CivArchive
    Zavy's Cinematic Stills - SDXL - v1.0
    NSFW
    Preview 8194444Preview 8194455Preview 8194471Preview 8194465Preview 8194468Preview 8194474Preview 8194470Preview 8194463Preview 8194475Preview 8194460Preview 8194467Preview 8194461Preview 8194462Preview 8194469Preview 8194466Preview 8194476Preview 8194464Preview 8194477Preview 8194509

    Please consider joining my Patreon so I can keep most of my work available for everyone - I'll be releasing early access models there for a cheaper monthly fee than buying them here individually, and also will be providing exclusive models and more!

    For business inquiries, commercial licensing, custom models, and consultations, please get in touch at [email protected] or [email protected]. You can also contact me here through CivitAI DM or join my Discord.


    Introduction

    This LoRA mimics the essence of a cinematic still. It typically produces a shallow depth of field, particularly when the subject, such as a person, is well-defined in the image. It furthermore offers lower contrast compared to SDXL finetunes (which often nearly burn their images), a distinct color range for shadows, and introduces a subtle haze. This results in a cinematic still that appears more natural and provides artists with greater flexibility for post-processing adjustments such as contrast manipulation, dehazing, texture enhancement, and saturation tuning.

    Triggers

    zavy-cnmtc

    Zavy-cnmtc is a semi-required trigger. Consider using other terms, such as: atmospheric haze, cinematic still, dynamic lighting, movie screencap, moody, film. These will help enhance the LoRA's effect, otherwise you may get a pretty bland image.

    Strength between 0.3 and 0.7.

    Description

    First upload, see description.

    LORA
    SDXL 1.0
    by Zavy

    Details

    Downloads
    5,974
    Platform
    CivitAI
    Platform Status
    Available
    Created
    3/19/2024
    Updated
    9/28/2025
    Deleted
    -
    Trigger Words:
    zavy-cnmtc

    Files

    zavy-cnmtc-sdxl.safetensors

    Mirrors

    Huggingface (1 mirrors)
    CivitAI (1 mirrors)