Legacy ImageMagick Discussions Archive

Hello.

Is Image Magick able to extract the;

1. Lens focal length in mm used to take the photo
2. Determine if a cropped or full frame sensor was used in the photo ? e.g. 1.6x, 1.5x
3. Determine if the photo was cropped or skewed with a Tilt-Shift lens, or in software ?

For Example:


Thank you IM Experts.

Best regards.

That metadata isn't in that file. (I already looked, curious about the pincushion distortion.)

"identify -verbose" will tell you everything that IM can read. I generally use exiftool, because that can see more.

Correct snibgo,

I am wondering if IM has the ability to determine #1-3 without metadata, I should have mentioned that.

From the image itself, ignoring metadata, it is not possible to determine the sensor size or lens focal length, or determine if the image has been cropped.

With intelligence and our knowledge of the subject, we can estimate if the field of view is "normal" or "wide-angle" or "telephoto". But these are only guesses about the image itself. A wide-angle photo can be cropped to be identical to be identical to a telephoto photo.

A tilt-shift lens might have two effects: shift can correct converging verticals etc ("perspective control"), and tilt can alter the plane of maximum sharpness. Detecting shift is not possible because a non-shift photo with a wider lens can be cropped to be identical.

Tilt could be used in this example to make the plane of best sharpness extend from the rooftop to the side of the road. Anything above or below that plane will be out of focus. In theory that can be detected. But not from such a small low-quality JPEG.

Similarly, detecting software transformations like skewing or perspective distortions is not possible from such a small low-quality JPEG.

Understood, thanks snibgo.

Similarly, detecting software transformations like skewing or perspective distortions is not possible from such a small low-quality JPEG.

What size/dpi would one need to analyze software transformations like skewing or perspective distortions ?

Regards.

I don't know. I've never examined an image, looking for evidence of software transformations like skewing or perspective distortions. But we might find stretching in some parts of the image and not others. For example if we point a camera up at a building we get converging verticals. We can correct this in software by stretching the top of the image to be wider. This might be visible as the stretching is horizontal only, not vertical.

But this effect would be difficult to detect in a small low-quality JPEG.