How to use the SD1.5 model

*This text was generated using Google Translate.* These are the results of my research and testing on how to use the SD1.5 model. The unique retro feel is irresistible.

#Touhou Project #Marisa Kirisame #Puella Magi Madoka Magica #Akemi Homura #Kaname Madoka #Akemi Homura #Kaname Madoka

Published on Aug 2, 2026

七夢

More to See

取扱説明書を作ります。参考画像１は、アニメ「魔法少女まどか☆マギカ」のキャラクター「鹿目まどか」です。参考画像２は、アニメ「魔法少女まどか☆マギカ」のキャラクター「暁美ほむら」です。この２人のキャラクターが解説役です。ちびキャラにデフォルメ化させて使用します。以下が取り扱い説明書の文章です。わかりやすく、ポスター形式１枚にまとめます。 Title: How to Play with SD 1.5 Models * These findings are based on my own research and testing, despite having limited experience generating images with SD 1.5. * Google Translate was used for the translation. (About Image Size) * I recommend generating images at "512x512" or "512x768" (portrait orientation). Generating images at larger sizes increases the risk of visual artifacts or distortion (resulting in a chaotic mess). Depending on the specific LoRA used, however, it may be possible to generate larger images without issues. * After generation, you can also experiment by using the upscaling function to increase the image size or enhance image quality. * It seems to struggle a bit with complex poses. * It is probably best to disable "fast generation" to keep costs down while you play. * Please refer to the image below to locate the upscaling function; it is indicated by the red square. 上記メッセージの後に、参考画像３をあてはめます。必ずそのままあてはめてください。※この行は指示ですので、文字として記載はしないでください。誰がみても分かりやすい、ポスター形式の資料としてください。高画質。高品質。文字は正確に記載します。

Show Parameters

Size: 512 x 512

Model & LoRA used

Similar Images

How to use the SD1.5 model

*This text was generated using Google Translate.* These are the results of my research and testing on how to use the SD1.5 model. The unique retro feel is irresistible.