Capture One

Latest Posts

Fuji X-Trans Sensor Excels in Capture One 7

For more than 10 years Capture One has been supporting various RAW files from major camera manufactures. These fully supported RAW files all come from cameras using the Bayer Color filter layout.

Over the years Phase One has constantly been working on improving the algorithms for converting the RAW data captured by the sensor to a full RGB file, a process often referred to as Bayer interpolation.

The quality that can be derived from the Bayer filter layout has been well proven throughout the years. Using the Capture One 7 processing engine it is possible to achieve an amazing level of detail and precise color from RAW files using this color filter layout.

The arrival of the X-Trans filter

In 2012 Fujifilm introduced a completely new Color filter layout named X-Trans. Fujifilm’s new color filter layout is based on a 6×6 pixel pseudo random color filter pattern (as opposed to the traditional 2×2). Despite a more challenging filter layout the engineers at Phase One have managed to make support for this new filter layout in Capture One 7.1, yielding natural colors and excellent detail rendition.

A camera with an image sensor that uses the Bayer Color filter layout can run the risk of producing moiré when photographing fine structures. Image content with specific types of fabric or distant building details are potentially problematic. Optical low pass filters, also called antialiasing filters, can be used to minimize the risk of moiré by slightly blurring the image, however this can result in an image with less sharpness.

In 2012, Fujifilm offered their own solution to deal with the above dilemma by introducing a new color filter layout named X-Trans. The theory behind the new layout was that by using a more random color filter, over a 6×6 pixel block, you effectively minimize moiré and can then remove the need for an antialiasing filter. Without antialiasing filter a sharper image can be achieved.

The X-Trans color filter layout with a 6×6 pixel pattern

You can, of course, never completely prevent moiré from happening with just a change to the color filters from the classical 2×2 Bayer layout to the new 6×6 X-trans layout.

A big challenge

The new 6×6 X-trans layout is a big challenge for all RAW converters. Many algorithms and process steps in the Capture One processing pipeline have been modified in order to handle this new pattern. As a result, Capture One now has a solution with which you can achieve excellent details and precise colors using these X-trans cameras.

JPEG file from camera RAW file in Capture One

The image above is a 200% zoom into a X-Trans file from a Fujifilm X100S. The lack of antialiasing filter coupled with a sharp lens results in a very crisp image. If you compare the JPEG with the RAW file from Capture One you will see that, by default, Capture One extracts slightly more details.

More precise color rendition

The real benefit, however is a much more precise color rendition of the smaller details in the image. Look specifically at the colors of the curtains in the windows. In the RAW file you can see green and blue curtains while the JPEG file only shows gray curtains. Also notice the natural color variations in the bricks of the RAW file.

When using a sensor with a color filter array, moiré will occur when patterns in the image interfere with the pixel resolution. This is also the case when using an X-Trans sensor.

JPEG out of the camera RAW file in Capture One.

The image above is a 100% zoom into a picture of two blue shirts. This specific type of shirt fabric has always created moiré problems when testing cameras and it is no different using a camera with an X-Trans color sensor. When no antialiasing filter is used and you have a super sharp lens you can always find a distance (from sensor to subject) where moiré can show up.

Remember that if color moiré occurs you can use the Moiré Tool in Capture One 7 to minimize the color moiré. The Moiré Tool can be applied to the specific area using the Local Adjustments tool, allowing you to apply strong settings adjustments without affecting other parts of the image.

RAW file with moire RAW file after using the Moire Tool

After using the Color Moiré Tool on the RAW file, all color moiré has been removed. Although X-Trans color filter sensors promise moiré-free images, the reality is something else. Luckily, with the advanced Moiré Tool in Capture One 7 it is possible to remove the Moiré completely.

All the best,

Niels

Improve the Level of Detail by Fixing Chromatic Aberration

Almost all lenses show some kind of chromatic aberration, but wide-angle lenses in particular are prone to this issue.

Chromatic aberration is a type of distortion where a lens fails to focus all the wavelengths of light at the same point. This is visible on an image in the form of colour fringing, especially at high contrast edges. Chromatic aberration can also reduce the sharpness and detail of an image.

The combination of the sharpest lenses and strong chromatic aberration can be a big challenge for the Bayer interpolation algorithm in a RAW converter.

In Capture One 7.1.3 the Bayer interpolation algorithm has been optimized to better handle images with strong chromatic aberration, thus improving detail rendition and correcting the color shift.

