Image editing for beginners

Image editing for beginners

For ease of reading and viewing, click on the button which closes the left hand column. It's on the top right of the page, next to the RSS button.

Introduction

In this article we will discuss what can be done after taking the shot, how to do it and what effects can be obtained. The discussed methods and examples are program independent - there is a multitude of image editing programs out there, from very basic to very sophisticated software like the expensive Adobe Photoshop or the free Gimp software. We will concentrate on the basic editing changes that can be made by novices and that can be accomplished with nearly every editing program you come across. Image editing can be very rewarding because it expands your creative options and it can also improve your images. Throughout the article you will see images that show a particular type of adjustment, often as a combined before/after image. They demonstrate visually the effect of a particular adjustment.

"Purists" may argue that the camera should deliver a correct image all by itself. They need to accept the fact that a in-camera generated image file always has been edited in some way or another - with parameters largely beyond your control. That may turn out quite well, but it also occasionally fails completely. So why not take full control yourself deciding what editing is needed? Editing has always been part of photography. In the past people have been using different development methods, used different chemicals to tone prints, applied selective exposure of parts of the print in the darkroom etc. Software based editing is the logical consequence of storing digital data!

Due to the size restrictions for images on the forum and because some adjustments are quite subtle, differences may not always be clearly visible. A large number of the images therefore contain a hyperlink: if you click on one of those with the mouse a new window is opened, showing a larger image or a sequence of images. You are strongly advised to use this option: the effects of adjustments will be far better visible in a dedicated window. Images containing a hyperlink have a asterisk added to their number (like image #3* ). With some browsers you may need to click again on the image in the new window to get it displayed at actual size.

This article is primarily aimed at relative beginners and novices. That does not mean it can not be of interest to more experienced users. The adjustments discussed here are all rather basic, but that does not make them unsuitable for use by people that may already have quite a bit more knowledge. I therefore suggest to them to glance through the article - you may well pick up a few ideas. People in a hurry can skip the beginning of the article and proceed to section 3 right away.

Even if you are a absolute beginner, it is important to pay some attention to your monitor settings. Most monitors are delivered with rather high contrast and brightness settings. That is fine for general office work, but less so when judging images. Asking to calibrate your monitor would be too much at this stage, but maybe you can check the settings of your monitor here. For more detailed information about setting your monitor and colour management see the article written by MichelvA.


Contents

• 1 Image editing for beginners
  
  • 1.1 Camera settings: RAW and JPG
    
  • 1.2 Editing modes: parametric (non-destructive) and direct (irreversible)
    
  • 1.3 First things first
  
  
• 2 What to change and in which order?
  
  • 2.1 Workflow
    
  • 2.2 Conversion and editing
    
  • 2.3 Histogram
  
  
• 3 Making adjustments
  
  • 3.1 White balance adjustment
    
  • 3.2 Tone adjustments
    
    • 3.2.1 Exposure
      
    • 3.2.2 Contrast
      
    • 3.2.3 Tone control sliders
      
    • 3.2.4 Tone curve
    
    
  • 3.3 Presence adjustments (clarity, vibrancy and saturation)
    
    • 3.3.1 Clarity
      
    • 3.3.2 Vibrancy
      
    • 3.3.1 Saturation]
    
    
  • 3.4 Detail adjustments (sharpening and noise reduction)
    
    • 3.4.1 Sharpening
      
    • 3.4.2 Noise reduction
    
    
  • 3.5 Optical corrections (distortion and chromatic aberration)
    
    • 3.5.1 Lens correction
      
    • 3.5.2 Chromatic aberration
    
    
  • 3.6 Special effects - adding vignetting and grain
    
  • 3.7 Cropping and straightening
  
  
• 4 Local adjustments
  
  • 4.1 Graduated filter
    
  • 4.2 Adjustment brush
    
  • 4.4 Spot removal
  
  
• 5 Some additional remarks
  
• 6 What is the best software for you?

