[MMH3] Exposure Slider

Am besten für Anfänger. Leistungsstarke Workflows, geringe Lernkurve.

PirateDiffusion Telegram

Pro studio (SpyGlass) Web

Erstelle Videos, Bilder, individuelle Chat-Personas und mehr.

Durchsuche Modelle, rendere und verfeinere von jedem Browser oder Handy aus.

Discord bot (Vector)

Füge Vector zu deinem Server hinzu oder tritt dem Vector Discord bei. Erschaffe alles!

Unified Editor Web

Editor für Bilder und Videos: rendern, bearbeiten, hochskalieren und mehr auf einer Seite.

Civitai Web

Bezahlen pro Nutzung

Direkt auf Civitai erzeugen; jedes Bild wird mit Buzz-Credits bezahlt.

Lokale Apps

Einrichtung erforderlich

Richte lokale Programme ein, wenn dein Computer leistungsstark genug ist.

ModelDuel Web

Bezahlen pro Nutzung

Erzeuge denselben Prompt gleichzeitig auf bis zu 50 Checkpoint-/Basismodellen.

GPU in der Cloud mieten

Graydient Cloud GPU

1,195 Downloads · 99 Likes auf Civitai

Pauschalpreis, unbegrenzte Nutzung. Dieses Modell mit einem Klick installieren.

RunPod Cloud GPU

Abrechnung pro Minute

GPU pro Minute mieten und Modell selbst einrichten.

Erstellt von

alcaitiff

auf Civitai

By - alcaitiff an exposure slider for H3 - see the civitai page for full details on usage

Notizen des Erstellers

Changes the exposure from dark to bright, with a range from -6 to +6. As always, results can vary depending on the video and prompt, so keep that in mind when using the slider. This was trained at rank 1 using only the last two layers, specifically to keep the slider focused on exposure and avoid changing other aspects of the video. That also makes the LoRA ridiculously tiny: only 213 KB. One thing to keep in mind is that the model itself is very capable and multimodal. If changing the lighting makes something else in the scene logically different, it may sometimes adapt other details accordingly. For example, if an outdoor scene becomes extremely dark, the model might interpret that as nighttime or assume there is some kind of roof or obstruction causing the darkness. Likewise, increasing the exposure can reveal or emphasize details that make sense for the new lighting. If you explicitly describe what you want in the prompt, the model should follow it. So if you want a specific lighting situation without changing the scene, describe it clearly in the prompt. In short: the slider controls exposure, but the model may use its understanding of the scene to make lighting-related adjustments that it considers reasonable.

SHA256

530334BB668FC222A179D55097632EF398288E497728518B096FB69E38C8D9F7

Ranking badge

Updated Oct 22, 2026

Made this model? Show this badge on your page.

Paste it into your Hugging Face card or Civitai description.

Change colors / personalize / get code

Also ranked: #19 Hottest MiniMax H3 model · #19 Hottest MiniMax H3 LoRA

Not this creator?

Get your own badge

Beliebtheitstrend von [MMH3] Exposure Slider

Heat-Score seit 18 Sep 2026

Heat-Score

8,420°

Diese Woche

▲

6.2%

8,407° damals, 8,420° jetzt

Diesen Monat

n/a

Noch nicht veröffentlicht/erfasst

Vergangenes Jahr

n/a

Noch nicht veröffentlicht/erfasst

7,000°

Grundwert

Die Skala reicht von 0° bis 9.999°. Alles unterhalb von 7,000° ist für Modelle reserviert, über die noch keine Quelle berichtet hat. Diese tragen dort eine Platzhalterzahl, damit die Listen vollständig bleiben. 7,000° ist der Punkt, an dem gemessene Daten beginnen: Ein Modell mit echten Daten, aber fast keinen Downloads, sitzt auf dem Grundwert, und die Messungen füllen die Grade darüber. Der Grundwert sagt nichts über das Modell aus; nur die Grade darüber tun das und trennen ein Modell vom anderen.

