अब आप न्यूज्ड हिंदी में पढ़ सकते हैं। यहाँ क्लिक करें
Home » Tech » Best Negative Prompts in Stable Diffusion: Harnessing the Power of Negative Prompts

Best Negative Prompts in Stable Diffusion: Harnessing the Power of Negative Prompts

Negative prompts are items that you want the image generator to omit when generating images.

By Newsd
Updated on :
Best Negative Prompts in Stable Diffusion
Best Negative Prompts in Stable Diffusion

Best Negative Prompts in Stable Diffusion: You can create images with Stable Diffusion using only text prompts, but if you want them to appear spectacular, you must use negative prompts. Negative prompts are items that you want the image generator to omit when generating images. You can use them to eliminate specific elements, designs, and environments when generating the images you desire.

If you’re unsure of what negative prompts to use and when to use them, we’ve compiled lists of negative prompts that you can add to Stable Diffusion images of various genres.

General negative suggestions for improved photographs

To generate higher-quality images with Stable Diffusion or any other image generator, you can use the following guidelines, which are applicable to all image types.

  • ugly tiling
  • disfigured deformed
  • low-grade pixelated fuzzy granules
  • out-of-frame text watermark signature with grainy text
  • disproportioned poor proportions
  • Gross percentages
  • Poor anatomy body outside the frame
  • identical trimmed trimmed draught
  • extra limbs
  • extra limbs extra legs
  • extra digits
  • extra appendages
  • Long neck mutation disfigured limbs with mutations.
  • poorly drawn visage
  • poorly drawn ankles
  • poorly drawn palms
  • Absent palms Absent limbs Absent legs
  • missing digits
  • fused digits
  • awkward position out of focus
  • poor definition
  • dreary background with no content renders an artificial engine

Best Negative Prompts for Various Occasions

When producing images with a specific style or setting, it may be beneficial to include negative prompts that are more pertinent to the type of artwork you’re about to generate. You can use these keywords to generate images with other AI image tools if you know how to include them in the generator you select.

Some of these keywords may appear more than once, as they are applicable to multiple categories of images.

For scenery

When constructing images of a landscape, natural beauty, or scenic view with Stable Diffusion, the following terms can be used as negative prompts.

  • simple background is tedious
  • plain standard
  • homogenous uncreative
  • unattractive
  • opaque
  • grayscale (opportunity)
  • black and white (optional)
  • distorted details details are low grains
  • cloudy and hazy (optional).
  • gloomy (optional)
  • blurry portrait oversaturated
  • weak contrast
  • underexposed
  • underexposed lowres low resolution
  • close-up macro
  • surreal multiple vistas
  • multiple positions

For urban vistas and street views

You can generate images of cities, streets, structures, and monuments using the following negative keywords.

  • permeated with poor contrast
  • underexposed overexposed lowres low-resolution low-quality
  • persons (obligatory)
  • Cars and pets are optional.
  • Optional: creatures
  • JPEG artefacts solid backdrop normal background asymmetrical buildings
  • close-up macro surreal multiple perspectives
  • multiple positions
  • Creepy frightening hazy grainy surreal sky bizarre hues
  • deformed architecture

For images featuring portraits of humans and animals

You can use Stable Diffusion to generate photographs of humans and animals that appear realistic if you include these terms as negative prompts.

  • disfigured  disproportionate  poor morphology bad proportions
  • unsightly beyond the border
  • Unfocused mangled asymmetric cross-eyed despondent adolescent plastic animal
  • enhanced depth of field
  • cloned facial features
  • shattered eyeshadow
  • asymmetrical ear lobes
  • imperfect vision
  • hair that floats (optional)
  • unnatural conjoined
  • missing limb
  • missing arm
  • poorly drawn visage
  • missing leg
  • poorly drawn ankles
  • poorly drawn palms
  • floating appendage disconnected limb additional limb
  • malformed appendages
  • malformed wrists
  • badly sculpted features
  • poor facial characteristics
  • imperfectly rendered limbs
  • double visage
  • Unbalanced physique Unnatural physique Lacking physique Juvenile (optional)
  • Long torso impedes aged obese caricature
  • 3D
  • strange colours
  • unnatural skin tone
  • unnatural epidermis
  • rigid visage and fingers
  • skewed irises
  • moustache (opportunity).
  • moustache (optional)
  • persons with bizarrely shaved heads

To create photorealistic images

To generate photorealistic images in Stable Diffusion, you must eradicate styles and elements that are incompatible with realism. Thus, you can use these keywords as negative cues to ensure that the generated images appear as natural as possible.

  • unnatural
  • implausible caricature, painting, and drawing Unreal engine
  • black and white (not required)
  • black and white (optional)
  • oversaturated
  • Low saturation, surreal, underexposed, and overexposed JPEG artefacts.
  • conjoined aberrations
  • multiple altitudes
  • glaring illumination anime drawings
  • distorted video game Photoshop UI inventive disintegrated rotten additional windows

Regarding paintings, drawings, and graphics

When creating paintings, illustrations, digital art, pop art, or drawings, you can use the following negative prompts.

  • Inadequate art photorealism
  • realistic stock photographic image
  • poor fan art
  • sloppy art text piled with ambient noise
  • poor design
  • human (opaque)
  • inaccurate plants
  • inaccurate 3D rendering of the heavens
  • Unsophisticated unprofessional artist
  • less imaginative
  • lowres grainy

These are the only negative image generation prompts permitted for Stable Diffusion.