1 Image editing for beginners

Before getting into the details of image editing, we need to discuss a few related subjects that potentially can have a large influence on what you can do when you edit your images: camera settings and available editing modes.

1.1 Camera settings: RAW and JPG
On the camera you can choose the format in which shots taken will be filed on the memory card. Most cameras offer 3 options: RAW, JPG or both at the same time. RAW means in essence a matrix, a array of data keeping track of the luminances measured and the filter mounted per photosite on the sensor. A RAW file is not yet a image file, but sophisticated software is able to construct the image based on the data in the file. A JPG file is a real image file - the RAW data has been converted by some clever mathematics in the camera resulting in a file that can be shown on the camera screen, a computer, a smartphone, copied to social media accounts or printed.

It is important to clearly understand the difference between a RAW file and a JPG file. A RAW file still needs to be converted into a different file format before it can be shown on screen or printed, a JPG file is already a image file. So why bother with RAW and the extra effort needed to convert it to a image file when a JPG direct from the camera can be copied and printed immediately? Simply put: you can extract a much better image quality when converting the RAW file later on with dedicated software, a so-called raw-converter.

There are a few reasons why more sophisticated cameras are capable of storing RAW data. The first is that all the data recorded at the time of capture are stored in that file - available for later use. The second reason is that a JPG file already has been processed by the camera. You have very little control over how that is done and over the adjustments carried out during conversion. Because in-camera conversion needs to be completed before you can take the next shot, speed is prioritised over precision. A third reason to prefer a RAW file over a JPG file is that a JPG file is compressed to reduce file size. Compression of a JPG file comes at a cost: some data is left out and lost forever. You still can edit a JPG file, but you have less information to start with. Editing a JPG file is like driving a sportscar and never getting past third gear - you will miss some of the performance the camera can deliver. And each time you save a JPG after making changes, the file is compressed again - which can result in additional loss of data.

Nowadays the larger JPG files made by the camera can be quite good, but when you want to edit the file, having all the originally recorded data available, gives you a far better chance to obtain the result you want. If up till now you only shot JPG and are quite happy with it, you can continue to do so. If however you want to carry out adjustments like we will discuss further on in this article, I strongly suggest you shoot RAW plus JPG for a while to experience the differences when editing both file types.

1.2 Editing modes: parametric (non-destructive) and direct (irreversible)
Editing software can make adjustments to a image in two ways: parametric and direct. Parametric editing means that the changes you introduce are not directly written into the original file when saving it, but are kept separately as a set of edit instructions to be applied when showing the RAW file on screen in a editing program, when printing the image directly from the editing program or when the file is subsequently saved in the JPG format. A more common name for this type of editing is non-destructive editing. The beauty of this approach is that the original data are not changed and can be used over and over again and no information gets lost.

The way the editing instructions are stored may differ, depending on the editing program. In Lightroom the editing changes are stored preferably in the so-called catalog - in fact a database with editing instructions per file and lots of other data pertaining to that file, like EXIF data, keywords added, file location etc. Other editing programs store the edit instructions in a sidecar file, with the same filename as the original file and with the extension XMP in most cases (Lightroom also can store data in sidecar files if you want to). When using sidecar files you thus end up with two files - one with the original contents that the camera produced and a additional file that contains instructions on what changes you want to apply to the original file when printing in or when showing it on screen. That works well, but you must make sure that when opening the RAW file later, the editing program is able to find the appropriate XMP file containing the instructions. How it works, is shown in this video.