+942°

Civitai-Downloads

1,195 Downloads dieser genauen Version auf Civitai. Downloads sind die wichtigste Zutat und werden auf einer Log-Skala gezählt: Zehnmal so viele Downloads erhöhen den Wert um 300°, sodass ein Modell mit einer Million Downloads nicht tausendmal heißer ist als eines mit tausend. Die sechs Download-Quellen teilen sich diesen Wert anteilig entsprechend ihrer Anzahl.

+300°

Likes

99 Likes und Daumen hoch über die Quellen hinweg. Ein Like zählt halb so viel wie ein Download, auf derselben logarithmischen Skala.

+78°

Verbreitungsrate

Etwa 10 Downloads pro Tag, gemessen zwischen zwei Abfragen der Quellen, nie geschätzt aus einem Durchschnitt über die gesamte Lebenszeit. Die Geschwindigkeit zählt ein Viertel des Download-Werts, sodass ein frischer Anstieg sichtbar wird, ohne dass ein eintägiger Spitzenwert jahrelange Nutzung übertrumpft.

+100°

Familien-Bonus

+100°, weil dies eine LoRA einer Familie ist, die diese Seite gerade besonders fördert. Das ist eine begrenzte Unterstützung innerhalb des Scores und wird hier offen gezeigt; sie kann ein Modell nie verbergen oder einen tausendfachen Unterschied bei den Downloads ausgleichen.

Detaillierte Aufschlüsselung anzeigen

Jede Zahl auf dieser Seite kann überprüft werden. Der Heat-Score dieses Modells beträgt 8,420°, und hier wird er Schritt für Schritt aufgebaut, mit dem laufenden Gesamtwert nach jedem Schritt: der Grundwert, die Grad, die jede Quelle hinzufügt, der Bonus für die Art des Modells und was das Alter seiner Gewichte abzieht. Die Farben entsprechen den Bereichen der obigen Grafik. Das Glossar am Ende erklärt jeden Begriff.

1
Starte beim Grundwert.
Warum 7,000° und nicht null: Die Skala reicht von 0° bis 9.999°, und der Bereich unter 7,000° ist für Modelle reserviert, über die noch keine Quelle berichtet hat. Diese tragen dort eine Platzhalterzahl, damit die Listen vollständig bleiben und niemals über einem Modell mit Daten stehen. 7,000° ist der Punkt, an dem gemessene Daten beginnen. Ein Modell mit echten Daten, aber fast keinen Downloads, sitzt hier; alles, was folgt, ist Messung, und der Bereich über dem Grundwert enthält zehn zehnfache Schritte bei Downloads zu je 300°. Der Grundwert ist für jedes Modell gleich, er entscheidet also nichts; die Grade darüber tun das.
7,000°
laufende Summe
2
Füge die Downloads hinzu.
1,195 Download-Äquivalente insgesamt, von Civitai-Downloads. Zehnmal die Downloads ergeben denselben 300°, also zählt die Größe, aber sie dominiert die Bewertung nie.
+942°
7,942°
laufende Summe
3
Füge die Likes hinzu.
99 Likes und Daumen hoch, mit der halben Gewichtung eines Downloads.
+300°
8,242°
laufende Summe
4
Füge die Verbreitungsrate hinzu.
Momentan etwa 10 Downloads pro Tag, gemessen zwischen zwei Messungen, mit einem Viertel der Gewichtung.
+78°
8,320°
laufende Summe
5
Familien-Bonus.
Ein offener, begrenzter Schub für LoRAs einer Familie, die diese Seite gerade fördert.
+100°
8,420°
laufende Summe
=
Der Heat-Score.
Wo dieses Modell auf jeder Liste dieser Seite steht. Nichts anderes beeinflusst ihn.
8,420°

