<h2 id="unipixel-style-experiment">UniPixel Style 🎨 [Experiment]</h2>--[<a target="_blank" rel="ugc" href="https://civarchive.com/articles/20213">Experiment Logs 📙</a>]--[<a target="_blank" rel="ugc" href="https://civarchive.com/models/2024568?modelVersionId=2291322">Unipixel tool 🛠</a>]--[<a target="_blank" rel="ugc" href="https://civarchive.com/articles/20103">Pinboard Updated🥝</a>]--This LoRA is trained to have thematic as well as style flexibility while keeping consistent pixel sizes enough for it to be scaled back to pixel format.<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/e75d5486-ddb6-4abc-b2bf-f7264c2dc3bc/width=525/e75d5486-ddb6-4abc-b2bf-f7264c2dc3bc.jpeg" /><hr /><h3 id="usage-tips-(anima-v0.3)-:">Usage Tips (Anima v0.3) :</h3><ul><li>I personally use it at the weight of ( 1.2 ) so i recommend that</li><li>the LoRA needs to have its trigger words "<code>nv-p1xel4rt,</code>" in the prompt</li><li>score tags matters A LOT with this LoRA so try to stick to <code>score_6,</code> or lower<ul><li>the lower you go the less color depth you'll get</li></ul></li><li>mind the negatives prompt that you use it might include the pixel effect as a false negative.</li><li>keep in mind that some style tags might break the lora like (pixel art, 3d, or even theme tags)</li><li>so far only tested on <a target="_blank" rel="ugc" href="https://civarchive.com/models/2458426?modelVersionId=2836417">Anima Preview Base 3</a> (used in the showcase pics)</li></ul><hr /><h3 id="usage-tips-(il-mini-ver)-:">Usage Tips (IL mini ver) :</h3><ul><li>this one have lets pixel density then the regular ones (trained on 8bit, 12bit pixel-art)</li><li>I personally use it at the weight of 1 so i recommend that</li><li>the LoRA needs to have its trigger words "<code>nv-mp1xel4rt,</code>" in the prompt</li></ul><h3 id="usage-tips-(il-regular-ver)-:">Usage Tips (IL regular ver) :</h3><ul><li>I personally use it at the weight of 1 so i recommend that</li><li>the LoRA needs to have its trigger words "<code>nv-p1xel4rt,</code>" in the prompt</li><li>so far only tested on <a target="_blank" rel="ugc" href="https://civarchive.com/models/827184?modelVersionId=1761560">WAI-NSFW-illustrious-SDXL V14</a> (used in the showcase pics)</li></ul><hr /><h3 id="usage-tricks-:">Usage Tricks :</h3><ul><li>control your pixel density (more pixels or less pixel per subject) by<ul><li>using different SDXL supported resolutions (refer to <a target="_blank" rel="ugc" href="https://civarchive.com/articles/20213">Experiment Logs</a> )</li><li>or prompting for the subject to be closer or further away</li></ul></li><li>you can easily get character sprites by adding "<code>simple background, white background, full body,</code>"</li><li>you prompt for props by using the prop name along with "<code>still life, no humans, simple background, white background,</code>"</li><li>you can get the image back to true pixel format by<ul><li>resizing it down by exactly 8 times using "nearest neighbor" sampling method at any photo-editing software (like Photoshop)</li><li>using the [<a target="_blank" rel="ugc" href="https://civarchive.com/models/2024568?modelVersionId=2291322">Unipixel tool</a>] i posted</li></ul></li></ul><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/073c45f1-4b70-46ea-a6e2-c39b71e467a1/width=525/073c45f1-4b70-46ea-a6e2-c39b71e467a1.jpeg" />

Files with SHA256: 17c16c0a48e70dba107fbca52b7eb54671c609f0a3ecfd85f2c5387837ec99bf