If you’re working with a model designed to generate anime images, chances are that it supports adjusting viewing angles. Mastering control over these angles can significantly enhance how well the final image aligns with your creative vision. Let’s explore how to use this feature effectively.
Overview
We will discuss tags that control the viewer’s position. Consider this as the position of the photographer taking a picture of the subject in the image.
| Prompt tags to use | Position of the photographer |
| straight-on | In front of the subject |
| from side | On the side of the subject |
| from behind | Behind the subject |
| from above | above the subject (most likely front and above) |
| from below | below the subject (most likely front and below) |
In below examples, animaPencilXL_v100.safetensors was used to generate images, but you can also use bluePencilXL_v050.safetensors. Just be sure to select an anime model. All styles were deselected.
straight-on
Positive prompt: 1girl, 20 yo woman, a castle and a puppy, straight-on, sweater, jeans, best quality
Generation parameters
{
“Prompt”: “1girl, 20 yo woman, a castle and a puppy, straight-on, sweater, jeans, best quality”,
“Negative Prompt”: “nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name”,
“Fooocus V2 Expansion”: “”,
“Styles”: “[]”,
“Performance”: “Speed”,
“Resolution”: “(1152, 896)”,
“Sharpness”: 2,
“Guidance Scale”: 4,
“ADM Guidance”: “(1.5, 0.8, 0.3)”,
“Base Model”: “animaPencilXL_v100.safetensors”,
“Refiner Model”: “None”,
“Refiner Switch”: 0.5,
“Sampler”: “dpmpp_2m_sde_gpu”,
“Scheduler”: “karras”,
“Seed”: 330627371048157193,
“LoRA 1”: “sd_xl_offset_example-lora_1.0.safetensors : 0.1”,
“Version”: “v2.1.864”
}
from side
Positive prompt: 1girl, 20 yo woman, a castle, from side, sweater, jeans, best quality
Generation parameters
{
“Prompt”: “1girl, 20 yo woman, a castle, from side, sweater, jeans, best quality”,
“Negative Prompt”: “nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name”,
“Fooocus V2 Expansion”: “”,
“Styles”: “[]”,
“Performance”: “Speed”,
“Resolution”: “(1152, 896)”,
“Sharpness”: 2,
“Guidance Scale”: 4,
“ADM Guidance”: “(1.5, 0.8, 0.3)”,
“Base Model”: “animaPencilXL_v100.safetensors”,
“Refiner Model”: “None”,
“Refiner Switch”: 0.5,
“Sampler”: “dpmpp_2m_sde_gpu”,
“Scheduler”: “karras”,
“Seed”: 330627371048157193,
“LoRA 1”: “sd_xl_offset_example-lora_1.0.safetensors : 0.1”,
“Version”: “v2.1.864”
}
from behind
Positive prompt: 1girl, 20 yo woman, a castle, from behind, sweater, jeans, best quality
Generation parameters
{
“Prompt”: “1girl, 20 yo woman, a castle, from behind, sweater, jeans, best quality”,
“Negative Prompt”: “nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name”,
“Fooocus V2 Expansion”: “”,
“Styles”: “[]”,
“Performance”: “Speed”,
“Resolution”: “(1152, 896)”,
“Sharpness”: 2,
“Guidance Scale”: 4,
“ADM Guidance”: “(1.5, 0.8, 0.3)”,
“Base Model”: “animaPencilXL_v100.safetensors”,
“Refiner Model”: “None”,
“Refiner Switch”: 0.5,
“Sampler”: “dpmpp_2m_sde_gpu”,
“Scheduler”: “karras”,
“Seed”: 330627371048157193,
“LoRA 1”: “sd_xl_offset_example-lora_1.0.safetensors : 0.1”,
“Version”: “v2.1.864”
}
from above
Positive prompt: 1girl, 20 yo woman, a castle, from above, sweater, jeans, best quality
Generation parameters
{