Glossar: die hier verwendeten Begriffe

Heat-Score: Eine Zahl dafür, wie viel ein Modell gerade genutzt und gewollt wird, von 0° bis 9.999°. Jede Liste auf dieser Seite, die „heißeste“ sagt, ist danach sortiert und nach nichts anderem.
Grundwert: Die 7.000°, mit denen jedes Modell mit echten Daten startet. Die Skala reicht von 0° bis 9.999°; der Bereich unter dem Grundwert ist für Platzhalter reserviert (Modelle ohne Daten), damit ein Modell mit Daten immer über einem ohne steht. Der Grundwert ist für alle gleich und entscheidet nichts; die Grade darüber sind die Messung.
Download-Äquivalente: Alles, was der Score als Download zählt, zusammengefasst: Downloads selbst, jeder Torrent-Seeder als 1.000, jedes AI Horde-Bild und jeder Lauf auf einer Generierungsseite als eins.
Logarithmische Skala: Gezählt in Zehnerschritten, nicht einzeln: Der Sprung von 1.000 auf 10.000 Downloads bringt genauso viele Grade wie von 10.000 auf 100.000. So verdrängen Riesen nicht alles andere.
Likes: Daumen hoch und Likes von allen Quellen, jeweils halb so viel wert wie ein Download auf der Log-Skala.
Verbreitungsrate: Wie viele Downloads ein Modell gerade pro Tag bekommt, aus dem Unterschied zwischen zwei Messungen einer Quelle. Ein Viertel der Gewichtung des Download-Begriffs.
Gemessen und geschätzt: Gemessen bedeutet, dass es zwei Messwerte einer Quelle gibt und der Unterschied echt ist. Geschätzt heißt, es ist nur die heutige Gesamtsumme bekannt und die Vergangenheit wird über die Lebensdauer des Modells verteilt; die Seite sagt, welche Variante angezeigt wird.
Seeder: Jemand, der eine Kopie der Gewichte online hält, damit andere sie über einen Torrent herunterladen können, oder ein gehosteter Spiegel, der dasselbe macht. Jeder davon zählt als 1.000 Downloads.
AI Horde: Ein Freiwilligennetzwerk, das kostenlos Bilder auf gespendeten Computern generiert. Die Anzahl der Bilder, die es mit einem Modell erstellt hat, ist öffentlich und jedes zählt als ein Download.
Generierungsseiten: Seiten, die eigene Kopien eines Modells hosten und Leute damit generieren lassen, wie PixAI, SeaArt und Yodayo. Ihre Downloads zählen wie bei Civitai und jeder Generierungslauf als ein Download.
OpenRouter Tokens: Ein Textmodell, das über eine API bereitgestellt wird, wie Claude oder GPT, hat keine Gewichte zum Herunterladen, daher wird seine Nutzung in Tokens gemessen: die Textstücke, die es über OpenRouter gelesen und geschrieben hat, einem Gateway, das viele Apps nutzen, in den letzten 30 Tagen. Je 10,000,000 Tokens zählen als ein Download. Quelle: OpenRouter (openrouter.ai/rankings), CC BY 4.0.
Vollständiges Modell (Checkpoint): Die Gewichte selbst, die Datei, die eigenständig generiert. Sie bekommt einen Bonus, weil Add-ons davon abhängen.
LoRA und Embedding: Kleine Zusatzdateien, die die Gewichte von jemand anderem in Richtung eines Stils, Charakters oder Konzepts lenken. Sie können nicht alleine laufen.
Familie und Gewichte: Die Basis, auf der ein Modell aufgebaut ist, wie SD 1.5, SDXL, Flux oder Wan. Jedes Modell erbt das Veröffentlichungsdatum der Gewichte seiner Familie.
Alter: Abzüge je nachdem, wie lange die Gewichte der Familie schon draußen sind: 150° pro Jahr, bis zu 750°. Alte Downloads, die Jahre zum Ansammeln hatten, bleiben oben in der Liste eingefroren.
Familien-Boost: Ein kleiner, begrenzter Schub für LoRAs einer Familie, die diese Seite gerade vorantreibt, immer in der Aufschlüsselung aufgeführt.
Stand-in-Wert: Eine stabile Zahl unter der Mindestgrenze, die einem Modell gegeben wird, über das noch keine Quelle berichtet hat, damit die Listen vollständig bleiben. Sie rangiert nie über einem Modell mit Daten.
Neu! (Schonfrist): Ein Modell, das weniger als 7 Tage alt ist, zeigt Neu! statt einer Punktzahl und nimmt keinen Platz in den Rankings ein. Seine Punktzahl wird ab dem ersten Tag berechnet, aber ein Modell, das einen Tag alt ist, wächst schneller als alles andere, daher wäre ohne die Wartezeit das #1 Modell der Woche immer das, das gestern herauskam, statt das, das sich eine Woche gehalten hat.
Trend: Wie das Wachstum eines Modells über einen Zeitraum im Vergleich zum typischen Wachstum eines Modells im gleichen Zeitraum ist. Schneller als üblich bedeutet Anstieg, langsamer bedeutet Rückgang.
Typisches Modell: Die Mitte des Katalogs: Die Hälfte der Modelle mit einer Zahl wuchs schneller im Zeitraum, die andere Hälfte langsamer.

