refraction73_MG_1171

High frequency

refraction74_20100809__ICEjokulsarlon_0148

Medium Frequency

refraction6420080816_iceland_reykjanes_0496

Low Frequency

Frequency is a term that’s being used more and more. That’s because new tools offer you more control over frequency than ever before. Noise reduction, sharpening, and HDR all offer unprecedented control over the look and feel of detail in our images. Frequency is used to describe the amount of detail packed into a given area of an image. This is measured by the amount of tonal variation between rows or columns of pixels. Imagine measuring an image with a line that passes across it (horizontally from left to right or vertically from top to bottom). The mean or average tonal value along lines can be charted and then compared to values from other measurement lines, especially those nearest to each other.

In high frequency images you’ll find a great deal of variation between measurements; many lines, thick and/or thin, and lots of texture, coarse and/or fine, rendered with high contrast.

In medium frequency images you’ll find a modest amount of variation between measurements; clear contours with moderate to low amounts of texture in between.

In low frequency images you’ll find very little variation between measurements; characterized by smooth long gradations, often with reduced contrast or flat fields of color.

Many images contain a combination of high, medium, and low frequencies. When enhancing images you can choose to emphasize the dominant frequency or selectively enhance areas with different frequencies for even greater precision. Some software features provide ways to target these frequencies as you adjust them. When software doesn’t provide ways to target frequency, you can design an effect for that image area on a separate layer in Photoshop and mask it from other areas you don’t want to be treated in the same way.

In many cases, you don’t need to measure an image to decide what tack to take. By looking at an image with a discerning eye you’ll quickly be able to tell if and where an image contains high, medium, and/or low frequencies.

Simply being aware of and sensitive to frequency in images will encourage you to be more precise with your image adjustments. With a little extra care your images will all become stronger.

Read more about sharpening here.

Learn more in my digital photography and digital printing workshops.

Refraction LXIX 2

Of the three stages in a sound sharpening workflow – capture sharpening, creative sharpening, and output sharpening – creative sharpening is the stage that has the most impact.

The goal of creative sharpening is to give an image a specific look and feel. There are at least three things creative sharpening can do for your images. One, creative sharpening can prioritize; it can direct attention to specific areas of an image. Two, creative sharpening can enhance; qualitative aspects of images like texture and line, can be amplified to produce stronger responses. Three, creative sharpening can be used to accentuate different qualities of light; a great deal of detail is carried by the luminosity component of color and changing it changes the overall appearance of light within the image. Used consistently creative sharpening can produce a distinctive style that is more easily recognizable to viewers. (Remember, sharpening is a way to enhance details and it may also be used with its counterpart blurring to make effects appear even stronger by comparison.)

The decisions made during creative sharpening are largely subjective and based on the visual preferences of the individual doing the sharpening, not the characteristics of the tools used to produce an image. When performing creative sharpening there are essentially no rules. Only the image source, the software you choose to use, and, most importantly, your aesthetic sensibilities will determine the limits of how far you can go. If there are limits to how far you should go, those limits are only determined by consensus; in general most viewers don’t want to be distracted unnecessarily by sharpening artifacts – unless they are an integral part of the statement an image makes. During creative sharpening you can leverage any and all sharpening techniques. Creative sharpening can be as simple or as sophisticated as you choose. Increasing your sharpening skills will lead to enhancing your expression. The final determining factor during creative sharpening is that it creates a desired visual appearance.

Creative sharpening needs to be determined visually, while previewing an image on a monitor at 100% screen magnification, the magnification that most precisely displays low frequency detail such as texture and noise.

Creative sharpening is done after Raw conversion that includes conservative capture sharpening, typically in Photoshop, employing additional image layers, with masks. Creative sharpening is most frequently applied selectively, varying the amount and/or type (Clarity, High Pass, Unsharp Mask, Smart Sharpen, etc) of sharpening in different regions of an image.

While there are no standard formulas for creative sharpening that will apply equally well for all images and more importantly all users, this doesn’t mean you can’t automate many parts of creative sharpening, once you’ve created a sharpening recipe that pleases you. Recording your preferred sharpening routines as actions can speed up this process. It does mean that to get it really right you’ll want to modify the results of your recipes based on the characteristics individual images you’re processing – sometimes subtly and sometimes substantially.

Creative sharpening may need to be removed and reapplied if an image file is dramatically upsampled, as the resampling process can make sharpening artifacts not visible at smaller scales more pronounced at larger scales and in some cases exaggerate them.

