ComfyPrompt Logo
ComfyPrompt
홈ComfyUI
WebUI
미드저니
가이드

Workflow Summary for Psychedelic Butterfly, Enraged Bear, and Swirling Colors Generation ComfyUI 예시

This workflow facilitates the generation of psychedelic and colorful images of various subjects such as a butterfly, an enraged bear, and abstract forms. It begins with an Empty Latent Ratio Select SDXL to define the canvas size and shape, then uses a Checkpoint Loader to load a selected model with accompanying CLIP and VAE configurations. Text prompts are processed using CLIPTextEncode nodes, which are then passed through reroute nodes for positive and negative conditioning. KSampler Config defines the sampling process, while ControlNetApply applies specific stylistic controls to the conditioned inputs. The KSampler conducts the sampling process with given CFG and seed parameters, and the resulting latents are upscaled using LatentUpscaleBy. VAEDecode nodes convert latents into images, which are then processed with Canny edge detection and fed into another ControlNetApply for further refinement. The workflow handles both positive and negative prompts with specialized reroutes. Eventually, the final images are saved with the Save Images No Display node, incorporating a filename based on the checkpoint name, and are previewed with PreviewImage nodes. Notably, this workflow accommodates batch processing and allows for detailed customization at each stage, including steps, CFG scale, and seeds.

midjourney sref archive, best midjourney sref, midjourney sref example, 미드저니 프롬프트 예시, 미드저니 예시, 프롬프트 예시, 프롬프트 검색, 미드저니 프롬프트 레퍼런스

프롬프트 >
인기 ComfyUI Workflow 예시
Workflow Summary for Psychedelic Butterfly, Enraged Bear, and Swirling Colors Generation
정보
TestedTipsViews20
This workflow facilitates the generation of psychedelic and colorful images of various subjects such as a butterfly, an enraged bear, and abstract forms. It begins with an Empty Latent Ratio Select SDXL to define the canvas size and shape, then uses a Checkpoint Loader to load a selected model with accompanying CLIP and VAE configurations. Text prompts are processed using CLIPTextEncode nodes, which are then passed through reroute nodes for positive and negative conditioning. KSampler Config defines the sampling process, while ControlNetApply applies specific stylistic controls to the conditioned inputs. The KSampler conducts the sampling process with given CFG and seed parameters, and the resulting latents are upscaled using LatentUpscaleBy. VAEDecode nodes convert latents into images, which are then processed with Canny edge detection and fed into another ControlNetApply for further refinement. The workflow handles both positive and negative prompts with specialized reroutes. Eventually, the final images are saved with the Save Images No Display node, incorporating a filename based on the checkpoint name, and are previewed with PreviewImage nodes. Notably, this workflow accommodates batch processing and allows for detailed customization at each stage, including steps, CFG scale, and seeds.
프롬프트 결과 예시 이미지 (0)
이 프롬프트로 생성된 이미지의 다운로드 가능한 고품질 결과물의 예시 입니다.
프롬프트 정보