Hello,
Hi, I'd like to request a feature if it hasn't been already done.
use case is like this:
1)[user:] shot a RAW photo with 550D and then tranfer it to computer convert it as BMP and modify it by photoshop （save as BMP etc lossless compression format） etc.
2)[user:] transfer the modified BMP photo back to 550D with SD card.
3)[user:] click the re-construct button on Magic Lantern and select the BMP photo to be re-constructed.
4)[ Magic Lantern :] re-construct (redo jpge compression coding by the 550D jpeg compression algorithm and chip) this photo and write the compressed photo to SD card.
Is this feature useful?
Yes!
The photoshop jpge compression coding is different from 550D and other camera, there are softwares like JPEGsnoop that can check that if the Photo got modified by softwares, and if we can let a 550D re-construct a photo, that re-constructed will looks "REAL" when checking the jpge compression coding styles.
So you basically want a method to cheat, and transform a processed image to pose as if it had been created directly by the camera...?
If you are really eager to trick the system - something that is harmful for the art and industry of photography - you should figure out the exact jpeg compression details of canon cameras and compress into that from cr2 instead of requesting a feature that ethically does not fit ML.