5
\$\begingroup\$

I am trying to focus on the differences between digital zoom and cropping an image afterwards.

This is discussed in detailed here: Is digital zoom really useful? focused on the compression difference, i.e.: in case the camera is compressing the images, most cameras will probably do the cropping before compression.

But what about the exposure adjustment? Is this typically done using only the zoomed frame? I know this might depend on the camera but I would expect to be some tendency. Or whether is never the case.

There is one answer in that question that suggests this, but I wonder whether that is actually the case.

Please note that I'm not referring here to advantages or disadvantages because that is completely arbitrary. I am just interested in the differences.

Improve this question
\$\endgroup\$
3
  • 2
    \$\begingroup\$ You could figure this out for your camera specifically by printing a large, black rectangle with a smaller, white rectangle on a sheet of paper and seeing if it selects the same exposure when zoomed in on the white area as when zoomed out show both. \$\endgroup\$
    – Blrfl
    Commented Apr 25, 2018 at 19:20
  • \$\begingroup\$ Exposure meters and the algorithms that cameras use to process the information from them vary widely. There's no universal way that cameras with digital zoom calculate exposure. \$\endgroup\$
    – Michael C
    Commented Apr 26, 2018 at 4:37
  • \$\begingroup\$ In digital zoom: In best case the exposure will just be calculated for the cropped region and the cropped out region will be ignored. But this is strongly depending on the calculation algorithm. E.g. my Nikon provides different light measuring methods: spot, center, matrix, and some others i think. At center and spot with center focus point there will probably not be much difference, between digital zoom and full image. \$\endgroup\$
    – Horitsu
    Commented May 9, 2018 at 5:18

3 Answers 3

Reset to default
2
\$\begingroup\$

Yes, at least for a Canon IXUS 160 in center weighted average metering mode. I applied the idea from mattdm's answer and took as best I could the same photo at maximum optical zoom, and then maximum optical + 4x digital zoom.

For the photo at max optical zoom (with a large dark area in the frame), the camera chose 1/250s, f/6.9, ISO 800, while for the photo at max optical + digital zoom the camera chose 1/250s, f/6.9, ISO 200 - or 2 stops darker. This is obvious from the photos where there is a lot more detail in the my car's bonnet.

Photos below, pretty much straight out of camera apart from recompressed in Lightroom to get them below imgur's 2 Mb size limit. I make no apologies for them being very boring.

Max optical zoom (1/250s, f/6.9, ISO 800): enter image description here

Max optical + 4x digital zoom (1/250s, f/6.9, ISO 200): enter image description here

Improve this answer
\$\endgroup\$
10
  • \$\begingroup\$ Awesome! Empirical results ftw. \$\endgroup\$
    – mattdm
    Commented Jun 10, 2018 at 16:28
  • \$\begingroup\$ Indeed, experimental measurements have the final word. I would have added a post cropped picture without the zoom for completeness, but that's for perfectionism :P. \$\endgroup\$
    – myradio
    Commented Jun 11, 2018 at 7:33
  • \$\begingroup\$ Actually, IMO this is a bad comparison... it is comparing a bad exposure to a better exposure (of the car) and it is due to a metering error. If the exposure had not changed, the exposure of the car would not have changed it would be the same (i.e. spot metering). \$\endgroup\$
    – Steven Kersting
    Commented Jun 11, 2018 at 18:16
  • \$\begingroup\$ I think you're missing the point of the pictures - they clearly demonstrate that the exposure changes when digital zoom is used, which is the question asked. I make no claim at all to them being well exposed, good composition or anything else at all! \$\endgroup\$
    – Philip Kendall ♦
    Commented Jun 11, 2018 at 18:51
  • 1
    \$\begingroup\$ Indeed, @StevenKersting, that is the point. This pictures are showing that the exposure did change. I don't know why do you consider this to be metering error. If we trust this pictures, he put the camera in auto-mode and without moving it he took two pictures, namely with and without digital zoom. The metering was different because the camera then changes the exposure when on digital zoom. \$\endgroup\$
    – myradio
    Commented Jun 12, 2018 at 7:21
2
\$\begingroup\$

This would be easy to test: simply stand in a dark room on a sunny day and point your camera at a bright window. Does the exposure change as you zoom in to only include the window in the frame?

I don't have a camera with this feature to test, except my Android phone — and that definitely makes its decisions based on the displayed image. I found this in the manual for a Nikon Coolpix, though:

enter image description here

Depending on how we interpret that, this camera either a) just uses the visible portion, or b) attempts to replicate that as closely as possible by using metering modes which are biased towards the center of the image, even if that doesn't precisely match the zoom. (The text sounds most like "b", but... who knows how the communication from engineering to technical writers went.)

Improve this answer
\$\endgroup\$
-2
\$\begingroup\$

Digital zoom is just cropping... there is absolutely no difference whether done in camera or in post. And there is no need to compensate/recalculate the exposure as the light per sensor area received by the camera is the same.

Improve this answer
\$\endgroup\$
4
  • 1
    \$\begingroup\$ But what if you have a large dynamic range in the scene, and the center happens to be much brighter than the image as a whole? Imagine you're indoors shooting out an open window on a sunny day. If you don't zoom, the exposure will try to balance the indoor light and the outdoor (probably badly, leaving the outdoors overexposed and the walls still dark). If you use digital zoom so the entire frame just shows outdoors, will the camera be smart enough to change the exposure? I imagine it might vary from camera to camera. I don't have any cameras with this feature anymore so I can't test.... \$\endgroup\$
    – mattdm
    Commented Jun 10, 2018 at 14:17
  • \$\begingroup\$ I agree with @mattdm, certainly is not hard to imagine someone (some company) implemented that. Also, your statement in general is NOT true since unless you are storing raw images, the compresion will happen before cropping in one case and that is without any doubt a difference. \$\endgroup\$
    – myradio
    Commented Jun 10, 2018 at 14:21
  • \$\begingroup\$ mattdm, What you are explaining are just metering errors. I.e. do you want the outdoors properly exposed or the walls? Whichever you choose, there is no need to change the exposure settings as you change the zoom. I.e. choose proper settings for the outdoors portion (spot meter) and zoom in, the settings required to maintain the proper exposure will be the same. Allowing the camera to change the settings/do whatever it wants due to improper metering can/will result in all sorts of varied (bad) results. \$\endgroup\$
    – Steven Kersting
    Commented Jun 11, 2018 at 18:25
  • \$\begingroup\$ As I said in the other comment, this might be just semantics at this point. You call this things metering error. But where are you attributing that this error is coming from, when is is reproduced by changing the digital zoom settings? The point here is that the camera is changing the exposure setting when digital zoom is on, which was ultimately, my question. Whether that is a good idea to implement, a very silly one, useless or super useful is another matter entirely. \$\endgroup\$
    – myradio
    Commented Jun 12, 2018 at 7:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.