Jump to content

in camera lens distortion correction


fhmillard

Recommended Posts

If you are shooting JPEGS yes it works. How does it work? I assume Nikon programs the processor with a generalized

(averaged) profile for that lens model that corrects for the center to corner light fall off, chromatic aberration, and possibly

geometric ( pincushion, barrel type) distortion.

 

Personally I leave the in camera lens correction turned off and. Use the lens profiles in Adobe Lightroom 5. I believe

CaptureOne Pro and DxO's raw processing software have similar features.

Link to comment
Share on other sites

Frank: yes, distortion correction throws away some of the captured pixels. The result of correcting the

distortion is that not all the pixels that were captured by the sensor rectangle map to a rectangle. Either

you lose a few pixels or you have to "make up" pixels (such as with a content-aware fill). Unless the lens

was very distorted, you won't lose much. I'd fix the distortion if it's significant (if you have straight lines in

the image) and otherwise preserve per-pixel sharpness by leaving well alone.

Link to comment
Share on other sites

<p>Distortion correction may or may not throw away pixels. Of course, straightening severe barrel distortion will require some cropping after the correction. When I use DxO Optics Pro to correct my 15mm fisheye from fisheye to rectilinear, the field of view goes from around 180-degrees diagonally to around 160-degrees. Interesting to note, in DxO the corrected image has the same number of pixels as the uncorrected image. (I don't understand how they do that and it'll likely vary from software to software).</p>

<p>Keep in mind that these corrections are not limited to geometric correction. Many in-camera and external correction software corrects for geometric, chromatic aberration, vignetting, pincushion, softness, etc. at every aperture and every focal length for a given lens/camera combination. If you're not using some form of digital lens optimization, either in-camera or in Raw conversion, you're not getting the most out of your lenses.</p>

Link to comment
Share on other sites

<p>David: It's true that if the boundaries of the image are already rectangular (and there are lenses that exhibit "moustache distortion" for which some of the image is distorted and farther-out bits are not) then no pixels are likely to be mapped beyond the boundary, In the general case, <i>most</i> distortions that are actually correcting anything will change the rectangular capture area to something that's not rectangular.<br />

<br />

Once you have an image that's not rectangular, you have to do something if you want a rectangular result. The options come down to cropping away useful bits of image or trying to fill in the gaps so that the resulting image is the shape that you want, unless you're very lucky that the edges line up and the rest of the image doesn't.<br />

<br />

I would expect most remappings to retain the original image resolution unless requested otherwise - this more or less preserves the maximum image quality where the image is undistorted. Many of the output pixels will be interpolated between captured values, and some will be averaging several pixels, depending on whether the image is being enlarged or shrunk at this point. The quality of the reconstruction will vary, but in general resizing any image to a fractionally larger size will mostly reduce image quality - one reason that I have such a loathing for 1366x768 HDTVs. I'm not suggesting that this loss of quality is huge, however - though there's a significant amount of stretching in moving from a fish-eye to a rectilinear image that suggests the edges of the image are likely to be very blurred (more for a circular fish-eye than a diagonal one).<br />

<br />

Geometric distortion correction blurs image detail through interpolation. Lateral/radial chromatic aberration correction performs a similar scaling, but on a per-channel basis (and is inferior to a better-corrected lens in part because the resizing applies only to a three-channel sampling of the input spectrum, not to each wavelength independently). Longitudinal/axial chromatic aberration is mostly heuristic-based, since full correction requires three-dimensional information. Vignetting correction results in a discrete remapping of pixel intensities, which can cause banding. Softness correction - sharpening - relies heavily on the dynamic range of surrounding pixels, and therefore introduces noise-related artifacts (and also tends to emphasize the rim of bokeh).<br />

<br />

That's not to say that digital correction shouldn't be performed. Indeed, many modern optics are being designed to trade - especially - distortion for absolute sharpness (and cost and size), knowing that the resulting image can be corrected digitally; better to blur an image slightly in correcting the distortion than to have the lens soft in the first place. If any of these artifacts are visible in the image and distracting, better to correct them than not. But I'd also be aware that there's no such thing as a free lunch, and that, especially if distortion is the primary artifact, it can sometimes be better not to bother with the corrections if what you're correcting for is not detrimental to the image.<br />

