Testing the New Model

Thank you always for your support.

I am currently testing a new image generation model, and this process is taking some time. As a result, new artwork updates have been delayed. I sincerely apologize for the wait.

After several days of testing, I would like to share my current findings and plans for future operations.

Features and Issues of the New Model

• Excellent depiction of multiple people

Even when characters are in close contact, they do not blend together as much, and anatomical inconsistencies are almost nonexistent.

• Accurate depiction of hands and eyes

Since I previously spent a significant amount of time on corrections in these areas, this improvement could allow me to produce more artwork efficiently.

• High reproducibility of characters

The model captures intricate details of existing characters, allowing for highly accurate reproductions with minimal inconsistencies.

• Various "acts"

The model can depict complex compositions, poses, niche situations, and items that were previously difficult to express, making it possible to meet a wider range of requests.

• Difficulties in controlling art style

The output varies significantly between generations, making it challenging to maintain consistent quality and style. Additionally, despite my best efforts, I have concluded that fully replicating my previous art style is not feasible.

Future Policy

The new model offers many improvements over the previous one, but it also comes with its own issues, making it a trade-off in some aspects. However, since it significantly reduces the time needed for post-processing, I expect to be able to increase the frequency of updates.

That said, fully transitioning to the new model will inevitably result in a noticeable shift in artistic style. Therefore, I will strive to enhance quality and ensure that the final results meet the expectations of my supporters.

If you have any thoughts or feedback, please feel free to leave a comment.

Thank you for your continued support!