10. Techniques

10.1 Speeding-Up of Updating Preview
    In order to speed up the preview mode, this software first displays the outline preview, while executing the formal development in the background. The preview is updated one by one, from the region where the formal development is completed.
    However, when you want to adjust development parameters after checking the formal development result, please wait until the formal development preview is displayed.

    Here, we describe techniques for reducing time until the formal development preview mode is described.

    (1) Minimizing a window
      Minimize a window to display a portion of a picture. The region where the formal development is executed becomes smaller.
      Since this software is designed to develop and update only a displayed area in the preview window, when the window is minimized and only the focused portion is displayed, a time until the formal development preview mode completes can be reduced.

    (2) Increasing display magnification
      Increase a display magnification to decrease a region of a displayed picture. This has the same effect as the case of (1). We recommend this technique since a focused portion is enlarged and displayed.

    (3) Setting a parameter of false color control to 0
    (4) Setting the "Dev. precision" slider on the "Development" tab to the "Fast" side
      The development logic (SILKYPIX® 1.6 Development Engine), which is used in this software, can change the development precision continuously. In most cases, you can obtain better results by setting the "Dev. precision" slider to the "Quality" side. However, this means a heavy workload (Refer to '10.4 Relation between Development Precision and Picture Quality').
      For this reason, as similar to the case of false color control, set the slider to the "Fast" side while adjusting other parameters to reduce a time until the formal development preview mode completes.

10.2 Difference between Exposure Bias with Camera or at Development, and Application
    When executing the exposure bias on a camera, the light quantity given to an image sensor is changed.
    For example, when executing +1EV exposure bias, the light quantity given to an image sensor is double.
    On the other hand, in the case of exposure bias at the time of the development processing, the light quantity which is recorded in a RAW file increases up to double amount.
    The effect is the same as it on a camera, however, in the case of the exposure bias at development, a picture becomes rough because a noise component also becomes double.
    On the contrary, the exposure bias at development also has a merit. Consider the case that if executing the exposure bias on a camera. If it is too much, a highlight has gone too much.
    Regarding the portion that has gone, because of exceeding the image sensor's limit, even a RAW file does not contain information of that portion. It is not possible to save the portion even by executing the desensitizing at development.
    In some cameras, it is possible to execute the desensitizing about 1/2EV at development and to save the portion that has gone. Refer to '12.1 Desensitizing at Development.' However, note that usually it is not possible.
    In such a case, sensitizing is advantageous at the time of the development processing.
    Because there is no need to determine the exposure when being busy with taking a picture. You can determine later at development.
    For this reason, it is often said, "RAW should be photographed with underexposure." Refer to '11.2 Should RAW be photographed with Underexposure?" for details.
    When you are not able to determine the exposure bias amount because of backlight or other difficult conditions to take a picture or you are about to miss a photo opportunity because of taking too much time to determine the exposure bias amount, we recommend to take a picture by setting the exposure bias in underexposure as a highlight portion will not be gone. Refer to '11.1 Photographing without Backlight Compensation.'

    Furthermore, as an application of functions, it is also thinkable of a very bold filming method to gain shutter speed and an aperture value by intentionally setting underexposure toward the dark subject that an ISO film speed must be increased.
    Refer to '10.3 Difference between ISO Sensitivity Adjustment with Camera and Exposure Bias at Development and Application.'

10.3 Difference between ISO Sensitivity Adjustment with Camera and Exposure Bias at Development, and Application
    Normally ISO sensitivity adjustment with cameras is performed before digitizing the voltage outputted from an image sensor.
    For example, if the ISO sensitivity is double, information from an image sensor is also doubled then sampled.
    Or it is doubled after being sampled.
    For this reason, as being different from a case of executing the exposure bias at the time of the development processing, its result is almost the same as a case of sensitizing at the time of the development processing.
    Therefore, the result of taking a picture with ISO 400 is same as a case of taking a picture with ISO 200 that is one stop underexposure.
    There is only a minor difference between the information from an image sensor is doubled in a camera and recorded or the information is doubled at the time of the development processing.
    Therefore, in this case, the merit of sensitizing at the time of the development processing is much.
    When taking a picture with ISO 400, if a highlighted portion is gone, it is not correctable, however, with ISO 200 that is one stop underexposure, the possibility for correction becomes higher.