<br />

More to the point, if you're worried about capturing as much information as possible, turning the distortion correction off will achieve this. You may well want to correct things in software later, but at least you had everything that the camera captured as a starting point. Shooting raw is arguably the better starting point for this anyway, since that way you always have exactly what the camera captured and can choose how to correct distortion in your conversion software. For what it's worth, I have the lens corrections on in JPEG, but also capture raw files, so I know I can recover any cropped-out (or smudged) detail.</p>

Link to comment
Share on other sites

<p>yes,Andrew i noticed that when i took a few shots of a door correction off and on w/ 28mm f1.8, i wiil post later this week, after i get a good frame reference for comparison; the short of it is -- in camera correction appears to shift the subject, which is not surprising, but i have no quantitative evidence to supply now</p>
Link to comment
Share on other sites

<p>Andrew is right, all this manipulation can lead to unwanted distortion. Shooting in Raw and using the sharpest lens possible, shot at it's sweet spot aperture will result in best data to begin with; however, EVERY image is digitally manipulated as it's transferred from the sensor to the Raw file. It's filtered and processed. The same lens on different bodies will yield different results. Image problems are often the result of firmware problems with a particular body. Remember when the Canon 5D MkII was new and had purple fringing problems? A firmware update largely fixed that issue.</p>

<p>I like shooting in Raw and then having a powerful Raw conversion program that allows me to apply the corrections as I see fit. In general, I turn on all the default corrections in my DxO, but I'll occasionally turn one off or reduce its impact if I think that it's going too far. </p>

Link to comment
Share on other sites

<p>Nikon's automatic distortion correction doesn't correct distortion <em>all the way</em>; it just reduces it to a less noticeable level. I think the logic here is not so much that correcting the distortion fully will result in further loss in sharpness of areas (mostly corners) of the frame in the final image. DXO Pro seems to have quite well tabulated distortion data on supported camera+lens combinations and does distortion correction to a high level of precision but I don't otherwise like their software much and the fact that it can't run on edited NEFs. I've generally been happy with Nikon's approach of partial correction when I've needed it, though that's not all that often. I do use distortion correcion on the AF-S 80-400, for example.</p>
Link to comment
Share on other sites

<p>> I assume Nikon programs the processor with a generalized (averaged) profile for that lens model that corrects for the center to corner light fall off, chromatic aberration, and possibly geometric ( pincushion, barrel type) distortion.<br>

Most corrections are indeed profile-based. But correction of lateral CA is not, works also with 3rd-party lenses. There in also optional axial CA correction in CNX2 (Defringe in LR), not applied automatically (not as efficient and may have side-effects). <br>

> When I use DxO Optics Pro to correct my 15mm fisheye from fisheye to rectilinear, the field of view goes from around 180-degrees diagonally to around 160-degrees. Interesting to note, in DxO the corrected image has the same number of pixels as the uncorrected image. <br>

That's if you use same-aspect-ratio cropping. You can also choose the wider format that preserves a wider horizontal angle (the vertical one is more restriced), but the corner quality gets even more dismal due to the extreme bitmap stretching required. Straightening with the Panini projection (Hugin, Fisheye-hemi) requires less stretching but doesn't straighten the horizontals perfectly. </p>

Link to comment
Share on other sites

<p>Speaking of corner quality in de-fished fisheye shots, I just had a 50" print done of the shot below. The subject does lend itself well to hiding some of smearing distortion in the lower corners caused by the de-fishing process, but the overall print turned out stunning.</p>

<p><a title="Starburst by dcstep, on Flickr" href=" Sunstar src="http://farm6.staticflickr.com/5501/9379716012_633aed6dbe_c.jpg" alt="Starburst" width="800" height="534" /></a></p>

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...