Can Capture One see more than your camera?

There’s something odd about the way Capture One crops certain RAW files. Now and again, when you select the crop tool, you’ll see that Capture One has chosen a default crop that’s actually smaller than the full image area.

Capture One Pro 11.1 review

This might seem like a minor operational annoyance and that you just need to recrop manually, but there’s actually a bit more to it than this. It’s not that Capture One is failing to include the full image area in its default crop – it’s actually seeing areas of the image outside those captured by the camera.

This sounds like it shouldn’t be possible, but I’ve tested it and it’s true. By extending the crop marquee to the full image area you get an image that’s wider and taller than the one captured and displayed by the camera and with more pixels.

The sample image I’m using here is from an Olympus OM-D E-M1 which captures images with a native resolution of 4,608 x 3,456 pixels, or 15.9 megapixels. The recropped Capture One image measures 5,027 x 3,658 pixels, which is 18.4 megapixels.

Capture One has not simply resampled the existing image to a new size – it’s captured additional areas of the scene outside the frame displayed by the camera.

It sounds crazy but it’s true, and here’s a step-by-step walkthrough that shows you how it works.


Step 01: Watch your thumbnails

Here’s a black and white seascape I’m editing in Capture One Pro. Over on the right is the Browser panel, and there’s something interesting about the thumbnails its displaying – some of them have greyed out areas outside the main image.


Step 02: Select the Crop tool

The image I’m working on is one of these, and I can see the additional image area by swapping to the Crop tool. This is quite a significant additional area we’re talking about here, not just a few pixels around the margin, and if you’re shooting with a wideangle lens, any extra coverage will be very welcome.


Step 03: Recrop to the full image area

Here, I’ve dragged the edges of the crop marquee out to the edges of the image, and you can probably see how this has included new objects and details that weren’t in the original picture. This was taken with the Olympus 12-40mm f/2.8 lens at its widest setting, equivalent to a 24mm lens. I reckon with the extra image area recovered with this re-cropping, it’s closer a 22mm lens’s angle of view.


Step 04: Try it in Lightroom

Just to check that Capture One had really found extra image areas and not just defaulted to a tighter crop, I opened the same picture in Lightroom. The Lightroom images match Capture One’s default crop area (and the crop area of the camera’s JPEG image, captured at the same time). I switched to the Crop tool in Lightroom to see if the crop marquee can be pulled out to capture the extra areas found by Capture One, but in Lightroom they’re simply not there.


Step 05: Export the proof

As a final check I exported the default Capture One crop as a new file. It matched the pixel dimensions and the angle of view of the Lightroom image and the camera JPEG exactly (4,608 x 3,456 pixels). The exported image with the wider Capture One crop has a resolution of 5,027 x 3,658 pixels and clearly shows new image areas outside the angles of view captured by the camera and Lightroom.

So what’s the explanation, and why does this happen with some images and not others?

There is a pattern. Olympus cameras, unless I’m mistaken, embed distortion correction profiles directly in the RAW files, and most RAW converters will decode this information correctly without recourse to lens correction profiles. Apple Aperture would do this with certain camera RAW files, even though it did not have distortion correction tools in the sense that Lightroom or DxO Optics Pro does.

I can only assume that the camera records a larger area of the scene in the RAW file but that this is discarded by most applications (and the camera) as part of the automatic correction. My theory is that, somehow or other, Capture One is able to both access this larger image area and apply distortion correction at the same time. It defaults to the smaller crop area simply to reflect the camera’s original framing.

Some tests seem to confirm my theory. Shots taken at longer focal lengths with this lens show a much smaller recoverable area outside the default crop and, at maximum zoom, none at all. Clearly, the distortion correction (and the recoverable area) is greatest at the shortest focal lengths.


Step 06: Check with another camera

Olympus is not the only maker to hard-wire distortion correction into RAW files. I seem to remember Fuji does the same thing (which is why its zoom lenses always seem so remarkably distortion-free). So I tried this with my Fuji X30, checking results at different focal lengths – and, sure enough, the same thing happens here. This picture was taken at the lens’s minimum zoom setting and you can see Capture One has again cropped in to a smaller area than the image’s full size.

I adjusted the crop to take in the full image area, exported the image and got the same result as I did with the Olympus. The Fuji X30 is a 12-megapixel camera, but Capture One exported a 13.5-megapixel image with a wider angle of view than the camera itself captured or displayed.

This is very exciting! Unfortunately, it only seems to work with RAW files with built-in distortion correction, which only a few makers use, and only then at wider zoom settings when the correction is likely to be greatest.

It didn’t do a thing for any of my D7200 RAW files taken with any of my lenses, alas, but then Nikon doesn’t embed correction data in the same way that Olympus and Fuji do.

7 thoughts on “Can Capture One see more than your camera?

    1. Capture One will apply the lens corrections automatically, but as far as I can see it does this using correction data embedded in the RAW file, not by looking up the lens information in the EXIF data and applying its own correction profile. Not sure if this answers your question, though.

      What’s still unexplained is how Capture One can display (undistorted) image areas outside those displayed by other apps and the camera itself.

  1. Sounds like these cameras are resampling their captures. If E-M1’s actual physical usable pixel count is 18.4 Mpix, then it downsamples images that don’t need much distortion correction to 15.9 Mpix and simply crops the ones that are most distorted (and a combination of both between the two extremes). If the actual pixel count in camera is 15.9 Mpix, then it first upsamples images that need most correction to 18.4 Mpix and then embeds a crop back to 15.9 Mpix into the RAW file. This would be to ensure the default output size is always 15.9 Mpix. Which one would be your conclusion? Is there something else going on?

      1. Yes, that’s an interesting point – thanks. Olympus quotes a gross pixel count of 16.8 megapixels, so Capture One clearly isn’t just finding extra ‘unused’ pixels from the sensor data. As you say, perhaps the camera is upsampling to 18.4Mp in order to get enough headroom for its distortion correction. I’ll try to find out from Olympus.

  2. Hi Rod,

    Thanks for the interesting article. I have the same observations with my Fujifilm X-Pro1 images. The default image area of the RAF’s in Lightroom 6 and Aperture 3 is roughly the same, but in CaptureOne 7 it’s noticeably smaller. The pixels count is the same in all 3 convertors.

    The difference between Lightroom/Aperture on one side and CaptureOne on the other is much bigger with my XF 18mm than with my XF 27mm.

    I quickly checked a few of my Nikon NEF’s and there I see no difference in image area and pixel count.


  3. It happens exactly how you describe with my Leica 109, Very happy to find quite a significant growth in size on the wide images

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.