Dungeons N Waifu’s (New Version 2.2)
Dungeons N Waifu's (New Version 2.2)
- TypeCheckpoint
- ModelSD 1.5
- TagCHECKPOINTarmor artistic bard character character art concept concept art dark dark fantasy design digital illustration digital painting dragonborn drow druid dwarf elf environment concept art fantasy fantasy art fighter game character girl goblin high fantasy high quality illustration kobold monk oil painting orc photorealism PHOTOREALISTIC portraits realistic roleplaying games rpg sci fi semi-realistic sorcerer style tiefling ultra realistic warlock wizard
๐ฅ๐ NOW UPDATED TO V2.2! w/ BUILT IN NOISE OFFSET! ๐๐ฅโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
A handpicked and curated merge of the best of the best in fantasy world, character & creature design. Should serve as a bit of a swiss army knife / multi-tool to create any and all fantasy art, character concepts, portraits, designs, locations, & even a bit of n00ds. I’ve yet to find what this can’t handleโ๐จ
Specializes in painterly, semi realistic to realistic illustration, and artists based prompts, but can do design work, vector illustrations, logos, and a bit of anime as well.
Works best from 7-13.5 CFG and up 20-70 STEPS imho. YMMV!
SHOW ME WHAT YOU MAKE WITH THIS!! ๐
Note on the Versions: Keep in mind this merged model has been evolving.. So the 1.6 version is not very similar to the 2.2 version or visa-versa. Yes, they both produce good results but the result you get will be different. (I tried to illustrate this in the example images if you flip between versions.)
If you find yourself asking “which is the right version?” If you’re new to SD and want an easy ride choose version 1.0 or 1.6 both are consistent and work quite well. I still use 1.6 for some things from time to time. If you want darker more detailed environments and detailed shading choose 2.0 or 2.2 but realize that 2.0 & 2.2 are a little more demanding from the prompt maker in order to get the exact result you want.
There is no embedded VAE in this (other than ones included in the merges) but one of these should work fine: (click the links)
โ vae-ft-mse-840000-ema-pruned.ckpt
โ kl-f8-anime2.ckpt
*** Note: that the example images are provided as a basis of reference and raw output. In most cases untouched other than mild inpainting for the eyes or face in a few examples. *** THESE SHOULD BE EASILY REPRODUCIBLE RESULTS! IF YOU HAVE QUESTIONS PLEASE ASK ๐ช
Merge listing consists of:
===========
(Version 1.6)
cheeseDaddys_35
dreamshaper_33
unstableinkdream_v6
aToZovyaRPGArtistsTools15
dreamlikeDiffusion10
deliberate_v2
===========
(Version 2.0)
cheeseDaddysLandscapes_35 (Offset Noise)
dreamshaper_3.32 50%
UnstableInkDream_v6 50%
aToZovyaRPGArtistsTools15 50%
dreamlikeDiffusion10 50%
deliberate_v2 50%
revAnimated_reva1_FULL 20%
umiAIMythologyAndBabes_aphroditeRealisticV1 30%
dungeonsAndDiffusion_v3 20%
noiseOffsetForTrue_v10 50%
===========
(Version 2.2)
cheeseDaddysLandscapes_35 (Offset Noise)
dreamshaper_3.32 50%
UnstableInkDream_v7 50%
aToZovyaRPGArtistsTools15 50%
dreamlikeDiffusion10 30%
deliberate_v2 60%
revAnimated_v11 25%
umiAIMythologyAndBabes_aphroditeRealisticV1 25%
FaeTastic 60%
dungeonsAndDiffusion_v3 25%
===========
There are pruned & non pruned versions available if you use the pulldown
This is better for merging & testing. Otherwise use the pruned. ๐
๐ PLEASE NOTE: All of the originals have their own attributions so these must be followed to use this.
I was asked recently how to get higher quality images like the examples and about blurred eyes and faces:
================================================
Blurred eyes, or sometimes warped faces are not that uncommon in SD and sometimes require an inpaint to pull off. Most of my examples are not inpainted. But a few are.
First off when you run txt2img ALWAYS use HIRES FIX!
latent (nearest or exact) upscale / denoise 0.4-0.7 (Complete variable – depends on the type and complexity of the image. YMMV!)
Upscale by 1.5-2 but no more than that.
This will get you a pretty good base image to work with.
Afterwards.. You can always send your txt2img picture to inpaint on the bottom right under the result area > select the face or eyes > lower the denoise strength to 0.5-0.57. > make sure your seed is set to random.
masked content = fill / inpaint area = whole picture ====== for faces
masked content = fill / inpaint area = only masked ====== for eyes
With “whole picture” selected for larger inpaint target aka faces (or hands) you may need the denoise to be lower to pick up more of the original image. 0.53-0.56
Now with “only masked” and a smaller target like eyes you get a little more freedom so your denoise can be 0.54-0.57
Run a batch of 4-5 (or more) for the faces.. go through pick the best one.. if you’re lucky you got the eyes and face in one go… if not you may need to choose the best face.. send back to inpaint again.. select just the eyes.. and rerun a new batch (changing it to only masked and the settings described above for eyes.)
Finally when you get it all together send to extras:
Resize to 1.5-2
Upscale 1= Nearest
Upscale 2= R-ESERGAN 2x or 4x
But google & get yourself Remacri Upscaler and Lollypop Upscaler
4x_NMKD_Superscale-SP_178000_G
Upscaler 2 is what will get you there. If you want super sharp / realism images you choose NMKD
You want a good all around upscaler that can do nearly any subject you choose Remacri
If you’re working on toons, anime, animated, cell shaded you choose Lollypop
Each of these is unique and so is their outputs if you want to compare simply change the pull down and run off the various types and contrast / compare the finished results.
For the more advanced upscale: Get yourself Ultimate SD Upscale
(click the link for a basic tutorial on how to get it and how to use it.)
Good Luck & Have Fun!
================================================