Horrify Anything
5.0
1 reviewsDescription
Introducing Horrify Anything:
Here's my new workflow, perhaps it serves no purpose, or maybe it does! I've created it just for fun, and I'm sure you'll have fun using it too. If you're tired of perfect and cute images and want to step up to the supreme level of terror, then you'll love it, I bet you'll spend hours transforming your photos or images.
But a warning: this workflow is not suitable for easily impressionable or faint-hearted people XD XD
How it works:
It's a simple image-to-image process, it doesn't use any special checkpoints, so it works with any checkpoint. The more realistic the checkpoint, the more terrifying the generated images will be.
It doesn't rely on LoRA or IPAdapter, nor on ControlNet.
I haven't included any LLM for the prompt because there's no need to input prompts, unless you want to generate an image with text-to-image embedded.
All images generated with "text to image" and "image to image" will be terrifying in output, at least that's the goal.
Additional information:
The generation speed is particularly fast, as it's purely image-to-image, without any additional elements that slow down the process. Of course, I've implemented a good prompt, which shouldn't be tampered with unless you know what you're doing, otherwise you won't get the horrifying images anymore.
Where it works:
As the name suggests, Horrify Anything, so you can use it with a wide range of images. Naturally, with illustrations, the result may not be effective, but it still does something.
You can use close-ups of people, half-body shots, full-body shots, animals, objects, environments and landscapes, practically everything.
Even already quite scary images will be further worsened, because when it comes to the worst, there's no end. You can also use photos of abandoned houses, and always for the same principle, the result will be very unsettling.
Below are some examples of what I've managed to do, without doing anything XD
there is no end to the worst
abandoned house (one of my generation)
scary werewolf...
Now he is a werewolf demon, I assure you he is worse XD
a beautiful unicorn
but also not
DISCLAIMER
All input images are my property, they are personal generations made over time, so being the owner, I have right of use.
I release them free of charge on this page, on OpenArt.AI, as an accompanying workflow, so they can be freely downloaded and reused.
In case they are reused and posted on the internet, on any social, forum pages, personal pages, etc. , I only ask that you refer to this page as the source.
Discussion
(No comments yet)
Loading...
Reviews
No reviews yet
Versions (1)
- latest (2 years ago)
Node Details
Primitive Nodes (6)
Image Comparer (rgthree) (1)
Latent Scale to side (1)
Note (2)
Text (1)
Text box (1)
Custom Nodes (43)
ComfyUI
- LoadImage (1)
- CheckpointLoaderSimple (2)
- CLIPTextEncode (4)
- EmptyLatentImage (1)
- VAEDecode (2)
- VAEEncode (1)
- KSampler (2)
- CLIPSetLastLayer (1)
- Compare (1)
- If ANY execute A else B (1)
- mape Variable (26)
- Seed Everywhere (1)
Model Details
Checkpoints (2)
SDXL\juggernautXL_v9Rundiffusionphoto2.safetensors
revAnimated_v2RebirthVAE.safetensors
LoRAs (0)