Not all images need to be creatively sharpened. In high productivity workflows, where large volumes of images need to be processed quickly, creative sharpening is typically abandoned because it can’t be fully automated. But ‘hero’ images are another matter entirely. The images you care most about deserve creative sharpening – and for these it can make all the difference in the world.

Follow up with How To Avoid Common Over-Sharpening Artifacts.

Read more about sharpening here.

Learn more in my digital photography and digital printing workshops.

1_Sharpen_No

Before sharpening

2_Double_USM

Unsharp Mask only

3_Hybrid_USMHighPass

Unsharp Mask and High Pass filters combined

Different sharpening techniques make the world look different. A world of difference can be seen between the thin hard line of Unsharp Mask and the broad feathered line of High Pass Sharpening.

Can you choose a combination of both? Yes you can! You can choose the texture of one, the halo of another, and the line of yet another, applying them either globally or selectively. You can customize the look and feel of detail in any image or image area with astonishing precision and flexibility.

2b_Sharpen_DoublePass

Multi-pass sharpening combines multiple applications of sharpening on one layer.

Multi-Pass Sharpening

Results will differ if you filter the same image layer twice. Why? First, either the technique or the settings can be varied. Second, having been filtered once, the state of the pixels will have changed before a second pass is applied, generating a different final effect. Consequently, not only the type and amount of filtration matters, but also the order in which the filtration is applied.

One classic multi-pass sharpening technique involves filtering first with an Unsharp Mask setting using a low Radius (under 1.0) and a high Amount (300% or more) and second with an Unsharp Mask setting using a high Radius (approximately1.5) at an Amount of 100%. A variant of this technique adds a third pass of High Pass sharpening. Both methods use the first pass of Unsharp Mask to give the second pass of filtration more to bite into. The key to making any multi-pass sharpening technique successful is to produce a strong yet still convincing effect with as few, if any, unwanted artifacts as possible, either with or without masking.

Some routines will repeat filtration at a lower amount multiple times; for instance, a sharpening setting may be applied ten times at ten percent instead of one time at one hundred percent. The idea behind this approach is that you can achieve a more intense effect (crisper edges) with fewer artifacts (accentuated noise/texture). As it’s inefficient to perform these routines by hand more than one time, this type of approach is best handled by recording an Action that you can play for future uses, which may need to be modified if resolution varies substantially.

Are there benefits to filtering more than twice on the same layer? Maybe. Maybe not. You get diminishing returns with each additional pass of filtration. You may also run the risk of producing more unintended artifacts. Furthermore, as complexity rises your ability to both predict and interact with the final effect diminishes. In general, I recommend you be cautious of highly complex routines and urge you to ask yourself if you derive significant benefit from them.

3b_Sharpen_Hybrid

Hybrid sharpening combines different sharpening effects using separate layers.

Hybrid Sharpening

Sharpening results will also differ if you apply varied filtration techniques to separate layers. Here, the order of the layers in the layer stack matters.

To combine the effects of the different layers use blend modes: Darken will display the only values on a sharpening layer that are darker than values on layers below it, such as the dark line; Lighten will display the only values on a sharpening layer that are lighter than values on layers below it, such as the halo; Luminosity will display any values that change in brightness, but not hue or saturation, and may override any sharpening effects below them so consider separating one Luminosity layer into two layers, one on Lighten and the other on Darken, as their cumulative effect will enhance rather than override underlying effects.

High Pass sharpening layers (or any technique that reduces an image layer largely to gray values) combine easily with other layers using blend modes (typically Overlay); they do this so well that many times it doesn’t matter whether they are placed above or below other sharpening layers.

To reduce file size, you may decide to merge multiple sharpening layers into a single layer. While this makes a file easier to manage now, it reduces your ability to modify the sharpening effect in the future and to clearly track any effects or artifacts were produced. Weigh the pros and cons of this option carefully.

Selective Sharpening

By keeping sharpening effects on separate layers you not only preserve the future flexibility of the effects you create but you are also able to selectively control the effects and target specific areas of an image more precisely. There are three primary ways of doing this; blend modes; Blend If sliders; and masks. A layer’s blend mode controls the way its values combine with values in layers below it; access a layer’s blend mode at the top of the layer stack. A layer’s Blend If sliders let you quickly remove effects from highlights and/or shadow. Activate a layer’s Blend If sliders by double clicking on it – split the sliders for smoother transitions. A layer mask allows you to target different areas of an image. Add a mask to any layer by clicking the mask icon in the Layers palette and fill (either with selection or brush) areas you want to reduce an effect in with varying shades of gray, darker values reduce effects more.

