I can't say for sure, but the image is encoded in ID3v2.4. 0 bytes are added to the image text. And if the image is encoded, then an additional three bytes are added to the frame header. This is the size identifier of the uncoded image. It is calculated as : H1 + the size of the image before encoding. If the image is not encoded, then these three bytes are not inserted. Now about the encoding rules... When reading the source image, which is added to the TAG if the 255 character is found in the ASCII table and written after it 0 bytes, then another 0 byte is added after the 0 byte. And also, if 255 characters are found, and after it there are characters >=224 (ASCII), then after the second character it is written 0 bytes. It turns out that an image with an additional 0 bytes is written to the metadata. Its size is larger than the original. Therefore, the first size identifier is inserted as usual, and the second one is placed before the MIME, as already described above. I hope it is translated correctly and understandable.

link to the diagram : https://disk.yandex.ru/i/IKlpgNeWrAthtQ

I am from Russia and used an online translator