    So it is possible to execute the very bold filming method to gain shutter speed and an aperture value by intentionally setting underexposure toward the dark subject which the ISO sensitivity must be increased. See '11.3 Turning Exposure Bias Dial into ISO Sensitivity Dial.'

10.4 Relation between Development Precision and Picture Quality
    In the de-mosaic algorithm which is installed in this software, you can change resolution precision at the time of de-mosaic processing continuously.
    This is called development precision.
    If you increase the development precision, resolution becomes minute and a developing time gets longer.
    If you decrease the development precision, minute of resolution is lost, however, a developing time gets shorter.
    Therefore, in order to obtain good quality, we recommend you to set the "Dev. precision" to a certain large value (about 80).
    It is almost correct to think that if you set the "Dev. precision" higher and higher, a better and better quality picture can be created. If a picture contains a lot of noise, there is a possibility that a result will be just highlighting a noise.
    In this case, you should set the "Dev. precision" lower.

10.5 Applying Auto Exposure Bias and Auto White Balance to Multiple Scenes
    Applying Auto Exposure Bias and Auto White Balance to Multiple Scenes
      You can select any scene from multiple scenes and set its exposure and white balance to "Auto."
      Right after you take pictures and load them into a PC, in order to apply auto exposure and auto white balance, first you can set exposure and white balance of the first scene to "Auto" and then paste its development parameters on all other scenes. This is very convenient.
      The procedure is shown below.
      1. Open folder to which RAW files are loaded.
      2. Set exposure and white balance to "Auto."
        Select "Auto" in the exposure combo box.
        Select "Auto" in the WB combo box.

      3. Copy its development parameters.
        [Edit(E)] - [Copy development parameter(C)]

      4. Select all scenes.
        [Edit(E)] - [Select all scene]

      5. Paste development parameters.
        [Edit(E)] - [Paste development parameter(P)]

10.6 Modification of Overexposure
    When taking a picture of a subject with saturation, especially flowers etc., there is a case that a color of a bright portion turns white.
    In this manual, this phenomenon is called "Overexposure."
    Here, a method for modifying this overexposure is described.

    • Reason for Overexposure
      In the first place, why does the overexposure occur?

      When shedding light on the subject that has smooth surface, the subject looks white because of light reflection. This is not the overexposure.
      Using overexposure, the subject that looks red with human eyes appears white because of its bright portion.
      This occurs when its brightness exceeds the intensity level of brightness that can be represented at the time of taking a picture or the development processing.
      In a case of picture (not only a picture, but also printing, printer, monitor, and TV), there is a limit for presentable brightness.
      This limit becomes smaller as a color becomes deeper.
      Here this phenomenon is explained with R, G, and B. On PC monitors, colors are created by emitting light from tiny points of R, G, and B.
      Now, it is assumed that the most darkest state of a red point is expressed as R=0 and the most brightest state is expressed as R=255. Regarding G and B, assumption is the same.
      When R=0, a red point emits the most darkest light and when R=255, it emits the most brightest light on a monitor. Color and brightness are represented by a balance of how light emits.
      The brightness that a man can sense is a mixture of emitted brightness of R, G, and B. (*1)

      For example, in a case of white color, R, G, and B have the same emitted brightness. In other words, when R=100, G=100, and B=100, the color looks white.
      This means white color (achromatic color) has a representation range from R=0, G=0, B=0 to R=255, G=255, B=255.
      By converting to brightness, the representation range is from 0 to 255+255+255=765.
      Next, let's consider light red color (pink). In a case of light red, R has a larger value than that of G and B. For example, R=200, G=100, and B=100.
      At this time, the values of G and B are a half of the R value.
      When this ratio is maintained, the color looks the same.
      For example, in a case of darkened light red, the values are R=100, G=50 and B=50.
      When it makes brighter, the values are R=200, G=100, and B=100.
      When it makes more brighter, for example, in a case of the most brightest light red, the values are R=254, G=127, and B=127. (*2)

