade, a brass scoop and a folded filter. Hard shadow, graphic.

Capture using a Phase One XF IQ4 150MP medium-format camera with a Schneider Kreuznach 120mm LS Macro Blue Ring lens, ISO 50, f/11, focus stacking, warm HDR studio lighting with a large scrim and one hard kicker, 16-bit colour workflow, artisanal food-editorial retouching, ultra-high-resolution 4K product photography with razor-sharp kraft fibre and bean oil detail.

Place the product from the reference image into the scene below EXACTLY AS-IS. Its shape, its colours, its materials, its finish and every letter of its text must be 100% identical to the reference — do not recreate it, do not redesign it, do not restyle its typography, and never add any word that is not already on it. Only integrate it naturally into the scene, at true real-world scale, with lighting and reflections that match the scene.

SCENE — a strict overhead flat lay on [BACKGROUND COLOUR]: the [PRODUCT] lying flat dead centre, surrounded by [AN ARRANGEMENT THAT CLASSIFIES SOMETHING], [OBJECT 2] and [OBJECT 3]. Hard shadow, graphic.

Capture using a [CAMERA BODY] with a [LENS], ISO [ISO], f/[APERTURE], focus stacking, [LIGHTING SETUP], 16-bit colour workflow, [RETOUCHING STYLE] retouching, ultra-high-resolution 4K product photography with razor-sharp [YOUR MATERIAL 1] and [YOUR MATERIAL 2] detail.

Ten seconds, eleven shots, no people. Here the bag doesn't levitate: it stays put and it's the camera that moves. A soft object in zero gravity deforms, so you avoid it. Only the background changes at each cut, always inside the same palette — it's the most profitable format: no casting, no location, no filming permit.

SCENE CONTEXT
A 10-second vertical kinetic product spot for TERRES, a single-origin coffee. No people, no location. A single espresso and kraft colour world, the bag standing with the camera orbiting it in every shot, roughly one second per shot, the backdrop changing on every cut inside the espresso and kraft palette.

ACTIVE REFERENCES
@image1 - the TERRES bag. A 250g stand-up coffee bag in uncoated natural kraft paper with a matte finish, a folded tin-tie top, a small brass degassing valve, and an off-white label reading TERRES in a modern serif with ORIGINE ETHIOPIE beneath and TORREFACTION ARTISANALE below a fine rule. 100% matches the reference.

LOCATION MAP
Seamless studio colour fields only, no horizon line, a soft graded floor the bag stands on from the first shot. Each shot uses one backdrop from the espresso and kraft palette: smooth espresso gradient, warm kraft bokeh, crossing copper light bars, drifting coffee dust, or a deep brown field.

FIRST FRAME / BLOCKING
Frame opens in extreme macro on the kraft paper fibre grain filling frame, a warm kraft glow behind it.

FORMAT MODE
Timed multishot, eleven shots. Cuts only at the specified points, the camera does not cut on its own.

OPTICS
Physical cine lens, 180 degree shutter motion blur, macro-capable rendering.

CAMERA
0.0-0.8s - 12 degree FOV. Accelerating macro creep across the uncoated kraft fibres.
0.8s HARD CUT
0.8-1.6s - 18 degree FOV. Fast 90 degree orbit around the bag on a deep espresso gradient at 20 km/h.
1.6s SMASH CUT
1.6-2.4s - 12 degree FOV. High-speed macro, roasted beans bursting outward past the lens, oil sheen on each one.
2.4s HARD CUT
2.4-3.2s - 29 degree FOV. Push-in as a spiral of ground coffee coils around the bag on kraft bokeh.
3.2s WHIP CUT
3.2-4.0s - 18 degree FOV. Fast arc past the bag between crossing copper light bars.
4.0s WHIP CUT
4.0-4.8s - 12 degree FOV. Whip into extreme macro on the brass degassing valve, a highlight sliding across it.
4.8s SMASH CUT
4.8-5.4s - 107 degree FOV. Whip pan through drifting coffee dust, the bag flashing past.
5.4s HARD CUT
5.4-6.2s - 12 degree FOV. Locked macro, the off-white TERRES label, the serif wordmark razor sharp.
6.2s SMASH CUT
6.2-7.0s - 18 degree FOV. Fast tilt across the bag, the TERRES label filling frame.
7.0s MATCH CUT
7.0-7.8s - 29 degree FOV. The copper of the valve matches into a copper light bar, camera arcing at 30 km/h across a deep brown void.
7.8s HARD CUT
7.8-10.0s - 47 degree FOV. The bag settles into its final position on a polished copper reflective floor, camera pushing in steadily.