Wie sich die Beliebtheit dieses Modells im Vergleich zum Rest des Katalogs entwickelt hat: sein Wachstum bei Downloads, Likes und Nutzung in jedem Zeitraum, gemessen am typischen Wachstum eines Modells im gleichen Zeitraum, mit dem Heat-Score zu diesem Zeitpunkt. Über dem typischen Wachstum bedeutet ein Anstieg, darunter ein Rückgang. Die Vergangenheit wird aus den heutigen Gesamtzahlen, dem ersten Tag des Modells und seiner gemessenen Download-Geschwindigkeit rekonstruiert, wenn zwei Messungen vorliegen; ein Modell, das zu Beginn eines Zeitraums nicht existierte, zeigt n/a an.

Beispielbilder

Prompts und Einstellungen, geteilt von den Leuten, die diese Bilder auf Civitai erstellt haben. Wähle ein Bild, um zu sehen, wie es entstanden ist.

Diesen Prompt holen

Render by alcaitiff on

Einstellungen

Größe

832x1216

Prompt

Absolutely — I’d keep the same **slider-demo structure and progressive transformation**, but replace the age/VHS axis with a dramatic **underexposed jungle → blazing tropical sunlight** axis. The panther itself should remain the same animal throughout; what changes is the lighting, visibility, shadow detail, color, and environmental illumination. integrated_multimodal_description: 0.00-0.60s: A completely static frontal shot of a magnificent adult black panther standing centered in a dense tropical jungle. The panther faces directly toward the camera, calm and alert, with glossy black fur, powerful shoulders, bright eyes, and a strong recognizable silhouette. No clothes, accessories, collars, or artificial elements. The jungle is extremely dark and dramatically underexposed. Dense tropical leaves, vines, tree trunks, and vegetation are barely visible in deep shadows. Only a tiny amount of soft ambient light outlines the panther's face, ears, shoulders, and fur. A clean horizontal slider labeled "Exposure Slider" is positioned in front of the panther, with the black metallic knob at the extreme left at "Low Exposure". The image has a cinematic modern digital appearance, but the exposure is extremely low. Most of the frame is nearly black, with only subtle details visible in the deepest shadows. 0.60-3.70s: The camera remains completely locked and the composition remains identical. The panther slowly reaches forward with one paw and drags the slider continuously from "Low Exposure" toward "High Exposure". The transformation must affect BOTH the lighting on the panther AND the entire jungle environment simultaneously. The panther remains the exact same animal throughout the shot. Its physical identity, proportions, fur, face, and position remain consistent while progressively more light reveals its details. The exposure continuously increases from extremely dark to extremely bright. 0.60-1.30s — VERY LOW EXPOSURE: The panther remains mostly concealed by darkness. Only faint highlights reveal the contours of its face, ears, nose, shoulders, and glossy black fur. The jungle is almost completely hidden in shadow. Leaves and vines emerge only as vague silhouettes. As the slider moves, subtle additional light begins revealing individual strands of fur and small details around the panther's eyes and muzzle. The image becomes progressively brighter without changing the composition. Shadow detail slowly emerges. Deep blacks begin lifting toward dark charcoal tones. 1.30-2.10s — LOW TO MEDIUM EXPOSURE: The panther becomes clearly visible. Soft natural jungle light begins illuminating its face, chest, shoulders, and glossy fur. Individual fur strands become increasingly visible, especially along the edges of the ears, cheeks, neck, and shoulders. The panther's eyes become more clearly visible and reflective. The surrounding jungle progressively emerges from darkness: large tropical leaves, hanging vines, tree trunks, branches, moss, and layers of dense vegetation become recognizable. The background develops rich natural greens and subtle depth. The lighting remains soft and cinematic, as though sunlight is beginning to penetrate the jungle canopy. 2.10-2.90s — BRIGHT EXPOSURE: The panther is now fully visible and beautifully illuminated by strong tropical daylight filtering through the jungle canopy. Bright natural sunlight creates distinct highlights across the glossy black fur. Individual hairs become extremely detailed and sharply resolved. The panther's facial structure, eyes, whiskers, nose, ears, and muscular body are clearly visible. Sunbeams penetrate through the jungle canopy behind and around the animal. Leaves become vivid and clearly separated. Dense vegetation fills the background with rich green foliage, vines, branches, moss, and tropical plants. Small patches of sunlight appear on the jungle floor. The overall image becomes significantly brighter while preserving realistic contrast and natural black fur. 2.90-3.70s — EXTREME HIGH EXPOSURE: The panther is now dramatically illuminated by intense tropical sunlight. Strong golden-white sunlight pours through gaps in the jungle canopy, creating powerful shafts of light surrounding the animal. The black fur remains visibly black but develops brilliant specular highlights along the face, ears, shoulders, back, and individual fur strands. Every detail that was previously hidden in darkness is now clearly revealed. The jungle is transformed from an almost invisible dark environment into a spectacular, intensely illuminated tropical rainforest filled with bright green leaves, vines, branches, mist, and glowing patches of sunlight. Some highlights begin approaching the limits of exposure. Bright leaves and sunbeams become intensely luminous. Subtle lens bloom and natural atmospheric glow appear around the strongest light sources. The visual progression must feel like a continuous exposure slider rather than a change of scene. There must be NO cuts and NO sudden character replacement. The audience must clearly perceive the exact same black panther continuously transitioning from almost complete darkness into intense tropical illumination. 