      From this condition, if adjusting the color much more brighter, what will happen?
      On paper, the following values can be created: R=300, G=150, and B=150
      However, since R=300 cannot be represented in reality, the actual value of brightness is R=255.
      So the three values are R=255, G=150 and B=150.
      What about adjusting the color much more brighter?
      The values R=510, G=255 and B=255 become R=255, G=255 and B=255.
      This is pure white color.
      Roughly speaking, the overexposure occurs in this way.
      In the case of this light red color, the limit values that the color does not change are R=254, G=127 and B=127.
      And the brightness that human eyes can sense is 254+127+127=508.
      This means that although the brightness of white can be represented up to 765, the brightness of light red can be represented only up to 508.

      How about a case of deeper red color?
      For example, let's consider a case of deeper red with values as R=200, G=50 and B=50.
      The values of the most brightest deeper red that can be represented are R=252, G=63 and B=63.
      This means that its brightness is 252+63+63=378.

      Therefore, when a color is deeper and deeper, brighter representation cannot be possible.
      So when an exposure level is adjusted to a light color subject, the overexposure occurs at a deep color subject.
      This is the reason for the overexposure occurrence.

    • Modification of Overexposure 1(exposure bias)
      The Modification of Overexposure 1 is exposure adjustment. Since the overexposure is caused by adjusting the exposure level to a brighter color, you can modify the overexposure by adjusting a exposure level to the certain level with which the overexposure does not occur.
      However, in many cases, a light color subject looks too dark.

    • Modification of Overexposure 2 (saturation adjustment)
      The Modification of Overexposure 2 is saturation adjustment. When color saturation is lower, a brighter representation is possible. So by lowering the color saturation, you can modify the overexposure.
      However, in many cases, when lowering color saturation as fitting to overexposure with high saturation, a very light color picture is created because saturation at the other portions becomes too low.

    • Modification of Overexposure 3 (highlight controller)
      This software has a function to control the method for clipping the color of the region where brightness exceeds any representation limit of R, G, and B in a range.
      For example, let's consider the former example of the brightness with values of R=300, G=150, and B=150.
      If this software is running, wider range of data than the actual presentable range are processed. For this reason, this software can keep the real color values such as R=300, G=150 and B=150.
      However, when the values are outputted, they are clipped in the range of 255 for each R, G, and B.
      If only the R=255 is clipped simply, hue, saturation and brightness (lightness)  are out of balance.
      Its original brightness is 300+150+150=600.
      When the highlight controller is set to give brightness highest priority, the color clipping is executed with priority of brightness 600 instead of color.
      For example, the values are R=255, G=173 and B=172 (*3), and brightness can be maintained as 600.
      However, this increases overexposure level.
      When the highlight controller is set to give brightness lowest priority (give saturation highest priority),  the color clipping is executed with priority of color.
      Here we omit the explanation with numerical values in the case of priority of color. Since in the case of priority of color,  the default brightness is 64, you can remember that modification of overexposure is possible by setting the value from 64 to 0.
      In addition, in this case, you can also determine which one to select, priority of hue or priority of saturation. For details, refer to '4.10 Highlight Controller.'

    • Modification of Overexposure 4 (fine color controller)
      By using the highlight controller, you can control an overexposure region, however, this does not have sufficient effect to save a high saturated color completely.
      Rather, as remaining a highlighted portion, the function is very effective when you determine its representation.
      The fine color controller, function can provide more effective method.
      The method is to lower both saturation and lightness of the color with which the overexposure has occurred.
      By adjusting only the color with which the overexposure has occurred, it is possible to execute the modification of over exposure without giving any effect to the white portion which has low saturation. Refer to '4.11 Fine Color Controller' for the use of the function.

    *1 ... In reality, sensitivity of human eye differs according to a color. Although a color of R+G+B does not have the lightness which a human eye can sense, we simplified the matter for the sake of explanation.
    *2 ... Since the non-linear characteristics which is called γ (gamma) characteristics is applied to the actual RGB data (for example, sRGB data), the reality is not so simple as such. For the sake of explanation, on the assumption that the RGB value is a linear value and the values of R, G, and B are in proportion to the light quantity which is emitted from a monitor, the explanation is given.
    *3 ... Actual operation of the software is more complicated. It is because that the actual lightness is not as R+G+B (*1), and the γ (gamma) characteristics is applied (*2). In order to help users understand the functional operations, we just simplified the actual matters.

