Croping ratio 3/2 or 16x9

I use a compact Canon G7x. With this camera, we can choose the format, from 1x1 to 19x9. I am used to 3x2 format. I wanted to correct such a picture with Dxo. The picture was sized 5472x3548 by the Dxo metadonnées. As I wanted to adjust the cropping with 3/2 ratio or with keep the ration, I was surprised to see on the DxO screen the cropped picture much narrower than the original one, but was also sized 5472x3648 by DxO (see screen picture).

something was crazy. I also crop with no constraint, resulting file is sized 6260x3548, about 16x9 .
In conclusion as I understand, whatever the ratio taken, the Canon G7x takes a 6260x3648 picture, but the pixels out of the chosen ratio are hidden. When loading the picture with Dxo, the full picture is shown, but the camera choosen ratio is given in the metadonnées. This is confusing, we have a pictue, we think it is 3/2, but it is not possible to have the full picture in 3/2!

Two possible solution : 1-show the whole picture with is real ratio 6260x3648. We know it and if we need to have a 3/2 picture, we just have to crope it.
2-or you can show the 3/2 picture as before with 5472x3548 in metadonnees .


I have the same with my Sony RX100 V.
While I shoot with 3x2 ratio, I also have a greater image, and this allow to work on perspective for instance without loosing any visible pixel that I see in the viewer when shooting.
As the format chooses is 3x2, logically DPL shows by default this format, that’s why we still have some more hidden pixels…

It can be seen as an advantage, always shooting in panoramics mode and cropping with Dxo. Furthermore, it helps when we want to correct perspective, we need sometimes to have pixels outside the frame…
But it makes file about 20% heavier because, even hidden, pixels need megabits of memory. And this way of working should be official, means it should be agreed by Canon, Sony and other Nikons… Furthermore the right number of pixel should be indicated in the metadonnee

Good morning,

Let me ask @Marie to have a look at it.

Thank you
Svetlana G.


in PhotoLab when a RAW image is processed full file is used so images will always have same dimension whatever option of crop (ratio) or mode (L/M/S or dimension) has been set in the camera EXCEPT if the RAW file itself is cropped or reduced.
Use of crop from camera is in our backlog but there are always other tasks to do for the moment :frowning: .

An other thing you have to have in mind with cameras like Sony RX or Canon GX is the embedded distorsion correction for JPG images. They are part of cameras which apply some optical correction on their JPG images.Usually it concerns distorsion.
PhotoLab doesn’t apply same correction as the camera (because usually they are not good enough or incomplete) so what you have is

  • JPG image from camera has correction applied
  • RAW image processed in PhotoLab (without module) doesn’t have that correction
    And as correction from DxO Optics module is not the same as the one from the camera it’s possible you end up with different field of view (we let more than the camera).

So this adds to the fact your RAW image converted by PhotoLab won’t be the same as the JPG from the camera but you’ll have the freedom to get better.



And this perfectly fine for me.
Thanks @Marie


Thanks Marie for your explanation. When I use my G7x camera, I always use Raw+Jpeg at the maximum resolution. I use most of the time the 3/2 ratio, other format only for special purposes.
Raw picture , after processed with DxO, is used for poster, exposition… I got an other format out of Dxo, 1920px for internet , small pictures or pictures for newspaper. Jpeg, without DxO process are used when I need pictures quickly. They are sometimes used as a back up. As I undestand from your paper, the best way would be always taking picture at the maximum ratio (16/9) ans crop it with Dxo. It will give the best freedom in formart choosing, and the best quality of pictures

Hello @Levendeen,

format 16/9 is not the max ratio of any camera, it’s just a crop of 3/2 (higher part and lower part of the image are crop).
So the best way, for you, is to use the native ratio of the camera.


1 Like