3.70-4.20s — MAXIMUM EXPOSURE: The slider reaches the extreme right at "High Exposure". The black panther is now dramatically illuminated by powerful tropical sunlight. Its glossy black fur is covered with bright natural highlights, with individual hairs clearly visible. Its eyes reflect the intense surrounding light. The jungle behind it is extremely bright, lush, vivid, and sun-drenched. Multiple strong shafts of sunlight pass through the canopy. Bright green foliage surrounds the panther. Some leaves and sunlit areas bloom slightly from the intense exposure. The deepest shadows have been substantially lifted, while the brightest parts of the jungle approach intense white highlights. The entire scene communicates the maximum possible exposure while remaining recognizable as the same jungle and the same panther. 4.20-5.00s — FINAL PUNCHLINE: Immediately after reaching "High Exposure", a gigantic jungle-themed wooden sign suddenly swings down from above the canopy and settles behind and above the panther. The enormous rustic sign is made from thick weathered jungle wood, wrapped with vines and surrounded by tropical leaves. It displays the exact words: "Exposure Slider" The lettering is large, clean, and highly visible, naturally integrated into the wooden sign. Strong sunlight strikes the sign, producing bright highlights and subtle natural lens flare. A few leaves shake loose from the impact and fall around the panther. The panther remains completely calm and majestic in the foreground, staring directly toward the camera as if completely unimpressed by the ridiculous giant sign. The sign settles with a heavy wooden THUD followed by rustling leaves and jungle ambience. The final image remains locked in the same composition: the dramatically illuminated black panther standing in an intensely sunlit tropical jungle, with the enormous rustic "Exposure Slider" sign behind it. The contrast between the first and final frame must be extreme: almost invisible black panther + nearly black jungle + severely underexposed image becomes fully visible glossy black panther + brilliant tropical rainforest + intensely bright high-exposure image. overall_soundscape: The opening audio is extremely subdued, with quiet nighttime-style jungle ambience: distant insects, faint leaves moving, subtle animal calls, and the panther's quiet breathing. As the exposure increases, the jungle ambience gradually becomes richer and more present. Birds, insects, rustling leaves, distant monkeys, tropical wildlife, and flowing water progressively become more audible. Mechanical slider sounds synchronize precisely with the knob as it moves from "Low Exposure" toward "High Exposure". As maximum exposure is reached, the jungle becomes vibrant and alive with bright daytime ambience. The giant wooden sign produces a deep wooden impact followed by falling leaves, branches shaking, and surrounding jungle sounds. non_diegetic_music: A very subtle mysterious atmospheric melody begins in the extremely dark opening. As the slider moves, the music progressively becomes brighter, warmer, more energetic, and more expansive. Deep atmospheric tones gradually evolve into a playful cinematic jungle-inspired rhythm. At maximum exposure, the music reaches a bright, triumphant, slightly humorous peak. A short playful musical accent hits precisely when the gigantic wooden "Exposure Slider" sign falls into place. first_frame: Extremely dark, dramatically underexposed cinematic digital video. A magnificent adult black panther stands centered in a dense tropical jungle, facing directly toward the camera. The animal is mostly concealed in darkness. Only faint highlights reveal its face, ears, shoulders, glossy black fur, and eyes. The surrounding jungle is almost completely black, with only vague silhouettes of leaves, vines, branches, and tree trunks visible. A clean horizontal slider labeled "Exposure Slider" is positioned in front of the panther, with the knob at the extreme left at "Low Exposure". No clothes, no accessories, no artificial elements on the panther. last_frame: The exact same black panther in the exact same composition, now intensely illuminated by powerful tropical sunlight. Its glossy black fur is extremely detailed, with bright highlights revealing individual hairs across the face, ears, shoulders, chest, and body. The panther's eyes are clearly visible and reflective. The surrounding tropical jungle is brilliant, lush, vivid, and sun-drenched, filled with bright green leaves, vines, branches, mist, and strong shafts of sunlight penetrating through the canopy. The image is extremely bright and highly exposed, with intense highlights, lifted shadows, natural bloom, and subtle lens flare around the strongest sunlight. The slider knob is fully at "High Exposure". A gigantic rustic wooden jungle sign covered with vines and tropical leaves hangs behind and above the panther, displaying the exact words "Exposure Slider". The final frame must look radically different in LIGHTING and VISIBILITY while preserving the same animal, composition, environment, and camera position: extremely dark underexposed jungle becomes intensely bright tropical jungle. The panther itself must remain unmistakably the same animal throughout the entire transformation.