10.7 Creating Color Gamut and Adjustment Outside of the Color Gamut

    There are some cases that exceed the presentable color range. For example, a case of a color with high saturation, increasing  saturation of a color and so forth.
    By executing '6.1 Highlight/Shadow/Outside of the Color Gamut Warnings', it is possible to give a warning regarding the portion which exceeds the presentable color gamut.
    This software can handle two colorspaces such as sRGB and adobeRGB.
    In these spaces, all existing colors cannot be displayed.
    On the other hand, when taking a picture with a RAW file, a camera can capture the colors accurately that are in the color gamut which is much wider than that of the case that these colorspaces can handle.
    On this account, in a case of a flower etc. with high saturation, colors may exceed the color gamut (presentable color range ) that these colorspaces can handle.
    This software can handle the color gamut completely which a camera can capture and also this software can clip the color which exceeds the range at the final stage of the development processing.
    The warning function for out of color gamut gives a warning to a portion with a clipped color.
    Since the colors that are out of the color gamut are clipped and pressed into the color gamut of sRGB and adobeRGB, details are pressed as if they are like being crashed.
    Even as the overexposure does not occur, if you feel that details of the highly saturated portion are lost, you need to check whether colors are out of the color gamut or not.
    Even if the colors that are out of the color gamut become darkened by executing the exposure bias, a warning is still given, not like the case of the over exposure.
    Speaking with the values of R, G, and B, it is a highly saturated portion as if one or even two among R, G, and B takes a negative value.
    For example, although a color with values as R=255, G=0 and B=0 is the most saturated red that can be represented in this RGB colorspace, a camera has captured red color with much higher saturation. When it is developed, it may have values like R=255, G=-20, and B=-20. Of course, such a color cannot be outputted, and it is clipped. So details with high saturation are looked like being crashed.
    For example,
    even if an actual flower has subtle details of saturation between R=255, G=-20, B=-20 and R=255, G=-30, B=-30, its development result has exactly the same colors.
    However, since this software keeps the colors that are out of the color gamut, the details can be restored by lowering their saturation and developing them.
    You can try to lower the color saturation until the warning disappears by checking it on a monitor or by using the fine color controller.

    The colors in such a range are difficult for printing devices such as a printer and so forth. Especially, bright magenta and blue-red colors between red and blue are the most difficult colors for printing devices.
    In addition, a monitor for a PC can display bright colors with high saturation, however, these colors are difficult for a printer and other printing devices.
    A monitor is a light-emitting device but printed materials display colors by absorbing light.
    For this reason, when presenting colors with high saturation, the colors are darkened.
    As a result of printing, if the colors are pressed as if they are like being crashed, lowering lightness might be effective.

10.8 Sharpening Image

    A picture is not sharp. It looks like anyhow lacking vigor and sleepy.
    If you receive such impression, you can try to adjust parameters in reference with the following information.
    1. Exposure Bias
      In a case of developing a RAW file, you can discard the wide range brightness information which a camera captured and determine only the range in which a necessary portion is represented as a picture.
      In a case that a RAW file was not used, only the exposure at the time of taking a picture was used to represent the light which entered into a camera as a picture.
      However, in a case of a RAW file, these can be determined during a developing stage after taking a picture.
      Like the trimming operation which cuts out unnecessary portion, the executing the exposure bias means trimming the range of lightness which you want to represent in the direction to lightness.
      You may adjust the exposure bias until the portion of a picture which you want to represent reaches to the optimal lightness.
      We hope that your picture can restore its lightness and vigor.

    2. Tone Representation
      By using the tone representation, you can determine how to compress or expand the lightness which a camera captures.
      First, let's consider contrast.
      The operation which adjusts white portion more white and black portion more black narrows the range in which you want to represent The contrast operation is to expand the narrowed range to the device presentable range. A picture might have vigor by this operation.