ACTION
Kraft fibres fill frame. The camera orbits the bag. Beans burst outward. Ground coffee spirals around it. The brass valve catches light. The label fills frame. The bag settles on polished copper.

PHYSICS
The bag is 250 grams standing on its gusseted base, the kraft creasing where it folds. Beans have real mass and bounce. Ground coffee disperses as fine dust.

LIGHTING
Hard specular key from camera-left at 4400K raking across the bag, a tight rim light on the right edge, copper practicals in the background. White balance fixed at 4400K. Atmospheric haze at 25%.

COLOR GRADE
60:30:10 - deep espresso dominant, kraft secondary, copper accent.

AUDIO
Non-diegetic kinetic design: sharp whooshes on each cut, beans tumbling, paper crinkling, a grinder burst, a low sub hit on the final landing.

STYLE
Photorealistic commercial cinematography, 24fps smooth motion, fine film grain, 8K detail, high-energy kinetic pacing.

OUTPUT SETTINGS
9:16 vertical, total duration 10 seconds.

POSITIVE LOCKS
The bag design stays identical to @image1 in every shot: natural kraft paper, folded tin-tie top, brass valve, and the off-white TERRES label with ORIGINE ETHIOPIE and TORREFACTION ARTISANALE legible. The set contains only seamless colour fields and the reflective floor. Every shot contains visible camera movement.

SCENE CONTEXT
A 10-second vertical kinetic product spot for [BRAND NAME], [WHAT THE PRODUCT IS]. No people, no location. A single [COLOUR 1] and [COLOUR 2] colour world, the [PRODUCT] standing with the camera orbiting it in every shot, roughly one second per shot, the backdrop changing on every cut inside the [PALETTE NAME] palette.

ACTIVE REFERENCES
@image1 - the [BRAND NAME] [PRODUCT]. [FULL PRODUCT DESCRIPTION — format, material, closure, functional details, and every word printed on the label in order]. 100% matches the reference.

LOCATION MAP
Seamless studio colour fields only, no horizon line, a soft graded floor the [PRODUCT] stands on from the first shot. Each shot uses one backdrop from the [PALETTE NAME] palette: [BACKDROP 1], [BACKDROP 2], [BACKDROP 3], [BACKDROP 4], or [BACKDROP 5].

FIRST FRAME / BLOCKING
Frame opens in extreme macro on [THE MATERIAL GRAIN OF THE PACKAGING] filling frame, [WHAT GLOWS BEHIND IT].

FORMAT MODE
Timed multishot, eleven shots. Cuts only at the specified points, the camera does not cut on its own.

OPTICS
Physical cine lens, 180 degree shutter motion blur, macro-capable rendering.

CAMERA
0.0-0.8s - 12 degree FOV. Accelerating macro creep across [THE MAIN SURFACE].
0.8s HARD CUT
0.8-1.6s - 18 degree FOV. Fast 90 degree orbit around the [PRODUCT] on [BACKDROP 1] at 20 km/h.
1.6s SMASH CUT
1.6-2.4s - 12 degree FOV. High-speed macro, [WHAT IS INSIDE THE PACK] bursting outward past the lens, [ITS SURFACE DETAIL] on each one.
2.4s HARD CUT
2.4-3.2s - 29 degree FOV. Push-in as [A SPIRAL OR RIBBON OF THE PRODUCT'S CONTENT] coils around the [PRODUCT] on [BACKDROP 2].
3.2s WHIP CUT
3.2-4.0s - 18 degree FOV. Fast arc past the [PRODUCT] between crossing [ACCENT COLOUR] light bars.
4.0s WHIP CUT
4.0-4.8s - 12 degree FOV. Whip into extreme macro on [THE FUNCTIONAL DETAIL], a highlight sliding across it.
4.8s SMASH CUT
4.8-5.4s - 107 degree FOV. Whip pan through [DRIFTING ATMOSPHERE], the [PRODUCT] flashing past.
5.4s HARD CUT
5.4-6.2s - 12 degree FOV. Locked macro, the [LABEL], the wordmark razor sharp.
6.2s SMASH CUT
6.2-7.0s - 18 degree FOV. Fast tilt across the [PRODUCT], the [LABEL] filling frame.
7.0s MATCH CUT
7.0-7.8s - 29 degree FOV. The [ACCENT COLOUR] of [WHERE IT SITS ON THE PRODUCT] matches into a [ACCENT COLOUR] light bar, camera arcing at 30 km/h across a [DARK VOID COLOUR] void.
7.8s HARD CUT
7.8-10.0s - 47 degree FOV. The [PRODUCT] settles into its final position on a [REFLECTIVE FLOOR MATERIAL] reflective floor, camera pushing in steadily.