Render by PotatoChipOwO on

Einstellungen

Sampler

res_multistep

Schritte

Seed

798274899036910

Größe

640x864

Prompt

subject_definitions: <Subject 1> is the chibi mascot girl defined by <Picture 1> (full-body turnaround sheet — inherit the character design only, never reproduce the side-by-side two-view layout or its white background) and <Picture 2> (front upper-body close-up): big blue eyes, long blue hair with cyan-tipped gradient, fin-like ornaments on the sides of her head, white frilled maid headdress, orange jack-o'-lantern pumpkin with a carved smiling face balanced on top of her head, white blouse with a dark-blue bow, dark-blue corset and dress with white hem ruffles and gold embroidery, white apron with a blue whale emblem, a blue whale tail behind her, dark-blue mary-jane shoes. She is tiny: standing she is only half as tall as the computer monitor. <Subject 2> is the workspace in <Picture 3> — wooden desk, black flat-panel monitor, black full-size keyboard, white wired mouse — layout retained, but its white studio background and daylight are discarded: the scene is a dark late-night room lit by overhead fluorescent light. <Subject 3> is the round coin in <Picture 4>: distressed bright-blue metallic surface with worn gold-brass patches, raised gold letters reading "TOKEN", reeded edge. <Audio 1> is the voice-timbre reference for <Subject 1> (S1). summary: [reference generation + audio reference] In a late-night fluorescent-lit room, the tiny <Subject 1> hides behind the monitor, peeks out with a giggle, trips on the keyboard while running to it, then cheerfully asks the viewer "Trick or Token!"; the viewer's right hand offers a <Subject 3> coin, which <Subject 1> sits down and bites a corner off, chewing happily. One unbroken take in first-person point of view with smooth camera motion. retention_analysis: <Subject 1> (appears throughout): fully_preserved — face, hair, pumpkin, costume, whale tail, proportions from <Picture 1> and <Picture 2> are retained; the turnaround layout and white background are not reproduced; her scale is half the monitor's height. <Subject 2> (appears throughout): partially_preserved — the desk, monitor, keyboard and mouse layout from <Picture 3> is retained; the background and lighting become a dark room under cool fluorescent light. <Subject 3> (appears from the coin offer onward): fully_preserved — distressed blue metal, reeded edge, and raised gold letters reading "TOKEN" are retained. <Audio 1>: reference — its vocal timbre guides the giggle and dialogue delivery of <Subject 1> (S1) without copying the original signal. detailed_description: Realistic rendered style, single unbroken take with no cuts and no shot changes, the camera moving in one continuous smooth motion the whole time. The entire video is the first-person point of view of a woman with bare arms (S0, never seen): her face never enters the frame, only her bare arms and hands, entering and leaving from the bottom edge, may appear. Scale is locked: <Subject 1> stands only half as tall as the monitor, so the woman's hand is roughly the size of her whole body. Exactly two characters exist in the scene: <Subject 1> and the woman. [Shot 1] The view looks straight ahead at the monitor of <Subject 2>, its screen glowing with an open AI chat window showing plain chat bubbles with soft, unreadable text, the keyboard and mouse below it; <Subject 1> is hidden behind the monitor. She walks out from the left side of the monitor, leaning her upper body into view; the woman's bare hands move down and out of the bottom of the frame. <Subject 1> smiles and giggles: <d>[English] Hehe.