It is also possible to store edit instructions directly into a specific type of RAW file. If you convert the RAW file into a DNG file introducing various changes at the time of conversion, the edit instructions may be stored separately in that DNG file. With some conversion software you can also include the original RAW file into the DNG file, for retrieval of the original data later. Doing so will make the DNG file substantially larger. Although the DNG file format is based on a fully published standard, applications have some options how to use them and may do so differently. Adobe and DXO do handle DNG files quite differently – something to take into consideration if you want to use DNG files. Some of the differences are explained in this video (RAW vs DNG - A Practical Overview of the Differences). More (and more complex) information on parametric editing and file formats can be found in the following links:

http://dpbestflow.org/image-editing/parametric-image-editing
http://dpbestflow.org/file-format/raw-file-formats
http://dpbestflow.org/file-format/file-format-overview
http://www.cambridgeincolour.com/tutorials/RAW-file-format.htm

When doing direct editing, changes made are saved directly into the file. When you open the file later, most times you can not go back to the original state. If the program you use has both options, parametric is the preferred choice. If only direct editing is available you better save adjustments made in a copy of the original file, so you still have the original data available when needed (and you sure will, because you will occasionally want a different rendering, e.g. both a color and a black and white image....)!

1.3 First things first
You may wish to start editing right away after you have made some shots of something you care to remember in years to come. Don't! Take some time to put the images somewhere where you will be able to retrieve them easily and make a backup of them on some other medium and store that backup copy somewhere else! In the event that the unforeseen happens and the original files get lost forever, you still will be able to load the copies into your computer and use them again. Disaster will strike eventually and you better prepare for that.

Once you have made a backup, you can start to browse what you have got. Some of the images will be quite good, some acceptable and some simply will not be sharp because of camera shake, a shutter time that was to long for the fast moving subject or because of less then perfect focusing. Be critical and discard any image that has shortcomings that no processing would be able to rectify. It is no use trying to edit a image that can not be brought into a acceptable state – it merely is a waste of time.

After that first step you can start thinking about what needs to be changed to improve the images left. To be able to do that you need 2 distinct skills: being able to preview in your mind how a image could look when processed and being able to use the tools to accomplish that. When starting out, that is a daunting task. You do not yet very well know how to use the various tools available and you are also not yet very good in imagining how a image could look when processed appropriately. The only way to improve your skills is practice - you will learn a lot by doing. In this article I will discuss some of the more popular common tools available and I will show examples that demonstrate the effect a particular type of adjustment can have on a image.

2 What to change and in which order?

Editing software offers a bewildering amount of tools that need not be used all every time. The more common types of global adjustment possible more or less fall into the following categories:

• white balance adjustment
• tone adjustments (exposure, contrast)
• presence adjustments (clarity, vibrancy, saturation)
• detail adjustments (sharpening, noise reduction)

The following adjustments may also be needed, but less frequently:

• optical corrections (distortion, chromatic aberration)
• special effects (vignetting, grain)
• cropping and straightening

All the above adjustments are global adjustments that effect the whole of the image although not necessarily every part of it to the same extend. After having done all the global adjustments, you may find that some local changes also need to be made. Some of the above adjustments can also be applied locally. After carrying out all the steps you think necessary, you can save a fully adjusted copy for viewing on screen, for use on a website or social media etc or to be sent to a printer.

2.1 Workflow
The various adjustments need not necessarily be done in the above indicated order, but over the years that order has emerged as being the most logical. It is rather difficult to make adjustments to a image where the WB is off, so correcting that is something that most times should be done first. After that you can make changes to the exposure and adjust contrast if you feel the need for it. If the WB, exposure and contrast are ok, you can use the various presence adjustments to make subtle adjustments to the image and finally apply some sharpening and noise reduction where needed. When you have made all the global adjustments you find necessary, you may also decide to do some local adjustments. Not all programs offer that option - some only have a limited number of options, others like Photoshop offer many more.

