>>30819

The problem with generating an image of a woman on a horizontal roasting spit is the sheer amount of time required. Even if the prompt is written to be physically accurate and sounds innocuous to the image generator, the error rate remains high.

It took two months to produce 500 images of women on horizontal roasting spits—featuring various ages, hair colors, and lengths, and backgrounds ranging from forests and beaches to schoolyards—to get the wording right, regenerate images, perform post-processing, and test different levels of realism.

I suspect you don't have that kind of time. Even with the correct prompt, the AI ​​might only render one out of 500 images reasonably well; this means post-processing is essential. For instance, the angle at which the spit enters the body is often incorrect and requires adjustment in a graphics program. I invest a great deal of time in this—effectively treating it as a full-time job, albeit an unpaid, voluntary one.

For a good presentation, I recommend a reclining pose on a serving platter; it takes less time to generate such images. To achieve the roasted skin texture, use this phrase element repeatedly: `((gold-brown-dark-roasted-Skin: 1.5))`. The parentheses and number indicate the element's weight within the prompt. Write a descriptive text of at least 40 sentences and repeat the bracketed phrase at least three times. Note that the more cartoonish your character is, the harder it is to achieve this skin texture.

Good luck. I look forward to the continuation of your story.