ACTION
[THE ELEVEN BEATS RESTATED IN ONE SHORT SENTENCE EACH, IN THE SAME ORDER].

PHYSICS
The [PRODUCT] is [ITS REAL WEIGHT] standing on [WHAT IT STANDS ON], the [PACKAGING MATERIAL] creasing where it folds. [HOW ITS CONTENT BEHAVES WHEN IT MOVES]. [HOW THE FINEST PART OF IT DISPERSES].

LIGHTING
Hard specular key from camera-left at [COLOUR TEMPERATURE] raking across the [PRODUCT], a tight rim light on the right edge, [ACCENT COLOUR] practicals in the background. White balance fixed at [COLOUR TEMPERATURE]. Atmospheric haze at [PERCENTAGE].

COLOR GRADE
60:30:10 - [DOMINANT COLOUR] dominant, [SECONDARY COLOUR] secondary, [ACCENT COLOUR] accent.

AUDIO
Non-diegetic kinetic design: sharp whooshes on each cut, [THE PHYSICAL SOUNDS THE PRODUCT ITSELF MAKES], a low sub hit on the final landing.

STYLE
Photorealistic commercial cinematography, 24fps smooth motion, fine film grain, 8K detail, high-energy kinetic pacing.

OUTPUT SETTINGS
9:16 vertical, total duration 10 seconds.

POSITIVE LOCKS
The [PRODUCT] design stays identical to @image1 in every shot: [THE FEATURES THAT MUST NEVER CHANGE, AND EVERY LINE OF TEXT ON THE LABEL, LEGIBLE]. The set contains only seamless colour fields and the reflective floor. Every shot contains visible camera movement.

The exact opposite of hyper motion: camera on dolly and crane, never handheld, four beats of two and a half seconds. One place, one person, one progression — here the bag gets sealed, then it gets used. This is what you sell for the most, because it looks like a real ad, and the ABSOLUTELY NO PACKSHOT line stops the model falling back on a studio background.

SCENE CONTEXT
A 10-second vertical commercial spot for TERRES, a single-origin coffee. Early morning in a small neighbourhood roastery as it opens. Composed camera on a dolly and a crane, never handheld. Four beats in a structured arc, roughly two and a half seconds each.

ACTIVE REFERENCES
@image1 - the TERRES bag. A 250g stand-up coffee bag in uncoated natural kraft paper with a matte finish, a folded tin-tie top, a small brass degassing valve, and a rectangular off-white label reading TERRES in a modern serif with ORIGINE ETHIOPIE beneath and TORREFACTION ARTISANALE below a fine rule. 100% matches the reference.

LOCATION MAP
A small roastery: a drum roaster, a steel bench, a zinc counter, sacks of green beans, one tall window. The set contains only what is listed here.

FIRST FRAME / BLOCKING
Frame opens on the drum roaster turning as steam rises in the cold morning air.

FORMAT MODE
Timed multishot, four shots. Cuts only at the specified points, the camera does not cut on its own.

OPTICS
Physical cine lens, 180 degree shutter motion blur, clean spherical rendering.