In this article I will assume you do your editing starting with a RAW file. If you start out with a already processed image (JPG or TIF file) most adjustments can be made the same way, but sometimes with some restrictions. If those restrictions do result in quite visible effects, I will mention them. The examples shown usually come together with a screenshot showing what setting has been changed. Most of those screenshots are taken from Adobe Camera Raw (ACR), Adobe Photoshop CS5 or DXO Pro 9. That is because I have access to those programs, not because they are the best available. Other editing software will be able to accomplish the same effects, and the way those effects are obtained is usually quite similar. The examples are primarily there to illustrate visually what a certain kind of adjustment brings about, so you get a idea of what a change in exposure, contrast, clarity etc looks like. That images are accompanied by screen shots is done to draw attention to the fact that changes made often are reflected in the histogram.

2.2 Conversion and editing
Strictly speaking conversion and editing is not the same. Conversion means converting the RAW file into another format - the data may be ordered differently, but they are not changed. When also adjustments are made at the time of conversion (for example a change in contrast), that would be called editing. RAW files are converted with a so-called raw-converter, like ACR, Lightroom, Aperture, RawTherapee, Capture One, DXO etc. All those converters let you make changes at the time of conversion - and thus also could be called edit applications. Other programs like Photoshop, Photoshop Elements, and GIMP can not convert RAW files but need a other program to do that before the converted file can be loaded. Thus, Photoshop can be regarded as a quite elaborate plug-in for ACR and other raw-converters.

2.3 Histogram
In image #1 a so-called histogram is shown.

image #1:



The histogram is a chart where on the vertical axis the frequency of a particular luminance is shown and on the horizontal axis the luminance itself ranging from 0 or 0% (pure black) to 255 or 100% (pure white). The histogram can show the information in various forms, in the example the luminance together with the separate red, green and blue channels are shown. The most important parts of the histogram are the leftmost and rightmost part. If you see a spike there, that indicates that either the blacks no longer show any detail, or the highlights are blown out. Both is better avoided – be it at capture time or as a result of later processing. When you edit a image, the changes applied show up in the histogram. It pays to check the changes in the histogram frequently, because it gives a visible clue whether you have kept the tones in the image within the appropriate range. There is no such thing as a "perfect" histogram - the distribution of luminances will vary depending on the contents of the image. As long as you avoid spikes at the extremes, you will be fine. Much more information on how a histogram reflects changes made to a image, can be found in the following links:

http://www.cambridgeincolour.com/tutorials/histograms1.htm
http://www.cambridgeincolour.com/tutorials/histograms2.htm

3 Making adjustments

Now let us see what happens when we apply various adjustments. We will start with the most important adjustments: setting white balance, exposure and contrast.

3.1 White balance adjustment
When the colours are quite different from how you remember them when shooting, it may be necessary to adjust the WB. Most programs offer 4 different ways to do that. You can move two sliders that can change the colourcast from blue to yellow and from green to magenta, you can choose one of the presets (daylight, cloudy, shade, tungsten etc), you can type in a value for the colour temperature or you can use a eyedropper tool to sample one of the colours in the image that you want to be neutral (meaning to contain equal amounts of red, green and blue). All methods have the same effect, the colour rendering of the image is shifted based on the adjustment you made. Setting the WB is not merely exact science, it also has a element of personal preference. You may prefer a rendering somewhat warmer or cooler then would be technically correct and there is nothing wrong with that. The adjustment options are there to help you achieve the look you want.

When using the eyedropper tool to set the WB, it is important that the area you select as a reference, contains more or less equal amounts of all the 3 different colour channels (red, green and blue). Whether that is the case is usually indicated, either in percentages (0-100%) or numbers (0-255). See image #2.

Image #2:



In image #3* you see a before & after shot. The original was shot with the in-camera WB set to 5500 K. In this particular situation (the interior of a rather dark pub) the image was rendered far too blue. Setting the WB to tungsten in the editing program improved the rendering substantially.

Image #3*:



In image #4 both the histogram and the WB adjustment sliders are shown of a JPG file. You still can adjust the WB, but your choices are more limited. You now can only choose between "as shot", "auto" and "custom" and use the sliders or the eyedropper, but you can no longer type in a value for the colour temperature. The reason for it being that in the JPG file there is already a fixed WB set and only relative changes are possible.