</d> She runs around the monitor toward the keyboard, trips on the keyboard and tumbles forward, her face bumping into the keys; she stands up, bends down and gives her skirt one pat, then straightens up with a happy expression, looking up toward the viewer, raises both hands and cheerfully moves her lips as she says: <d>[English] Trick or Token!</d> The woman's bare right arm rises from the bottom of the frame holding the <Subject 3> coin — the coin looks large in <Subject 1>'s world — and offers it to her; <Subject 1> takes it with both hands and the right hand releases and drops out of the bottom of the frame. <Subject 1> sits down on the keyboard and brings the coin to her mouth with both hands, bites a corner off it, and chews happily with the missing corner visible. She keeps her eyes closed and keeps chewing the food in her mouth, and the video ends. overall_soundscape: Mechanical keyboard typing at the start over a quiet late-night room tone with a faint fluorescent hum. A small giggle from <Subject 1>, tiny quick footsteps running around the desk, a stumble and a clatter as she trips and her face bumps the keyboard keys, fabric rustles as the skirt is patted, a small metallic clink as the coin changes hands, then a crisp bite-crunch followed by audible chewing to the end. non_diegetic_music: A cute, warm music-box and glockenspiel melody at a gentle bouncy tempo with soft plucked synth accents; it stops right at the end while the chewing continues in the closing silence.

Render by alcaitiff on

Einstellungen

Größe

2208x1024

Prompt

integrated_multimodal_description: 0.00-2.00s: A 25-year-old blonde woman wearing a flowing red dress holds a cellphone close to her face while recording herself on a sunny beach. The shot is a handheld selfie perspective, framing her face and upper body as she smiles naturally into the phone camera. Ocean waves and bright blue sky are visible behind her. She lowers the cellphone while continuing to smile, then places it carefully upright on the sand with the camera facing toward the beach and ocean. 2.00-3.50s: From the cellphone's low position on the sand, the woman steps backward into the frame, turns toward the ocean, and begins running across the beach. Her red dress moves dynamically in the sea breeze as her blonde hair flows behind her. The cellphone remains stationary on the sand, creating a stable low-angle view of her movement toward the water. 3.50-5.50s: She runs faster toward the shoreline, reaches the shallow water, and joyfully jumps forward into the incoming wave. Water splashes around her legs and the red dress moves with the impact. The camera remains completely stationary on the sand, capturing her energetic jump against the bright ocean horizon. 5.50-6.00s: She lands in the shallow water with a playful splash, briefly turning her body toward the cellphone as another small wave rolls around her. overall_soundscape: Close handheld recording sounds at the beginning, followed by soft beach ambience, ocean waves, wind moving through the woman's hair and dress, footsteps on dry sand, then increasingly prominent splashing water as she enters the ocean. non_diegetic_music: Upbeat, carefree instrumental music with light percussion and bright acoustic elements, beginning softly during the selfie recording and gradually becoming more energetic as she runs toward the water.