CAMERA
0.0-2.5s ESTABLISHING - 63 degree FOV. Slow crane descent through the roastery at 2 km/h, steam rising, the roaster turning.
2.5s HARD CUT
2.5-5.0s PRODUCT IN USE - 47 degree FOV. Slow dolly-in as a man in his forties in a canvas apron folds the tin-tie top of the bag and sets it on the counter.
5.0s HARD CUT
5.0-7.5s LIFESTYLE - 29 degree FOV. Lateral dolly along the counter as he pulls an espresso, the bag standing beside the machine in the foreground.
7.5s HARD CUT
7.5-10.0s HERO - 63 degree FOV. Slow crane rise pulling back as he slides the cup across to a customer, the bag on the counter with its label legible.

ACTION
The roaster turns and steam rises. A roaster folds the bag shut and sets it down. He pulls an espresso beside it. He slides the cup across the counter.

PERFORMANCE
He moves with practised economy, checking things by habit, not looking at the lens. Nobody performs for the lens.

PHYSICS
The bag is 250 grams, standing on its gusseted base with a soft contact shadow, the kraft creasing where folded. Steam rises and disperses. Espresso pours with real crema.

LIGHTING
Cold morning light from one tall window at 5400K, warm tungsten from the work lamps, steam catching both. Atmospheric haze at 20%.

COLOR GRADE
60:30:10 - warm kraft dominant, espresso brown secondary, copper accent.

AUDIO
Diegetic only: the drum roaster turning, beans tumbling, a grinder, an espresso machine, quiet morning street. No dialogue, no subtitles.

STYLE
Photorealistic commercial cinematography, 24fps smooth motion, fine film grain, 8K detail, unhurried editorial pacing.

OUTPUT SETTINGS
9:16 vertical, total duration 10 seconds.

POSITIVE LOCKS
The bag design stays identical to @image1 in every shot: natural kraft paper, folded tin-tie top, brass valve, and the off-white TERRES label with ORIGINE ETHIOPIE and TORREFACTION ARTISANALE legible. The people keep the same faces, hair and clothing across every cut. Every shot contains visible camera movement. ABSOLUTELY NO PACKSHOT: the bag is never shown isolated on a studio backdrop and never appears without a person in the scene.

SCENE CONTEXT
A 10-second vertical commercial spot for [BRAND NAME], [WHAT THE PRODUCT IS]. [TIME OF DAY] in [THE PLACE] as [WHAT IS HAPPENING THERE]. Composed camera on a dolly and a crane, never handheld. Four beats in a structured arc, roughly two and a half seconds each.

ACTIVE REFERENCES
@image1 - the [BRAND NAME] [PRODUCT]. [FULL PRODUCT DESCRIPTION, INCLUDING EVERY LINE OF TEXT ON THE LABEL]. 100% matches the reference.

LOCATION MAP
[THE SET, DESCRIBED ONCE: the machines, the surfaces, the raw material in the room, the light sources]. The set contains only what is listed here.

FIRST FRAME / BLOCKING
Frame opens on [THE MACHINE OR GESTURE THAT DEFINES THE PLACE].

FORMAT MODE
Timed multishot, four shots. Cuts only at the specified points, the camera does not cut on its own.

OPTICS
Physical cine lens, 180 degree shutter motion blur, clean spherical rendering.

CAMERA
0.0-2.5s ESTABLISHING - 63 degree FOV. [THE SLOW OPENING MOVE, ITS SPEED, WHAT IT REVEALS].
2.5s HARD CUT
2.5-5.0s PRODUCT IN USE - 47 degree FOV. [THE MOVE] as [THE PERSON] [THE SINGLE GESTURE THEY MAKE WITH THE PRODUCT].
5.0s HARD CUT
5.0-7.5s LIFESTYLE - 29 degree FOV. [THE MOVE] as [THE WIDER MOMENT], the [PRODUCT] [WHERE IT SITS IN THE FRAME].
7.5s HARD CUT
7.5-10.0s HERO - 63 degree FOV. [THE FINAL MOVE THAT LEAVES THE PRODUCT IN FRAME WITH ITS LABEL LEGIBLE].

ACTION
[WHAT ACTUALLY HAPPENS, ONE SHORT SENTENCE PER BEAT].

PERFORMANCE
[HOW THE PERSON MOVES AND WHY]. Nobody performs for the lens.

PHYSICS
The [PRODUCT] is [ITS REAL WEIGHT], standing on [WHAT IT STANDS ON] with a soft contact shadow, the [PACKAGING MATERIAL] creasing where folded. [WHAT RISES, POURS OR DISPERSES IN THE SCENE].

