Not all cameras had the perfect units years ago, and not all still do.īut if a camera (or other software) rotates a JPEG after it’s been saved in that format, each decompression, rotation, and recompression progressively ruins the image. It’s possible to rotate JPEGs without adding additional image quality loss if the image resolution divides perfectly into the 8 by 8 or 16 by 16 grids that a given JPEG algorithm used. (“Lossless” compression is less efficient, but by identifying redundant patterns in a hunk of data, it allows an exact reproduction of the original data from the compressed file.) “Lossy” file formats like JPEG use approximations via mathematical formats to describe regions of an image some camera makers’ raw formats are lossy as well. If the camera were to produce with data stored in the same orientation as you took the picture, except for the single orientation that matches how data comes out of the sensor array, it would need to create a second image file into which to map each pixel from one orientation to the other, then delete that original file.Įarlier cameras lacked the computational power to handle this rotation, which was also RAM intensive: The image would need to be rotated before the sensor data was converted to a file format that compressed image data, like JPEG.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |