As stated on the original 1.5 release, "Vulcan" is part of my brand new and personal color series where I aim to create an overall style but oriented mostly for certain color aesthetics so it can be used for characters, objects, landscapes, etc. This time is the turn of red/yellow dominance.
"Vulcan" aims to create globalized creations but with red and yellow colors (and in this case there won't be too much derivatives except orangish) and high saturated images. This LORA is extremely good for comics, certain graphic art, posters etc.
Main token is v5lcn and class is style so you can use v5lcn Style or just v5lcn. I recommend this time the first case.
How to use this LORA?
*************************
You can refer to the sample images I upload in the post since it changes behaviour in ComfyUI and Automatic. Even generating 512x768 (and Hi-Res) Vs. 1024 native.
- For characters a portrait of "concept" in the style of v5lcn BUT careful cause SDXL behaviour is different than 1.5
- For landscape and objects v5lcn style concept or a "concept" in the style of v5lcn
- You can control mono colors or influence them using "red" or "yellow" in the prompt. And in the case of SDXL, even more In most cases you will get a mix but if you want certain focus remember you can use that.
Which are the best values?
*******************************
- You can go from 0.7/0.8 onwards till 1 and 1 onwards for a more "v5lcn style".
If you are focusing a global dominance of black/red/yellow and found "annoying" colors, rise the the LORA to 1.X or try to simplify the prompt. This LORA is good with few words.
- To control the saturation is recommended to start from a CFG scale of 5 and play from there. Remember the red and yellow words. They're sub tokens that works better even in SDXL.
Which models are good for this LORA?
********************************************
- Tested only in SDXL 1.0 but should work on everyone, I think. Test it and comment how it went ;).
This LORA is GOOD for:
******************************
- Highly saturated/contrasted art and red/yellow landscapes. - Comic R&Y style in duotones contrasted and saturated. - Graphic design style for your creations. - Duotone creations with saturated and lined values.
- Saturated and high contrasted colors in general with highly R/Y dominance.
- Globalized creations (such as really anything you want). Just remember what happens with anime models.
This LORA is NOT designed for:
***********************************
- Realistic things (maybe) since it's trained on certain drawings and illustrations. Should be difficul to guide it to a realistic basis.
IMPORTANT!
I wrote an article here and it's like my finetunning diary for the ones interested with releases upcoming, trainings done, exclusive patreon models previews, personal thoughts, etc. So if you want to see what I'm planning or what I'm into, feel free to take a look.
As always thanks to everyone reviewing, giving feedback , commenting and supportin us in Patreon. Keep supporting us so we can keep creating and delivering quality content.
Hi! Releasing the SDXL ver. of Vulcan.
As stated on the original 1.5 release, "Vulcan" is part of my brand new and personal color series where I aim to create an overall style but oriented mostly for certain color aesthetics so it can be used for characters, objects, landscapes, etc. This time is the turn of red/yellow dominance.
"Vulcan" aims to create globalized creations but with red and yellow colors (and in this case there won't be too much derivatives except orangish) and high saturated images. This LORA is extremely good for comics, certain graphic art, posters etc.
Main token is v5lcn and class is style so you can use v5lcn Style or just v5lcn. I recommend this time the first case.
How to use this LORA?
*************************
You can refer to the sample images I upload in the post since it changes behaviour in ComfyUI and Automatic. Even generating 512x768 (and Hi-Res) Vs. 1024 native.
- For characters a portrait of "concept" in the style of v5lcn BUT careful cause SDXL behaviour is different than 1.5
- For landscape and objects v5lcn style concept or a "concept" in the style of v5lcn
- You can control mono colors or influence them using "red" or "yellow" in the prompt. And in the case of SDXL, even more In most cases you will get a mix but if you want certain focus remember you can use that.
Which are the best values?
*******************************
- You can go from 0.7/0.8 onwards till 1 and 1 onwards for a more "v5lcn style".
If you are focusing a global dominance of black/red/yellow and found "annoying" colors, rise the the LORA to 1.X or try to simplify the prompt. This LORA is good with few words.
- To control the saturation is recommended to start from a CFG scale of 5 and play from there. Remember the red and yellow words. They're sub tokens that works better even in SDXL.
Which models are good for this LORA?
********************************************
- Tested only in SDXL 1.0 but should work on everyone, I think. Test it and comment how it went ;).
This LORA is GOOD for:
******************************
- Highly saturated/contrasted art and red/yellow landscapes.
- Comic R&Y style in duotones contrasted and saturated.
- Graphic design style for your creations.
- Duotone creations with saturated and lined values.
- Saturated and high contrasted colors in general with highly R/Y dominance.
- Globalized creations (such as really anything you want). Just remember what happens
with anime models.
This LORA is NOT designed for:
***********************************
- Realistic things (maybe) since it's trained on certain drawings and illustrations. Should be difficul to guide it to a realistic basis.
IMPORTANT!
I wrote an article here and it's like my finetunning diary for the ones interested with releases upcoming, trainings done, exclusive patreon models previews, personal thoughts, etc. So if you want to see what I'm planning or what I'm into, feel free to take a look.
As always thanks to everyone reviewing, giving feedback , commenting and supportin us in Patreon. Keep supporting us so we can keep creating and delivering quality content.
David.