Skip to main content

⚠️ Please note that this topic or post has been archived. The information contained here may no longer be accurate or up-to-date. ⚠️

Lens correction tool

Comments

5 comments

  • Ian Wilson
    Moderator
    Top Commenter

    I'm not entirely clear what you are asking. The manufacturer profile applies correction data embedded in the image file by the camera when the photo was taken. It is specific to the particular lens in question, not just a generic profile for that manufacturer. 

    What you refer to as "the specific lens model profile" is presumably the profile developed from testing of the lens by Capture One's developers. They don't always do that for all lenses especially modern ones that can embed a manufacturer profile in the image file. 

    If there is both a manufacture profile and a Capture One produced profile, the amount of correction (and pixels "lost") may be more or less in the manufacturer profile.

    And you are not necessarily losing pixels based on the nominal sensor size. I'm looking at an image taken at the wide end of the Nikkor Z 24-200 zoom on a Nikon Z6ii. The image is shown as being 6048 x 4024 pixels in Capture One. This equals the Max Image Resolution published on the Nikon website. But if I crop unconstrained I can pull it out to 6382 x 4058 "recovering" some pixels cut out at the edge by the lens correction (manufacturer profile). If I turn the lens correction in Capture One off completely (changing from Manufacturer Profile to Generic) I can pull it out even further to 6794 x 4195. This is more than the "sensor resolution" of 6062 x 4041 published on the Nikon website. And if I turn off the option to hide distorted areas in Capture One, I see this.

    The sensor has recorded pixels outside 6764 x 4195, but more so in the corners. And they are not necessarily "lost" because the lens correction profile pulls them in somewhat.

    So to sum up, the published spec of the camera is 6048 x 4024, and that is the size of the image I get in Capture One with the lens correction on. The sensor of the camera actually captures a bit more than that. I can sometimes get away with recovering some extra pixels by cropping into the "corrected" area, if the 24mm wasn't quite wide enough. In an image like this one, it might look OK. But in an architectural shot, the distortion would show.

    Ian

    1
  • FirstName LastName

    Thanks for replying!

    The image below is taken at 40.5mm of the Simga 28-70mm f2.8 DG DN on a Sony a7IV. The camera spec is 7008x4672. There is both a manufacturer profile and a Capture One produced profile. When I select the manufacturer profile, I get 7008x4672. When I select the Simga 28-70mm f2.8 DG DN profile, I get 6891x4594. Why is the amount of pixels different between these two profiles?

    There's also a difference in distortion and light falloff as well. Which profile should I use and what would be the recommended numbers in order to produce an image with no visible distortion, sharpness falloff, and light falloff? The default is set at 0 for light falloff which seems off. 

    0
  • Ian Wilson
    Moderator
    Top Commenter

    The pixel sizes are different because the Manufacturer Profile and the Sigma 28-70mm profile developed by Capture One are applying slightly different corrections. (Presumably if Capture One's engineers tested the lens and decided that their profile would be just the same as the Manufacturer Profile, they wouldn't bother to provide their profile.)

    I took a screenshot of this and looked at it in Capture One. I rotated it very slightly (+0.23 degree) to get the outer edge of the door frame absolutely vertical and then put a guide up to it. As far as I can see, the edge of the door frame exactly follows the guide, with no curvature either outwards or inwards, in both cases. 

    (However, with the colour of the wall and the colour of the door frame being very similar, it is not easy to tell.)

    But for distortion, as far as I can see, there is very little difference. But this is with the zoom at 40.5mm. At the widest (28mm) you would probably see more distortion, and which profile would correct it better only you can say.

    As for light fall-off, the light switch (or whatever it is) top left looks a lot brighter in the first image than in the second, but which correctly matches how it should be is less easy to say. I put a colour readout half way down the wall and another in the lower left corner. The colours in the first example look like this

    and in the second example, they look like this.

    I'd say the second one makes the brightness more nearly the same. But you seem to have the light fall off right up to 120 in both cases. If you set it a little lower, perhaps the first one would be more equally bright, and the second one would be a bit darker in the corner.

    Really, I'd say that it doesn't much matter which of the two profiles you use. The number of pixels difference is not very great. You just need to decide which gives you images that look better to you, and use that one.

    Ian

    0
  • FirstName LastName

    I see. Thanks!

    0
  • JF
    Up-and-comer

    For my Nikon lenses, I much prefer the manufacturer profile (embedded in the images direct from the lens) over the Capture One profiles, particularly for getting the light falloff correction right.  That difference could be specific to my lenses, I don't know, but it's highly in the manufacturer's interests to get their profiles right because it strongly affects how people perceive their images.  Plus, those are probably the same corrections they apply to in-camera JPEGs too.

    2

Post is closed for comments.