Image #4:



3.2 Tone adjustments
Tone adjustments have influence on how lighter and darker parts are shown and how they are rendered relative to each other.


3.2.1 Exposure
The first adjustment that can be made is a change in exposure. Changing the exposure will lighten or darken all parts of the image with a equal amount. You may ask: "why would I need to adjust the exposure when I expose correctly?" There are various situations that may call for a adjustment after the shot has been made. The first being that you or the camera got it wrong - no matter how good the metering system in the camera works and how good your judgement whether some compensation might be needed is, results will not be perfect each and every time. You may also have decided to deliberately choose a exposure that will not render the image the way you saw the scene and thus you may need to introduce some correction after the shot. That can frequently happen when you expose to the right (ETTR). More on ETTR can be found in this link.

Image #5*:



In image #5* 3 images are shown. The leftmost image had the exposure reduced with half a stop, in the rightmost image the exposure was increased half a stop and the centre image is as how the image was shot. The original image was shot on a overcast day around noon in late September. The left image is to dark, whether the original image or the image that was lightened a bit is to be preferred, is a matter of taste. The original image to me best represents how it was; if you fancy the cars, the rightmost shot may well be the one you prefer because it shows more detail.

3.2.2 Contrast
Changing the contrast setting changes the difference between the lighter and darker parts of the image. When adding contrast the image gets more "punch" and increasing contrast also has some effect on perceived sharpness. Lowering the contrast makes the image look "flat". In image #6 the contrast in the right image has been increased quite a bit. To a certain extend that has improved the image, it is somewhat more crisp. At the same time though the trousers of the lady no longer show detail - just adding more contrast to a image is a rather blunt instrument that does not always work well. In this example (image #6*) using a more subtle adjustment would have been better.

Image #6*:



3.2.3 Tone control sliders
In many editing programs you will also find more sophisticated ways to influence contrast. You may see adjustment sliders for highlights, midtones, shadows and blacks. They work predominantly within a certain luminance range having little effect on other parts of the image. The most useful slider in that group is the highlights slider. It can help to get back some information from the highlights that were blown.

Image #7*



In image #7* a difficult to adjust shot is shown. Early October, hard light coming in through the window, very dark shadows. A simple reduction of the contrast setting is of no help here - it would only flatten the image and not bring back any detail. Therefore the contrast has been adjusted by means of the tone control sliders. The highlights have been reduced quite a bit to get back some detail in the blown out highlights around the window. At the same time the midtones and the shadows have been made a bit lighter, the blacks have not been changed. The change in settings is visible in the screenshots - the changes are also reflected in the two histograms. If you take a closer look at the histogram you can notice that the spike at the right has been brought down (the highlight area) and that the left most part of the diagram has been shifted somewhat more to the middle of the histogram - the result of the changes in the settings of the midtone and shadow sliders.

3.2.4 Tone curve
The most sophisticated way to control the tonal range of the image is offered by the tone curve section of the editing program. The name for that type of adjustment varies somewhat: Curves in Photoshop, Tone curve in Lightroom and DXO. The way it works is the same - by adjusting the curve you can both adjust contrast and the relative effect on various parts of the image. And if you do weird things while changing the tone curve, the result will also be bizarre. It is important to fully understand what the tone curve represents and what effect a change in the tone curve will have on the image. In image #8 a tone curve is shown, together with the image it is based on.

