9 images; was trying to combine barbie+harleyquin together on this one. was just testing out the idea, i think theres potential there. anyways,
hope ya enjoy the late night post ! :)
2023-07-13 05:14:36 +0000 UTC View Post
10 images; i absolutely love the black & gold aesthetic, same scifi lora as my last post was used but weight at 50% instead of 100. anyways;
hope you all enjoy ! :)
2023-07-12 17:27:18 +0000 UTC View Post
8 images; tried to focus on fine face detail for this one, lighting, shadows etc. Tested out a new sci-fi theme lora aswell. anyways;
hope you all enjoy !
2023-07-12 08:13:37 +0000 UTC View Post
29 images; had fun making this one, outfit and theme came out surprisingly well!
hope you all enjoy ! :)
2023-07-11 11:48:10 +0000 UTC View Post
29 images; tried out a different workflow with this batch. I usually focus on 1 image at a time, upscale it etc, this time around I grouped each step together. Started with 200 raw images, only 50 turned out good, then only 29 survived the upscaling process. Tryin to make my computer more productive when im asleep as the upscaling can take anywhere from 3-6 mins per photo. anyways;
hope you enjoy ! :)
15 images; my goal going into this one was seeing how much skin detail i could get out of the ai model. went for a much more natural aged look aswell. Honestly tho gonna probably stick with the look from previous posts going forward, just wanted to test it out. anyways;
hope you all enjoy ! :)
10 images; tech wise the photo-realism ai model i use got updated so should be some improvements going forward.
hope you all enjoy ! :)
2023-07-09 21:00:03 +0000 UTC View Post
13 images; just spent time understanding cause and effect of technical tweaks on this batch. this character is always my go to since every ai model is filled with data on her.
also, I updated my megasync archive. hopefully it makes browsing through all the images easier/faster. anyways;
hope you enjoy ! :)
2023-07-09 01:45:56 +0000 UTC View Post
11 images; Think I figured out a better balance of LORA's for the SciFi Angel look.
This is literally my favorite aesthetic.
Hope you all enjoy ! :)
15 images; generation went well.
going to do some quality of life work on the patreon page, you'll notice the changes via naming scheme & tags.
hope you all enjoy ! :)
14 images; just organizing what my workflow will be for this character in the future. Seems like its easier to use an embedding rather than a lycoris. ai struggles with getting the chin right so it took longer to generate correctly, for the upscaling process i used something different which gives more of an airbrushed photoshop look. also had to photoshop the eyes to the correct color. anyways;
hope you all enjoy ! :)
11 images; more of a tech test. tried combining 4-5 lora files together, blending cyberpunk, overwatch and angelic themes. will come back to this model in the future.
Hope you all enjoy ! :)
27 images; I leaned into the SciFi aspect more this time, had fun with it. Character generates well but I lost quite a few images to the upscaling process. i'll be revisiting this character again in the future 100%. I find myself literally drawing hands into the pictures and having the ai do another pass to fix it. Kinda funny how I actually started drawing because of ai art. anyways,
hope you all enjoy ! :)
small edit; added 2 wallpaper images.
21 images; workflow went smoothly for this one.
Hope you all enjoy ! :)
9 images; I think I improved the likeness from my first post. used an embedding i bought from a random kofi store to pull this off :)
hope you all enjoy ! :)
18 images; completin requests!
generation went rather smoothly for this character
hope you all enjoy ! :)
18 images; just completting requests!
hope you all enjoy :)
16 images; completin requests! tech wise, there was only one lora i could find to replicate the likeness & oh my god it was annoying to work with. If the prompt was too long, or the wrong ai model used. she would instantly go from asian to sub-Saharan African. If I didnt use a specific sampling method, instant garbage, but i think i got a decent batch going for the first time using it.
hope you all enjoy ! :)
19 images; completin requests for long standing members!
hope you all enjoy ! :)
16 images; completing requests !
hope you all enjoy :)
16 images; completing a request for a long time supporter.
hope you all enjoy !! :D
2023-07-02 08:04:05 +0000 UTC View Post
11 images; honestly I'm not to sure about the likeness as ive never watched anything she's in at all. used a lycoris file & usually they have had accurate results. Probably gonna have to put this back in the oven and try again later, anyways
hope yall enjoy ! :)
2023-07-01 19:37:08 +0000 UTC View Post
11 images; completin requests. same workflow as previous post.
anime ai -> photo realistic -> upscalin -> photoshop -> back to upscaling -> finished product WOOOO
Hope you all enjoy ! :)
10 images; completin requests !
Hope you all enjoy :)
11 images; completing requests! tech wise nothing different from last post.
hope you all enjoy ! :)
2023-06-30 21:28:41 +0000 UTC View Post
8 images; tech wise went from anime->photorealistic->upscaling.
Hope you all enjoy !
12 images; finally made some improvements from my first post with her
hope you all enjoy ! :)
2023-06-29 21:43:28 +0000 UTC View Post
7 images; I think it turned out really well. lsdr upscaling really makes the red "pop" out more which I like. in my past attempts her forehead would come out 3x bigger, glad i got that issue fixed up now :)
hope you all enjoy ! :)
2023-06-29 18:36:12 +0000 UTC View Post
23 images; with the witcher season 3 coming out tomorrow, I thought it would be prudent to make a good set before someone asks me to. I like how this turned out, considering I only used an embedding to make the likeness work. got me rethinking my recent dependence on lycoris files.
hope you all enjoy ! :)
26 images; generation went rather well. figured out a way to render more detail in faces, think i did a good job here
hope you all enjoy! :)
