Anima version is available!
Nova Furry XL
A 2d/2.5d furry checkpoint model that can have great details on any type of furs, scales and feathers, which aims to be XL-fied Peaki Furry
Rules
You cannot use the generated images for commercial use if it's not edited (or just turning it to black and white)
Advertising is always welcome
Recommended settings:
Sampler: Euler a for best performance
Steps: 20-50 steps
(Pony)
CFG Scale: 6-7
Prompt: score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_furry, BREAK, furry, anthro, detailed face eyes and fur, {Animal Type}
Negative Prompt: blurry, jpeg artifacts, username, watermark, signature, normal quality, worst quality, large head, low quality, text, error, missing fingers, extra digits, fewer digits, bad eye
(Illustrious)
CFG Scale: 3-5
Prompt: masterpiece, best quality, amazing quality, very aesthetic, high resolution, ultra-detailed, absurdres, newest, scenery, furry, anthro, {Prompt}, BREAK, depth of field, detailed fluffy fur, volumetric lighting
Negative Prompt: human, multiple tails, modern, recent, old, oldest, graphic, cartoon, text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, lowres, bad anatomy, bad hands, missing fingers, extra digits, fewer digits, cropped, very displeasing, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, signature, watermark, username, simple background, conjoined, bad ai-generated
Description
Changes
Reduced overcooked visual
Further enhancement of coloring
NoobAI EPS v1.1 + Illustrious v2.0-stable DARE applied
FAQ
Comments (31)
The entire Nova suite is 🔥 although If we could have a Nova Disney/Pixar checkpoint my life would be complete. 👌
Any tips on how to achieve monotone fur, without any form of pattern or different colours? Prompts like "monotone fur", "monotone body" and negative prompts like "multicolored fur:2, two tone fur:2, countershade fur:2, dots, mottled, mottled fur:2, spotted fur" can't produce satisfying results reliably and don't do anything at all in most cases, at least in V15 or V15.5.
Hello!!! I use it this way and it works in most cases.
((monochrome red fur:1.5)), ((red skin:1.5)), ((red body:1.5)), ((red skin face:1.2))
@Sother Hey! Thanks a ton, that really seemed to be the trick.
Ohh, I do like the style tag adherence on 15.5; best of the batch so far from what came after 9.
bro why is there so much unhinged bullshit in the Gallery below
Use block user (three dots -> hide content from user) & enjoy
Some people's imaginations should just stay locked up in their minds.
@deadricpon3390 and throw away the key
Welcome to the internet
v15.5: After 5 months (v11) I decided to test this model again and it seems like the problem that I had with it are fixed now. Huge comeback. LORA compatibility is amazing and even just out of the box its amazing. Thank you!
So far the best looking Anime checkpoint I used, and I cycled through a lot. For some reason it is even better for non-furry content than NovaAnime or at least on par.
Anatomical accuracy and the ability to add more detail is definitely a step in the right direction for newer versions, but please prioritize compatibility and flexibility more. While the results are more stable, they lack flexibility of previous versions. it seems as if models past V12 or V13 used AI imagery as image references.
I'm just here to point out that going from v11 to 15 is such a huge downgrade. This is such an awesome checkpoint when it comes to 3D CGI. Even better than Nova3DCG. The generations went from being on point to looking more like your classic plastic 3D look. I think I'll stick to v11 for now.
Something happened since v13. The color scheme has changed, and in some rare cases, the anatomy broke (yokozuwari pose, for example). The same thing happened with Nova Anime starting with the same v13. The last good version, in my opinion, was 12. Although many prefer 11.
where i can use those version of the model? the telegram page i lurk updates often and i cant do it
On TensorArt, I think. The link is in the description.
I still use 12. Is 11 better?
@RottenRaccoon Looks about the same
I don't think v11 is better. I switched to v12 when it released and didn't notice any issues. So I had no reason to go back to v11.
@PalmtopTiger Still on 12?
@RottenRaccoon yes. I don't like the faded palette and strange anatomy in the newer versions. It's hard to explain, but sometimes the model can generate homunculi with unnaturally thin waists, large heads, and so on. It's as if it has a penchant for drawing lolis.
15.5 is probably better than 12. Probably.
Edit: Never mind. Smeared colors and sepia filter.
What even is the point of the first comment if you haven't compared them?
@HidFig I was busy comparing them. Unless I'm messing up the prompt:
- The borders between colors are blurred. This is most visible between the iris and pupil of the characters, as if they're crying. First found on version 13.5
- There is a sepia tone or blue light filter present in the gens. Think Hollywood Mexico.
Regarding the comments, I'm working on creating the IL v16.0 which features the following:
• Non-blurry edges for facial features (especially eyes)
• Better color handling (no sepia filter)
• Better posing (extreme, consistent and stable)
• Great style adherence
• Controlled details
• Anime / 3DCG style handling
• Great LoRA compatibility
It'll be available around next week
Thanks for the good work! I still think I'm not trying hard enough with the prompts but 13.5 was literally unusable because of the smeared colors (deleted all gens after comparing to 12, no brainer) and 15.5 is good until you notice the ambervision.
Hope to see a pixai port soon. ☺️
@Nexious Uploaded!
https://pixai.art/en/model/1979552307491464440
@Crody Thanks a bunch 🥰🫶
This model is already very very good. Yeah, 1024 iterations on medium shots or further out you get vague looking eyes, but it puts enough there for hires.fix or upscale to add detail properly. And the rest of it is just excellent. Kudos, this is a banger of a model.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.