When you combine different sharpening techniques you’ll find that when it comes to the appearance of detail you’ll have a wider variety of choices to choose from. This can affect more than just the look and feel of your images. You can also use it to guide the eye to specific image areas in different ways, producing a qualitatively different visual journey. Sharpening can make the world looks different. Master sharpening and you may even see the world differently. People who view your images certainly will.

Read more on image sharpening here.

Learn more in my digital photography and digital printing workshops.

1_sharp_refraction_LXXIII1 copy

2_sharpenedlayerstack copy

Layers have Blend Modes and can be masked

3_layerstyle copy

Double click a layer to activate its Blend If sliders  

There are many reasons to use layers when sharpening your digital images.

How do you do this? Simply duplicate the Background layer and sharpen the new layer.

Eliminate Saturation Shifts

Layers can be used to eliminate saturation shifts. Change the Blend Mode of a sharpening layer from Normal to Luminosity. Color noise will also be reduced this way.

Prevent Clipping

Layers can be used to prevent clipping in deep shadow detail (near-black) and bright highlight detail (near-white). As sharpening is a contrast effect, near-white and near-black values can be driven to pure white and pure black by it. There’s a cure. Double-click the layer to activate Layer Styles. Use the Blend If sliders to reveal the lost highlight and shadow detail in the background layer below the sharpening layer; zoom way into a highlight area, hold the Option/Alt key and drag the right arrow to restore highlights and the left arrow to restore shadows.

Precise Local Adjustment

Layers can be masked for greater control over confined areas in an image. To begin, add a layer mask. Select an area from which you wish to remove a sharpening effect, like a sky or other area of even tone, and fill the area with black. You can use this strategy to remove unwanted texture or noise from selected areas of an image. Gray values can be created on a mask with the Gradient tool or with a Brush tool to gradually reduce a sharpening effect. This often can produce a more strongly felt impression of space within an image. In anticipation of selectively modifying an effect, you may decide to sharpen an image more aggressively.

One approach to gaining additional flexibility with sharpening effects is to set a sharpening layer to 50% Opacity before applying the filter and then later adjust the opacity up or down to get more or less of the effect. This can be useful, but be mindful of its limitations. Reducing or increasing a sharpening layer’s opacity will provide an effect similar to adjusting Amount; more or less contrast is added. But modifying opacity can’t simulate the effects of different Radius settings—thicker or thinner contours.

Combine Multiple Types Of Sharpening

Use more than one duplicate layer and you’ll be able to combine multiple types of sharpening by simply reducing the top layer’s opacity. Unsharp Mask, High Pass, and Clarity all produce different effects that can be combined into still new effects. With these tools, you can craft a unique look and feel for detail in your images.

In addition to the flexibility of changing and/or removing and remaking sharpening effects layers’ features Blend modes, Blend If sliders, Opacity, and Layer Masks offer extraordinary control and precision. When you want to get sharpening effects really right, use layers.

Read more on sharpening here.

Learn more in my digital photography and digital printing workshops.

Quotes_Looking

Enjoy this collection of quotes on Looking.

“All of us are watchers – of television, of time clocks, of traffic on the freeway – but few are observers. Everyone is looking, not many are seeing.” – Peter M. Leschak

“We do a lot of looking: we look through lenses, telescopes, television tubes… Our looking is perfected every day, but we see less and less.” – Frederick Franck

“Seeing, in the finest and broadest sense, means using your senses, your intellect, and your emotions. It means encountering your subject matter with your whole being. It means looking beyond the labels of things and discovering the remarkable world around you.” – Freeman Patterson

“Joy in looking and comprehending is nature’s most beautiful gift.” – Albert Einstein

“All of the top achievers I know are life-long learners… Looking for new skills, insights, and ideas. If they’re not learning, they’re not growing … not moving toward excellence.” – Denis Waitley

“Thinking is more interesting than knowing, but less interesting than looking” – Johann Wolfgang von Goethe

“We learn more by looking for the answer to a question and not finding it than we do from learning the answer itself.” – Lloyd Alexander

“Poetry surrounds us everywhere, but putting it on paper is, alas, not so easy as looking at it.” – Vincent van Gogh

Read more

Reflection XII Adagio

high noise / low noise

noise_banding

column and row noise / severe under-exposure

noise_fixedpattern

hot pixel noise / very long exposure or very hot conditions

noise_bayer

Bayer pattern noise / substantial under-exposure

noise_random

random noise / high ISO

While it’s best to eliminate noise in images at the point of capture (by choosing optimum tools and making exposures), taking the steps to do this may be impractical and/or lead to unacceptable trade-offs, so you may need to make a compromise settling for reducing it (first during exposure and second during post-processing). But which compromises should you make? Knowing the types of noise that are produced in digital images and how they are produced will help guide you to solutions that will eliminate, reduce, or remove it.

