The effect of the model can be viewed based on the return image; I really like this myself.

The

XL model doesn't necessarily have to be 1024X1024. When drawing a picture of a person, I can open 1024 in height, and I often use 536 or 768 in width. This is an appropriate

ratio.

Also, there is no need to fix the image in HD; there is no need for this, just open the AD plug-in.

My swimsuit series TAG can produce great results using this model.

You can find TAG pictures at other Nekochan models, all of which are pleasantly surprised.

The strength of the

XL model is its understanding of semantics. It's not HD image quality, so it really doesn't need an aspect ratio

of up to 1024+.