LIGHTING
[THE NATURAL LIGHT] at [COLOUR TEMPERATURE], [THE ARTIFICIAL LIGHT], [WHAT CATCHES BOTH]. Atmospheric haze at [PERCENTAGE].

COLOR GRADE
60:30:10 - [DOMINANT COLOUR] dominant, [SECONDARY COLOUR] secondary, [ACCENT COLOUR] accent.

AUDIO
Diegetic only: [THE FOUR SOUNDS THE PLACE ITSELF MAKES]. No dialogue, no subtitles.

STYLE
Photorealistic commercial cinematography, 24fps smooth motion, fine film grain, 8K detail, unhurried editorial pacing.

OUTPUT SETTINGS
9:16 vertical, total duration 10 seconds.

POSITIVE LOCKS
The [PRODUCT] design stays identical to @image1 in every shot: [THE FEATURES AND EVERY LINE OF TEXT THAT MUST STAY LEGIBLE]. The people keep the same faces, hair and clothing across every cut. Every shot contains visible camera movement. ABSOLUTELY NO PACKSHOT: the [PRODUCT] is never shown isolated on a studio backdrop and never appears without a person in the scene.

You never generate a UGC video in one go. You build this still image first in GPT Image 2, you approve it, and only then do you animate it. Getting the still wrong costs $0.20; getting the video wrong costs seven times more.

The whole prompt is working to kill the pretty: no background blur, no colour grade, no studio light, sensor noise, botched framing, unretouched skin. It's exactly the opposite of the eight previous visuals — and that's why it's written in prose and not in JSON.

Use the attached image as the exact product reference.

THIS IS A FRAME GRABBED FROM A PHONE VIDEO, not a photograph. Vertical phone video, the phone propped on a tripod so BOTH of their hands stay free — there is never a phone in their hands.

A man in his forties with a shaved head and a greying beard, wearing a canvas apron over a tee. They are mid-sentence, caught talking — mouth open on a word, eyebrows slightly raised, eyes locked on the lens. Not smiling, not posing.

SETTING — his kitchen in the morning, a grinder and a kettle on the counter behind. It looks real and lived-in, not styled for a shoot, with ordinary everyday things around them.

FRAMING — a real front-camera selfie: the head fills most of the frame with the wide-angle distortion of a selfie lens. The crop is imperfect, the top of the head lightly cut by the frame edge, something intruding at the bottom corner. Slightly off-centre.

IMAGE QUALITY — raw front-camera video still: slightly soft, visible sensor noise, heavy phone HDR flattening the contrast, a window or lamp blown out to white. No cinematic depth of field, no colour grading, no studio light.

SKIN — real and unretouched: shine on the forehead and nose, visible pores, uneven tone, stray flyaway hairs. No beauty filter, no smoothing.

THE PRODUCT — the bag stands on the counter in front of them and they scoop beans from it with both hands. It keeps exactly the shape, the colours, the materials and every letter of the text of the reference image, at true real-world scale.

The image is the bare video frame itself: no interface, no icons, no counters, no buttons, no progress bar, no timecode, no captions, no watermark.

Use the attached image as the exact product reference.

THIS IS A FRAME GRABBED FROM A PHONE VIDEO, not a photograph. Vertical phone video, the phone propped on a tripod so BOTH of their hands stay free — there is never a phone in their hands.

[THE PERSON: age, hair, plain unbranded clothing]. They are mid-sentence, caught talking — mouth open on a word, eyebrows slightly raised, eyes locked on the lens. Not smiling, not posing.

SETTING — [A REAL ROOM WHERE THIS PRODUCT ACTUALLY LIVES], [THE TWO ORDINARY OBJECTS VISIBLE BEHIND THEM]. It looks real and lived-in, not styled for a shoot, with ordinary everyday things around them.

FRAMING — a real front-camera selfie: the head fills most of the frame with the wide-angle distortion of a selfie lens. The crop is imperfect, the top of the head lightly cut by the frame edge, something intruding at the bottom corner. Slightly off-centre.

IMAGE QUALITY — raw front-camera video still: slightly soft, visible sensor noise, heavy phone HDR flattening the contrast, a window or lamp blown out to white. No cinematic depth of field, no colour grading, no studio light.