“Prompt”: “1girl, 20 yo woman, a castle, from above, sweater, jeans, best quality”,
“Negative Prompt”: “nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name”,
“Fooocus V2 Expansion”: “”,
“Styles”: “[]”,
“Performance”: “Speed”,
“Resolution”: “(1152, 896)”,
“Sharpness”: 2,
“Guidance Scale”: 4,
“ADM Guidance”: “(1.5, 0.8, 0.3)”,
“Base Model”: “animaPencilXL_v100.safetensors”,
“Refiner Model”: “None”,
“Refiner Switch”: 0.5,
“Sampler”: “dpmpp_2m_sde_gpu”,
“Scheduler”: “karras”,
“Seed”: 330627371048157193,
“LoRA 1”: “sd_xl_offset_example-lora_1.0.safetensors : 0.1”,
“Version”: “v2.1.864”
}
from below
Positive prompt: 1boy, a 30 yo warrior, a castle, from below, blue eyes, full body armor, best quality
Generation parameters
{
“Prompt”: “1boy, a 30 yo warrior, a castle, from below, blue eyes, full body armor, best quality”,
“Negative Prompt”: “nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name”,
“Fooocus V2 Expansion”: “”,
“Styles”: “[]”,
“Performance”: “Speed”,
“Resolution”: “(1152, 896)”,
“Sharpness”: 2,
“Guidance Scale”: 4,
“ADM Guidance”: “(1.5, 0.8, 0.3)”,
“Base Model”: “animaPencilXL_v100.safetensors”,
“Refiner Model”: “None”,
“Refiner Switch”: 0.5,
“Sampler”: “dpmpp_2m_sde_gpu”,
“Scheduler”: “karras”,
“Seed”: 2269649633924201714,
“LoRA 1”: “sd_xl_offset_example-lora_1.0.safetensors : 0.1”,
“Version”: “v2.1.864”
}
Combining tags
Note that you can combine these tags. However, you might not achieve the desired result on your first attempt, so be prepared to experiment with several iterations.
from above & from side
Positive prompt: 1girl, 20 yo woman, a castle, from above, from side, sweater, jeans, best quality
Generation parameters
{
“Prompt”: “1girl, 20 yo woman, a castle, from above, from side, sweater, jeans, best quality”,
“Negative Prompt”: “nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name”,
“Fooocus V2 Expansion”: “”,
“Styles”: “[]”,
“Performance”: “Speed”,
“Resolution”: “(1152, 896)”,
“Sharpness”: 2,
“Guidance Scale”: 4,
“ADM Guidance”: “(1.5, 0.8, 0.3)”,
“Base Model”: “animaPencilXL_v100.safetensors”,
“Refiner Model”: “None”,
“Refiner Switch”: 0.5,
“Sampler”: “dpmpp_2m_sde_gpu”,
“Scheduler”: “karras”,
“Seed”: 1638067457742175727,
“LoRA 1”: “sd_xl_offset_example-lora_1.0.safetensors : 0.1”,
“Version”: “v2.1.864”
}
from below and from behind
Positive prompt: a gigantic scifi fighter, a skyscaper, from below, from behind, full body armor best quality
Generation parameters
{
“Prompt”: “a gigantic scifi fighter, a skyscaper, from below, from behind, full body armor best quality”,
“Negative Prompt”: “nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name”,
“Fooocus V2 Expansion”: “”,
“Styles”: “[]”,
“Performance”: “Speed”,
“Resolution”: “(1152, 896)”,
“Sharpness”: 2,
“Guidance Scale”: 4,
“ADM Guidance”: “(1.5, 0.8, 0.3)”,
“Base Model”: “animaPencilXL_v100.safetensors”,
“Refiner Model”: “None”,
“Refiner Switch”: 0.5,
“Sampler”: “dpmpp_2m_sde_gpu”,
“Scheduler”: “karras”,
“Seed”: 7103595447640470348,
“LoRA 1”: “sd_xl_offset_example-lora_1.0.safetensors : 0.1”,
“Version”: “v2.1.864”
}
Congratulations! You’ve successfully mastered the technique to adjust the viewing angles of subjects in your images. This newfound skill will undoubtedly enhance the creativity and composition of your artwork!