There are three types of noise; random noise, fixed pattern noise, and banding noise.

Random Noise

Random noise appears as both luminance (light and dark) and chrominance (hue/saturation) variations not native to an image but produced by the electrical operation of a capture device. The electrical signal produced in response to photons is comingled with electrical variations in the operation of the capture device. Random noise patterns always change, even if exposure conditions are identical. Random noise is most sensitive to ISO setting. Again, digital cameras have one native ISO setting; higher ISO settings artificially boost the signal produced by the sensor and the noise accompanying it. The results? You get a brighter picture from less light and exaggerated noise. Since the pattern is random it is challenging to separate the noise from the image, especially texture, and even the best software used to reduce it through blurring may compromise image sharpness; how much depends on the level of reduction.

Fixed Pattern Noise

Fixed pattern noise (“hot pixels”) is a consistent pattern specific to an individual sensor. Fixed pattern noise becomes more pronounced with longer exposures. Higher temperatures also intensify it. Since the pattern is consistent, it can be easily mapped and reduced or eliminated.

Column & Row Noise

Banded noise is introduced with the camera reads the data produced by the sensor. It’s camera-dependent. Banding noise is most visible at high ISOs, in shadows, and when an image has been dramatically brightened. This type of noise quickly becomes obvious and objectionable; the regular row and column patterns from the sensor quickly call attention to the capture device; it is challenging to reduce without severely compromising image sharpness.

Noise can be broken down into two kinds; chromatic (hue/saturation variances) and luminance (brightness variances).

noise_chrominance

chrominance / color noise

Chromatic Noise

Chromatic noise produces a more ‘unnatural’ appearance, it is easier to reduce without compromising image sharpness than luminance noise. Chromatic blurring is less noticeable than luminance blurring, as human perception tends to see color contained within contours, even when it is not precisely true. It’s a convenient optical illusion. Larger chromatic variances may result from bayer pattern demosaicing. (Digital sensors typically capture photons with an array of two green, one red, and one blue photosites that register separate luminance values for each site. This data is then processed, ‘averaged’ if you will, to generate a final color, such as brown or lavender, or even a specific green, red, or blue. If done under suboptimal conditions, such as underexposure, larger areas of color variances may occur and will require additional post-processing. Extreme amounts of chromatic noise reduction may results in reduced saturation, especially along contours separating strongly contrasting colors.

noise_luminance

luminance / light & dark noise

Luminance Noise

The presence of luminance noise is more readily accepted than chrominance noise. Luminance noise is harder to reduce than chrominance noise. Luminance information encodes contour, volume, and texture, key elements in representational images. Aggressive amounts of luminance noise reduction subdue image texture, creating a synthetic or overly smooth appearance, and blurs contours, lessening the appearance of focus. Camera noise reduction tends to be crude. Raw conversion software produces significantly improved results. For more extreme noise reduction, third-party software (such as Define, NoiseNinja, and NoiseWare) offer superior functionality and results.

Noise also varies in both magnitude and spatial frequency. Noise occurring over short distances has a high frequency (it’s ‘fine-grained’), while noise occurring over long distances has a low frequency. Noise magnitude, often described by the statistical measure of ‘standard deviation’, quantifies the variance a pixel will have from its ‘true’ value. Higher magnitude noise overpowers fine texture and becomes exceptionally difficult to remove.

The noise floor dominates other forms of noise. It is created by the type of read circuits in the sensor, the transistor characteristics, and support circuits such as the analog to digital converter. As light levels increase the noise associated with light (‘photon shot noise’) exceeds the noise floor. If the signal is increased by a factor of two (one f-stop), then the noise increases by a factor of one and the signal to noise ration increases by one. A higher signal to noise ratio makes noise less visible. When the signal exceeds the maximum value the sensor is capable of capturing (dynamic range is a measure of the largest ratio of the capture signal to the noise floor), the noise drops because the signal is pinned at the saturation value.

With a thorough understanding of what produces noise, how it is produced, what kinds and types of noise to be on the lookout for, you can take steps to reduce it at the point of capture. You want to start with as little noise as possible. If you want noise, you can always add it later, which gives you the possibility of customizing it with almost infinite precision. If you begin with noise in your originals, you’re locked in, and it can be challenging to reduce it without compromising image quality – sharpness, texture, saturation, and hue variety. Given that noise isn’t the only concern you balance, for some uses this may be an acceptable trade-off.

.
.

Subscribe

Get the RSS Feed  

Subscribe by Email