Concept Art Helper - Quick Designs

David here presenting you the final version (and exclusive) of my C.A.H. - Quick Designs LORA for Stable DIffusion XL. This is probably one of my best (if not the best work) and I suggest you to give this a try.

"Quick Designs" is a bit different of the things I use to train in my day by day and it's NOT intended to create amazing and stunning art pieces or perfect and accurate AI images. On the contrary, it avoids that. I bring you the power of the "speed painting" technique and the "unfinished brush-stroking" so you can use this LORA to prototype your own designs and finish them later on any Digital Painting software (Photoshop, Krita, etc.)

Basic Info:
I created this using several concept art sources (including personal works, yes, mine) so you can create quick concept art like you do normally on photoshop (specially photobashing) and after, overpainting them on a polishing phase.

Main token is jh**k3 and class is style so you can use jh**ke style or just jh**k3. I recommend the first option.

This LORA doesn't have powered subtokens but there is a tendency when you use word "grey background"/"white background" and "concept art" that the gets more power by itself.

You can use negative embeddings and words but this LORA has been proved to work without them in a normal way. Choise will be yours but I'd recommend to don't use it.

How to use this LORA?
**********************

You can refer to the sample images I upload in the post but:

- For characters * jh**k3 style Portrait/side portrait of concept... / a jh**k3 style concept...
- You can alter the pose and view using "front view", "side view", "back view", etc.

- For landscape and objects a [concept/object/whateveryouwnt] in the style of jh**k3

Which are the best values?
*******************************

- Extremely recommended 0.7-0.8 but you can use values from 0.6 onwards. Carefult with 1 due how AI see the brush stroking.
- CFG Scale in this case is a bit different than usual. CFG Scale recommended is 2-5. Due to the inner aesthetics and nature of the LORA.
- Clip Skip: I'd say 1 but you can go 2 too. Choice is yours.

Which models are good for this LORA?
********************************************

- Tested on: SDXL Base 0.9 and SDXL Base 1.0 (others will work EXCEPT PONY).

Extra Info:

This LORA is GOOD for:
******************************

- Character prototyping of several types.
- Blockout phase.
- Speed painting style artworks
- Brainstorming fast in iteration phase.
- Unfinished brush strokes typical of quick concept art and rough blocking.
- Sci-Fi/Fantasy oriented aesthetics (specially mechs)
- Turnarounds (with right prompting)
- Grayscale blockouts (pre-polishing)

THIS LORA IS NOT GOOD FOR:
******************************************

- Extreme detailed characters (some degree to detailed aesthetics though)
- hyper realistic close ups .
- Clean hands and faces on mid/long distance.
- One click -> finish piece (except if you aim Speed Painting style)