Image #8:



Apart from the endpoints the curve shown has 4 points that were added, and one point is selected (the black one). Beneath the curve you can see the input and output values for that particular point (78 and 50). The luminance values in the original image are numbered from 0 to 255, where 0 is pure black and 255 pure white. A tone curve thus is a chart showing how a image will be rendered with on the vertical axis the output value belonging to the input value on the horizontal axis. When the image is rendered as the camera made it, the tone curve will be a straight line - each and every input value will get the same value in the output (where output can be a on-screen rendering or the file send to a printer). When you change the curve by inserting a point and dragging it down, as in the graph, you will see that the output value no longer equals the input value. In this example the original value of 78 is brought down to 50, which means that it will be rendered darker then it originally was. Neighbouring values will also be effected, but to a lesser extend. When adding more points to the curve you can make quite subtle changes in specific luminance ranges. That is the beauty of the curves tool - you have far more control then when using the contrast slider or the tone control sliders because you can choose exactly where changes will occur.

Image #9*:



In image #9* you can see 5 different tone curves applied to the same image of a old farm tractor:


• A - linear full range curve
  
• B - full range curve with reduced contrast (flat)
  
• C - full range curve where the dark shadows are made black, the light grays made white and the tonal range in between squeezed into a higher contrast range
  
• D - a typical S curve that reflects the tone curve of classic film and that makes the image "pop"
  
• E - a unusual tone curve that can give very bizarre results


A is the tone curve for the image out of the camera, without any adjustment. B and C are shown for educational purposes only - they illustrate what happens when moving the endpoints of the curves. When editing you would not use them that way, but they show how contrast can be reduced or enhanced - the steepness of the curve in a certain range of luminances defines the amount of contrast within that range. D is a fairly typical curve for enhancing a image. In this case the curve only contains 2 additional points, if you want you can add a lot more to define a more complex curve. Each point that you add can be used to lighten or darken a specific part of the tonal range relative to the other ranges. In example E I have gone all out - only to show that you can do some really weird things with a tone curve, far beyond what normally would be needed for enhancing a image. Once you get a feel for what a particular change in the tone curve brings about, you will no longer really need the simple contrast slider nor the tone control sliders, although the highlight reduction slider may still proof very useful to recover blown out highlights and the contrast slider can still be used for a quick and dirty adjustment. Most editing programs offer a number of preset tone curves with names like linear, medium and strong contrast. You can use them as a starting point and make some subtle changes to them where appropriate.

In many editing applications you will also have the option to choose a different tone curve for the red, green, blue and composite channel. When you use that option you are not only changing the contrast based on the original luminance, but also based on the original colourmix. That can be very useful to avoid local colourcasts and get you a more subtle colour rendering. Using tone curve adjustments on a per channel basis is somewhat more complex, but it can give results not obtainable otherwise. Some times it may be necessary to adjust WB and/or exposure after making changes to the per channel tone curves. In image #10 you can see a example.

Image #10*:



In image #10* 3 different renderings are compared - the only difference being the tone curves. Rendering A is the original image with the WB based on the wall behind the stairs on the right with the help of the eyedropper tool. The WB may be set correctly, but there is a strong red/magenta cast all over the image. In rendering B that cast has been reduced by altering the tone curve for the red channel - by decreasing the output values for the midtones. The image is improved considerably, but there still is a magenta cast that is quite visible on the floor and on the faces. In rendering C the tone curve for the green channel was additionally changed with the result that in the midtones most of the magenta cast is gone. In this example the settings for the blue channel and the composite channel were not changed, only the green and red channels were set in such a way that that only certain parts in the image were influenced.

You thus have 3 different ways to adjust contrast, with different levels of sophistication (and ease of use). You can start out with using the contrast slider and progress towards the more sophisticated methods later on once you get a "feel" for what you are doing. You can also use them together if you prefer so. In some applications you may also come across a so-called levels adjustment tool. With the levels adjustment tool you can also adjust the way the darker and lighter tones are rendered. It can be regarded as a subset of what is possible with the curves tool and because of that is not discussed here. If you are going for all out sophistication, why not use the real thing?


3.3 Presence adjustments (clarity, vibrancy and saturation)
The presence adjustments do influence the way colours are rendered: vivid, subdued or completely faded.


