>>106153445
Nope, not noticeably sharper. Maybe the VAE needs an EQ tune-up like SDXL/Flux got. Those always produce more sharpness when used during inference.