Render by PrisonerResearchInstitute on

Einstellungen

Größe

1376x768

Prompt

Prompt and workflow embbed, Two of my black friends approve this

Render by alcaitiff on

Einstellungen

Größe

2208x1024

Prompt

A realistic candid handheld video in a city square. A woman sits on a bench eating a small snack while a pigeon stands directly in front of her staring intensely at the food. She slowly moves the snack away, and the pigeon follows it with its eyes. She moves it left, then right, and the pigeon tracks every movement with exaggerated seriousness. She eventually gives the pigeon a tiny piece and it immediately walks away as if nothing happened. Natural daylight, realistic bird movement, subtle handheld camera shake, deadpan comedy, ordinary everyday environment.

Render by alcaitiff on

Einstellungen

Größe

2208x1024

Prompt

A realistic candid video inside a bright apartment. A tiny puppy walks across a polished floor and suddenly notices its reflection. It cautiously approaches, tilts its head, takes one step backward, then playfully jumps toward the reflection. The puppy becomes increasingly confused and excited, wagging its tail rapidly. A person filming quietly laughs from behind the camera. Natural indoor lighting, realistic puppy behavior, handheld camera movement, imperfect focus, warm everyday atmosphere, genuinely cute spontaneous moment.

Render by alcaitiff on

Einstellungen

Größe

2208x1024

Prompt

integrated_multimodal_description: 0.00-1.20s: A young woman rides a skateboard rapidly along a smooth urban promenade, leaning naturally into the movement with confident balance. Her hair and clothing move with the wind. The camera tracks alongside her at low height, keeping her centered while the pavement flows beneath her. A large colorful street mural behind her clearly displays the exact words "Exposure Slider". 1.20-2.50s: She performs a smooth skateboard trick, briefly lifting the board from the pavement and landing cleanly while maintaining her momentum. The camera makes a subtle lateral tracking movement to follow the trick. 2.50-3.50s: She accelerates forward and looks toward the camera with a playful smile as the skateboard rolls rapidly away. The words "Exposure Slider" remain clearly visible in the background. overall_soundscape: Rolling skateboard wheels, subtle pavement vibration, light wind, and distant urban ambience. non_diegetic_music: Energetic upbeat electronic music with crisp percussion and a short rhythmic accent synchronized with the skateboard trick.

Render by alcaitiff on

Wie man eine LoRA installiert: Ordner, Dateiprüfung und erste Nutzung in ComfyUI, Forge, Automatic1111 und Draw Things

Einstellungen

Größe

2496x1216

Prompt

A realistic candid handheld video in a neighborhood park during late afternoon. A young woman holds a dog's leash and starts walking toward home. The dog suddenly sits down and refuses to move. She gently tries to encourage it, but the dog dramatically lies flat on the grass as if completely exhausted. She says: "Good girl". She looks at the dog, looks toward the camera, and starts laughing. The dog remains completely committed to not moving. Natural golden-hour light, realistic grass and trees, authentic animal behavior, casual clothing, spontaneous comedy, documentary-style realism.

Lokal ausführen

Wie man installiert

24. Sep 2025

Civitai öffnen Herunterladen

MMH3-ExposureSlider-V1.safetensors

0 MB · SafeTensor · bf16

Downloads werden von Civitai bereitgestellt. Manche Ersteller verlangen zuerst eine Anmeldung bei Civitai.

Mehr von diesem Ersteller

[KREA 2] Detail Slider

KREA2

[KREA 2] Weight Slider

KREA2

[KREA 2] Warm Light Slider

Alle Modelle von alcaitiff

Ähnliche Modelle

Minimax H3 Photorealistic Image Generator (LORA + Workflow)

H3 Fight Flow V1 | Continuous Fight Interaction LoRA

Dispatch Style Lora H3

MMH3 Locked Stabilization (Camera locked to target)