Chromatic aberration Fixed chromatic aberration

The sample image above is zoomed to 300% view. The image has been shot using a sharp lens with strong chromatic aberration. The image on the left is without the removal of chromatic aberration and the image on the right is the same image but after the removal of chromatic aberration in Capture One 7. After fixing the chromatic aberration not only the visual color shift is fixed but also the definition of all the tiny lines and structures are highly improved.

A normal image sensor in a camera only measures one color per pixel. The most common layout of color filters (named Bayer layout) uses 2 green, one blue and one red pixel for every 2×2 pixels. This layout was invented and patented by Bryce Bayer from Eastman Kodak in 1975. The process of estimating the 2 missing colors for every pixel in the sensor is often named Bayer interpolation.

Bayer color filter layout on an image sensor

When strong chromatic aberration occurs, the red, green and blue color information in the image can be misaligned by up to a few pixels. This misalignment can confuse the interpolation algorithm. Imagine you have a thin one pixel wide line in an image and you have strong chromatic aberration, then one set of color pixels will see the image of the thin line where another set of color pixel don’t see anything. This is in essence the challenge when doing the Bayer interpolation of an image taken with strong chromatic aberration.

Chromatic aberration Fixed chromatic aberration

This is a closer zoom that shows the image before and after the removal of chromatic aberration.

In Capture One 7 we have included automatic lens corrections for a large number of lenses. If an image is taken with one of the known lenses, Capture One 7 will automatically compensate for the chromatic aberration from the lens.

In the Lens Tool the Profile shows the automatically selected Lens
Profile. By default it compensates for chromatic aberration.

When using a non supported lens, Capture One 7 still gives you the option to fix the chromatic aberration in the image and therefore also the potential to increase the detail rendition. In the Lens Tool you only need to check the Chromatic Aberration checkbox.

When checking the box, the application will analyze the image to determine the behavior of the chromatic aberration. After the analysis is complete, the image will then be re-interpolated taking into account the information gathered from the analysis.

Setting the checkmark for chromatic aberration initiates an analysis of the image to determine
a proper fix for the chromatic aberration. The result is a more natural looking image with a highly improved level of detail.

The small icon to the right of the chromatic aberration label will become visible after the chromatic aberration analysis has completed. Clicking this allows you to re-analyze the image. If you have copy-pasted lens correction adjustments including the chromatic aberration information from another image, clicking the small icon will also show which other image the information is form.

All the best,

Niels

Speedup Image Processing with OpenCL

NOTE: This article discusses an outdated version of Capture One. To learn more about our latest version, click here.

An option to significantly speedup processing has been added to Capture One since version 6.0. You can take advantage of this speedup option by using graphics card accelerated processing through an interface called OpenCL. The idea is to make it possible to fully utilize the processing capabilities of the graphics hardware to get a faster application.

In Capture One 7 the graphics card is used for accelerated processing not just to the display, but also when processing images to files.

The speed advantage of using OpenCL can be huge. When using one of the newest graphics cards on the market in a Mac Pro computer, you can process a full 80 megapixel raw file in less than 2 seconds.

What is OpenCL?

OpenCL is an open standard defining how to interface and utilize all the processing capabilities within graphics cards (GPU) and multiprocessor CPUs. OpenCL is short for Open Compute Language.

OpenCL essentially enables effective utilization of thousands of compute units within modern graphics cards. For Capture One, only relatively new GPU cards have enough computing power and onboard RAM to be utilized effectively.

In Capture One version 7.1.2 and 7.1.3 the default setup is not to use OpenCL for processing. To enable the processing you need to switch to Auto in the preferences as shown below.

When Auto is selected Capture One will test the capability of the graphics card, and if the card proves to be powerful enough Capture One will use it for processing.

Performance

Utilizing OpenCL makes it possible to accelerate your processing significantly. For an iMac with the integrated graphics card from NVidia (GTX 680MX) you gain more than a factor of 3 times in processing speed as shown in the benchmark below. On the graph, benchmark 100 is the speed when processing using only the CPU and not OpenCL.

From the figure we see both Canon “cr2” and Nikon “nef” files will process more than 300% faster than the CPU, when utilizing the graphics card on a 4 CPU core iMac. An IQ180 raw file will process close to 250% faster.

The benchmark numbers depend on your graphics card, number of CPU cores and speed of your SSD disk. In terms of the interactive speed using one of the recommended graphics cards, it is possible to get real-time feedback (30fps) for almost all slider interaction in Capture One 7.

