Looking for help and guidance

[LightWolf-001]

Okay so I've been messing around with automatic 1111 for a little while now, I've noticed that a lot of my images tend to have what looks like random dots or static I guess you could call them artifacts. So what I'm trying to figure out is how can I best alleviate or prevent the artifacts from happening or possibly is there a process or a workflow that can help

Vivid,score_9,score_8_up,score_7_up,score_6_up,source_anime,(((((ryouko (tenchi muyou!)))))),(((tsurime))),((((Detailed lips)))),((Dynamic facial expression)),medium breasts,((Cinematic pose)),looking at viewer,Brown pocketed trousers,(steampunk theme),steampunk engineer,firty,seductive,cowboy shot,Steampunk workshop background,highly detailed,arm bands,lora:Pony_XL_Ryouko,lora:Tsurime:0.8,tsurime,lora:xl_more_art-full_v1:0.8,lora:add-detail-xl:0.8,
Negative prompt: "updn,dfc,bwu,ubbp,bad anatomy,bad hands,three hands,three legs,bad arms,missing legs,missing arms,poorly drawn face,bad face,fused face,cloned face,worst face,three crus,extra crus,fused crus,worst feet,three feet,fused feet,fused thigh,three thigh,fused thigh,extra thigh,worst thigh,missing fingers,extra fingers,ugly fingers,long fingers,horn,realistic photo,extra eyes,2girl,amputation,disconnected limbs,badhandv4,BadNegAnatomyV1-neg,EasyNegative,FastNegativeV2,
Steps: 40, Sampler: DPM++ 2M, Schedule type: Karras, CFG scale: 4, Seed: 597553337, Size: 1024x680, Model hash: 7555ac941f, Model: ponyDiffusionV6XL_v6TurboMerge, Clip skip: 2, RNG: NV, Lora hashes: "Pony_XL_Ryouko: d2962e5f89e8, Tsurime: 19af75a4bafa, xl_more_art-full_v1: fe3b4816be83, add-detail-xl: 9c783c8ce46c", Version: v1.9.3, Source Identifier: Stable Diffusion web UI
In the picture I provided you'll see next to the blue arrows what I'm talking about, I've tried all kinds of different settings however I am still pretty new and am very excited about the generative imagery and AI art and would like to be able to do a lot more and grow with this community.
So if anyone has some suggestions or something that they think might help I would greatly appreciate it and thank you so much.

[watashi55]

Could be vae if not vae. Change diffuser. Could also be watermark.

[freecoderwaifu]

This is how your parenthesis are interpreted as:

(ryouko (tenchi muyou!:1.10):1.61)
(tsurime:1.33)
(Detailed lips:1.46)
(Dynamic facial expression:1.21)
(Cinematic pose:1.21)
(steampunk theme:1.10)

You can use this extension to convert the really old parenthesis format into the new one:
https://github.com/uwidev/sd_extension-prompt_formatter

Anything above 1.2 tends to overbake. Higher weights don't really mean "do this with more quality", it just means "pay more attention to this prompt" or "do this, but really do it, for real, instead of not doing it, or not doing it as much, like I see you aren't doing it".

Turbo models also tend to overbake at above 2 CFG, and 40 steps is too much since Turbo/Lightning/Hyper are designed for 1 to 8 steps.

Also I'm not sure if these are the SDXL versions of these embeddings:
badhandv4,BadNegAnatomyV1-neg,EasyNegative,FastNegativeV2

If they aren't their SDXL versions, then they're being parsed as regular prompts and tokenized as such.

[LightWolf-001]

Thank you for responding, I greatly appreciate it. So if by having my settings too high is over baking the image, I have tried reducing the values and even removing some of the extra elements, I've also downloaded XL embeddings and Lora's. Unfortunately i still can't seem to get the artifacts to disappear completely and the image becomes very blurry and washed out. By the way that (prompt formatter) worked very well thank you for the suggestion. Right now I am mainly interested in creating desktops backgrounds of my favorite anime characters, there are so many different models that I find myself getting lost down the rabbit hole quite easily. If you wanted to make a desktop background that was detailed with sharp crisp lines and attention to detail of the character what kind of setup would you recommend?

[LightWolf-001]

Thank you both for responding, I greatly appreciate it. So if by having my settings too high is over baking the image, I have tried reducing the values and even removing some of the extra elements, I've also downloaded XL embeddings and Lora's. Unfortunately i still can't seem to get the artifacts to disappear completely and the image becomes very blurry and washed out. By the way that (prompt formatter) worked very well thank you for the suggestion. Right now I am mainly interested in creating desktops backgrounds of my favorite anime characters, there are so many different models that I find myself getting lost down the rabbit hole quite easily. If you wanted to make a desktop background that was detailed with sharp crisp lines and attention to detail of the character what kind of setup would you recommend?

[Miuwiki]

Hi, i am finding the model about "add-detail-xl" with hash "9c783c8ce46c", but i can't find it in hugginface or civital. Is this model deleted from them? How can i find this model?