I trained this LoRA to create images that mimic the effect that camera sees both under and over water which mostly called as over-under or split-shot underwater photography.
I called it as Waterline.
Trigger is waterline style. But it also needs a bit definition to give good results. Please check the example images to see what works as prompt.
It's some kind of beta now and i need to collect more training photos and also i will include the successful outputs from the current version to make it more stable.
For now, it doesn't always give perfect output but i still wanted to share to see if will get any interest.
Hope you like it.
Description
This is a beta version. Next update will be more stable.
FAQ
Comments (11)
Thanks for this Lora! I love it!
Actually i thought you disliked it when i see your rating, but good to hear that you liked it. Thanks for sharing your thoughts and contribution.
@CGArtist I love it! I do think there's some room for improvement.... it's not easy to get the waterline split consistently. I'm looking forward to future version <3
@swarmnet372 You're certainly right. Just because of this i was thinking to not make it public, but now i'm happy i did and i will work on it to make next version both flexible and stable. Thank you again!
This is a great idea for LoRA, some models already can do this but the results are somewhat inconsistent. Well done!
Thank you for sharing your thoughts! Currently, if the prompt data also was able to give the actual effect without LoRA, then you get two water lines in the same image with LoRA which seems weird. But i will try to find a way to make next version both flexible and stable.
beat me to it. I was working on this same LORA. found it very tricky to train. My goal was to have ti generate a character half submerged half above. which is even harder than a pure split. I'll try out this LORA of yours. thanks for releasing it.
check my profile pic subs for a image I made using pure luck and prompting
I love this idea! I had some interesting results. Unfortunately I failed to combine an everyday scene - like a living room or office with this effect. I really appreciate your work and would be happy, if you kept enhancing this Lora. Thanks a lot!
thanks for this LoRA. It can be a bit finicky and it took a fair amount of work with ChatGPT modifying the prompts but I think I got several good images. thanks again! :-)
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.