Initialization of OpenCL

To increase robustness in Capture One 7.1.3, we have added the OpenCL initialization as a separate process. This ensures that future problems during OpenCL initialization will not affect the Capture One application. Further more, this initialization only happens when the graphics hardware changes.

When the preferences for hardware acceleration are set to auto, the application will test the capabilities of the graphics card. OpenCL is not enabled (in auto mode) if one of the following conditions is met for the graphics card:

Less than 1GB of on-board Video RAM
Too few GPU cores
Too low processing improvement compared to the system CPU performance
Mac OS 10.6

Drivers

As of June 2013, four major companies have developed desktop drivers for OpenCL: NVidia, AMD, Apple and Intel. OpenCL is also becoming an interface for other computer architectures going forward.

For Mac OS X 10 the OpenCL drivers are part of the operating system, but for Windows these drivers are installed as part of the display drivers. For your convenience here are the links to the OpenCL enabled drivers for Windows.

The Windows driver requirements

For AMD/ATI graphics: Catalyst Software suite 13.4 or higher – pick: AMD Catalyst™ Software suite.

For NVidia graphics: NVidia Display driver version 320.18 or higher – pick “download”.

Intel Iris Pro Graphics: Version 15.31.9.64.3165 or higher.

Recommended graphics and acceleration cards for “Best Performance”

As of June 25, 2013 using one of these GPU cards will experience the very best performance:

Mac: Windows:

AMD HD7950 AMD HD7970 GHz edition

Great performance is likewise expected for:

Mac: Windows:

NVidia GeForce GTX 680 NVidia GTX Titan

AMD FirePro NVidia 680GTX

NVidia Quadro 5000 NVidia Quadro 6000, Tesla

ddd

Final words

When working with Capture One 7 certain image settings will bypass the use of OpenCL and use the normal processing pipeline: these settings include the use of spot removal tool and non RGB proof profiles.

Apart from these settings most users with capable graphics cards will see a performance increase when working with Capture One 7 and I recommend that you go into your preference and enable it today.

Regards,
Lionel

How to Correct Perspective with the Keystone Tool

Capture One Pro 7 has a remarkable tool for correcting perspective errors. In addition to complete manual control, Capture One Pro 7 gives you a unique and convenient tool that you can use to rectify the keystone error of an image. Keystone corrections in Capture One simulate a true optical keystone correction, similar to what you can achieve with a technical camera. Once the tool is applied in Capture One Pro 7, perspective errors will be corrected perfectly.

Both vertical and horizontal keystone can be corrected. These corrections can be applied at the same time or independently from one another.

Using the Keystone Tool in Capture One

By default, the Keystone Tool is located in the Composition Tool tab.

By dragging the Vertical, Horizontal, Amount and Aspect sliders you can control the keystone effects in an image. While dragging any slider, a grid is automatically placed over the image to help you align horizontal and vertical aspects.

However, for a faster and extremely accurate correction it makes sense to use the Keystone Cursor tools. These can be selected at the bottom of the Keystone Tool, or from the cursor toolbar. When using the Keystone Cursor Tools we not only correct the keystone but we also correct for any horizontal misalignments of the camera.

The Keystone Tool can correct either vertical keystone, horizontal keystone or a combination of both. On our sample image we need to correct only for vertical keystone, but the principle is the same for images requiring horizontal keystone too.

Begin by selecting the appropriate keystone cursor. You will see vertical lines, horizontal lines or a combination of both imposed over the image depending on which cursor you choose. As I am correcting only for vertical keystone in the example below, we see only vertical lines.

Once selected, simply align the two lines against the sides of the object to be corrected. Like this:

Notice that the ‘Apply’ button has appeared in the centre of the screen. Click this to complete the correction.

In a snap, we have an excellent correction of the vertical keystone.

If we examine the Keystone Tool once more, we can see that values have been calculated for the Vertical and Horizontal fields. By default the ‘Amount’ field is always set to 80% when using the Keystone Cursor Tool for vertical corrections as this gives a more natural looking correction than the ‘perfect’ 100% correction. However, you can adjust this amount to your taste.

If you think the correction has changed the appearance of the aspect ratio a little too much, you can use the Aspect slider to edit the effects.

If the correction needs to be quite strong, you may find the option ‘Crop Outside Image’ useful in the Crop Tool. This will allow you to drag the crop beyond the boundaries of the image area, in case you want to use something like Content Aware in Photoshop to add more to the image.

All the best,

Niels