MiniMax H3 Image can directly produce 4096×4096 images, about 16 million pixels. This number sounds large, but it needs to be converted into specific applications.
print.The commonly used standard for prints is 300 dpi, 4096 pixels divided by 300, which is about 13.65 inches, which is about 34.7 centimeters square. A piece of A4 paper has a short side of 21 cm and a long side of 29.7 cm. The 4K square image is cut into an A4 vertical version for printing. The accuracy is sufficient.
If it is a poster or roll-up poster that can be viewed from a distance, 150 dpi is generally acceptable, and at this time it can be printed to about 69 cm square. No matter how large an outdoor advertisement is, it still requires a special design draft or further enlargement.
E-commerce details page.Main product images on mainstream e-commerce platforms are mostly 800 to 1500 pixels square, with 4096 pixels being extra. The meaning of redundancy lies in cropping: a 4K image can be cropped into several images including the main image, detail enlargements, and partial close-ups. Each image has sufficient resolution, which is equivalent to being generated once and reused in multiple places.
social media.Most platforms will compress the uploaded images, and you may not be able to see the difference when sending 4K original images. The value of 4K here mainly lies in the later stage. It can be cut into vertical, horizontal or square versions without fear of blur.
The first frame of the video.4096 pixels wide, more than the 3840 pixels of 4K video. The first frame used to make Tusheng videos has a margin for definition, and the final video resolution is determined by the video model.
A small reminder: H3 Image's 4K is square 4096×4096. When doing horizontal or vertical layout, the actual pixels used will be less than this, so calculate the ratio before typesetting.
MiniMax’s H3 Image can now experience 4K rendering on official partner sites PixPix and PixTV. For e-commerce graphics, go to PixPix. For video first frames and storyboards, go to PixTV. Once the image is out, calculate it based on the dimensions above to see if it is enough.