SKIN — real and unretouched: shine on the forehead and nose, visible pores, uneven tone, stray flyaway hairs. No beauty filter, no smoothing.

THE PRODUCT — [WHERE IT STANDS AND THE ONE TWO-HANDED GESTURE THEY MAKE WITH IT]. It keeps exactly the shape, the colours, the materials and every letter of the text of the reference image, at true real-world scale.

The image is the bare video frame itself: no interface, no icons, no counters, no buttons, no progress bar, no timecode, no captions, no watermark.

The approved still becomes the first frame of the video. The phone stays on a tripod, both hands are free, one single take with no cut — and the person says exactly two lines: one at the start, one at the end. The two middle beats are marked SILENT. It's that silence, written down in black and white, that stops the model inventing dialogue.

Continue this phone video from the exact frame shown. The phone stays propped on a tripod the whole time, so BOTH of their hands stay free — there is never a phone in their hands. ONE SINGLE CONTINUOUS TAKE, no cuts, no scene changes.

They speak ONLY these two lines, each said ONCE and never repeated:
Line 1, at the very start: "I was buying supermarket coffee for fifteen years."
Line 2, at the very end: "Now I can't go back. That's the honest truth."
Between the two lines they say NOTHING at all — silent, only acting.

Beat 1 — already mid-sentence, delivering line 1 straight into the lens with a small eyebrow raise.
Beat 2 — SILENT: they lift the kraft bag with both hands and turn it so the TERRES label faces the lens and stays readable, then set it back down. Eye contact kept, lips pressed together, no words.
Beat 3 — SILENT: they scoop a handful of beans from the open bag and let them fall back in.
Beat 4 — leans back, shrugs one shoulder and delivers line 2 with a small closed-mouth smile.

Keep the raw look of the frame: the head stays large with selfie wide-angle distortion, the crop stays imperfect, the window stays blown out white, the image stays slightly soft with phone HDR flatness and sensor noise. Natural body sway throughout, never still. Real unretouched skin.

The product stays exactly as in the frame in every beat: the natural kraft paper, the tin-tie top, the brass valve and the off-white TERRES label, sharp and unchanged, never distorted, never resized and never losing its text.

Close-mic phone audio with the voice up front, faint room tone, no music. Raw iPhone footage, TikTok storytime energy, no-filter realism. Vertical 9:16. NO SUBTITLES, NO TEXT OR EMOJIS ON SCREEN.

Continue this phone video from the exact frame shown. The phone stays propped on a tripod the whole time, so BOTH of their hands stay free — there is never a phone in their hands. ONE SINGLE CONTINUOUS TAKE, no cuts, no scene changes.

They speak ONLY these two lines, each said ONCE and never repeated:
Line 1, at the very start: "[WHAT THEY USED TO DO INSTEAD — the objection the viewer already has]"
Line 2, at the very end: "[WHAT CHANGED, SAID PLAINLY — never a slogan]"
Between the two lines they say NOTHING at all — silent, only acting.

Beat 1 — already mid-sentence, delivering line 1 straight into the lens with a small eyebrow raise.
Beat 2 — SILENT: they lift the [PRODUCT] with both hands and turn it so the [LABEL] faces the lens and stays readable, then set it back down. Eye contact kept, lips pressed together, no words.
Beat 3 — SILENT: [THE ONE REAL GESTURE OF USE, DONE ONCE].
Beat 4 — leans back, shrugs one shoulder and delivers line 2 with a small closed-mouth smile.

Keep the raw look of the frame: the head stays large with selfie wide-angle distortion, the crop stays imperfect, the window stays blown out white, the image stays slightly soft with phone HDR flatness and sensor noise. Natural body sway throughout, never still. Real unretouched skin.

The product stays exactly as in the frame in every beat: [THE FOUR FEATURES AND THE LABEL THAT MUST NEVER CHANGE], sharp and unchanged, never distorted, never resized and never losing its text.

Close-mic phone audio with the voice up front, faint room tone, no music. Raw iPhone footage, TikTok storytime energy, no-filter realism. Vertical 9:16. NO SUBTITLES, NO TEXT OR EMOJIS ON SCREEN.