      Now, where is a boundary for determining contrast emphasis? That is contrast center.
      If your picture is bright, you can move the contrast center up and make contrast around the bright portion. If it is dark, you can move the contrast center down and make contrast around the dark portion.
      How is it?
      Hopefully the problem is improved.

      If flare appears on a picture or if you receive impression that it is dusty, you can try to increase a black level.
      Do you feel that a picture is tightened?
      If you feel that a picture seems to be a sleepy because of backlight, a distant view and so forth, you can tighten black color by increasing the black level.

    3. Sharpness
      If a picture has been adjusted up to this level, perhaps the picture is cool now.
      As the rest, focus is almost ...
      From now on, let's adjust sharpness of outline by enlarging a focused area in a picture (more than 100% display magnification).
      First of all, let's maximize the sharpness.
      The outline of a picture becomes clear, and you might have well-shaped impression.
      However, if you look at it very carefully, you might discover that a noise was also increased simultaneously and the outline which was originally very clear seemed to be unnatural because it was too much emphasized.
      In order to delete this unnatural impression, you should reduce the sharpness. Now adjustment is over.
      Refer to '10.9 Creating Extremely Clear Image' for improving the sharpness setting.

10.9 Creating Extremely Clear Image

    In order to create a picture with high resolution, you can adjust sharpness. We call this sharpness control.
    Simply saying, if you select a strong setting in the "Sharp" combo box or increase "Outline emphasis" in the "Sharp" tab, the outline looks very clear.
    However, a noise component is also increased simultaneously.

    When executing the sharpness control, this software can execute a noise reduction process simultaneously. This is called noise canceller.
    There is a reciprocal relation between these two.

    1. When sharpness (outline emphasis, detail emphasis) is increased, a noise is also increased.
    2. When the noise canceller is enhanced, resolution is lost.

    Besides this function, this software has the noise reduction adjusting function which can cancel a noise in a positive way on the RAW level (There is a "Noise reduction" slider in the "NR" tab).
    In addition, there are two functions called false color control and development precision. These two can be operated simultaneously. However, these are not related to reduction directly.

    Here, we describe how to optimize these parameters.
    By executing the following procedure, you can obtain an optimal combination of parameters in a relatively easy way.

    0. Before executing the procedure shown below, you have to adjust white balance, exposure bias, tone and color in advance.

    1. First, turn off the sharpness and the noise canceller.
    2. Move the "Noise reduction" slider as you check details of a picture.
      By adjusting the "Noise reduction" slider, high frequency noise decreases, however, the details are easy to be pressed.
      If you press the details here, you are not able to restore them later when you adjust the sharpness. Therefore, you must adjust the "Noise reduction" slider properly as the details are not pressed.
      In addition, when you adjust by increasing the "Noise reduction" slider, on the contrary, low frequency color noise appears conspicuously. You have to adjust the false color control simultaneously.
      The adjustment by moving the "Noise reduction" slider must be completed, before you adjust the development precision.

    3. Adjust the sharpness and the noise canceller until you obtain the optimal resolution.
      When you adjust the noise canceller, you must maximize the canceller intensity first. Then you can adjust a noise level. Next, you can make fine adjustments with the canceller intensity. This adjusting order makes you feel comfortable.
      While increasing the sharpness, if a false outline is noticeable around the outline, you can increase the false outline control.

    There is a reason for this effective order.
    White balance, exposure bias, tone and color are parameters that are associated with the color direction. On the other hand, development precision, false color control, noise reduction, sharpness and noise canceller are parameters that are associated with the resolution direction (frequency aspect).
    For this reason, first, you have to determine the basic parameters in the color direction. Then you can execute adjustment according to the picture processing order.
    Regarding the processing in the resolution direction (frequency aspect), it is executed in the following order.

    RAW --> Noise Reduction --> De-mosaic (development precision, false color control) --> Sharpness (outline emphasis, detail emphasis, false outline control), Noise canceller (noise level, canceller intensity)

    In other words, determining the parameters in the color direction (0), turning off the processes in the later stage (1), according to the processing order, adjusting noise reduction, false color control (2), finally optimizing sharpness and noise canceller (3).

Copyright©2004-2006, Ichikawa Soft Laboratory. All rights reserved.