3.3.1 Clarity
Clarity (Micro-contrast in DXO) adjusts local contrast. If you move the slider in the "more" direction the result is that small details become better visible. In image #11* a example is shown. The effect is quite subtle, adding the maximum amount may give nasty artifacts.

Image #11*:



3.3.2 Vibrancy
Vibrancy increases or decreases colour saturation but has little effect on skin tones and is especially visible in blue tones. It is very suitable to enhance the rendering of a blue sky. If you overdo it, you end up with a "postcard look". Notice that in the example (image #12*) the saturation setting has not been changed - the increased saturation is the sole result of adding quite a bit of vibrancy, especially visible in the blues and to a lesser extend also in the yellows.

Image #12*:



3.3.3 Saturation
Saturation acts on all the colours equally, making them more saturated or subdued. When increasing saturation too much, the image starts to look very unnatural, reducing saturation at first gives a more subdued look and when pushing down saturation completely you end up with a black and white image. In image #13* a example is shown. The rightmost part of the image with the saturation reduced completely is indeed a black and white image, but does not look like a nice black and white image from the film era. It is rather dull and misses punch. Just reducing saturation to zero does not yield a black and white image we like. The reason for that is that a black and white image obtained this way treats all colours in the image equally. In classic black and white photography that was not the case - classical film emulsions had different sensitivities for different colours and when in a particular situation you wanted a different rendering a glass filter was used during capture that reduced the intensity of some colours and enhanced others before the light hit the film. In the scene shown in image #13* you would most likely have used a yellow or a red filter to enhance contrast and get more detail in the clouds. If you want a more "realistic" black and white image you better use one of the commercial or free plug-ins and presets that do change the way colours are rendered in black and white. Most of them do quite a good job and are able to mimic various classic films including the grain that came with them. Satisfactory conversion to black and white can be quite tricky and will not further be discussed here.

Image #13*:




3.4 Detail adjustments (sharpening and noise reduction)


3.4.1 Sharpening
The term sharpening is somewhat misleading. You can not sharpen a image after capture. What you can do, is increase contrast around edges, which gives the impression of a sharper image. What we do when sharpening is trying to increase the perceived sharpness, in fact introducing a form of optical illusion. Sharpening controls usually comprise settings for the amount of sharpening, where to use it (the radius setting that defines the thickness of the edges to be sharpened), a threshold or detail setting that decides how much details are sharpened and a mask setting that you can use to influence whether all of the image will be effected or just certain parts of the outline of elements in the image. At first it will be rather difficult to choose the optimum settings. The best way to get a feel for it is to look at the image set to a 100% view (1:1) or even larger and then watch what happens when you move the various sliders. The high magnification ratio is necessary to be able to clearly see the effects of sharpening on screen.

DXO comes with a additional setting called edge offset. When you change that setting you apply more or less sharpening towards the edges of the image, useful if you want to sharpen relatively soft corners somewhat more then the centre of the image. Especially when shooting wide open you may see a need for that. Lightroom has a feature called masking that can show you visually how the image will be effected. When you move the masking slider in the Sharpening section of the Detail panel between 0 and 100 while pressing the [Alt] button (on a PC), you are shown a black and white rendering of the image where the white areas are effected by the sharpening and the black areas are not. It is a very useful visual equivalent of the DXO threshold setting.

Image #14*:



Sharpening is a rather complex subject that is dealt with far better and in more detail in some other articles on Dyxum, see for example Sharpening-A brief introduction written by Micholand.
Basically there are three kinds of sharpening you should be aware of: capture sharpening, image sharpening and output sharpening. Capture sharpening is needed when converting a RAW file into a image file to compensate for the inherent fuzziness of the RAW file. If you edit a JPG file generated by the camera, that kind of sharpening has already been applied. Image sharpening is the sharpening that you apply because you want the image you view a bit sharper, output sharpening is some extra sharpening needed when you want to print the image. You can get a impression of the output sharpening needed when the editing program you use has the option to show the image in proof mode on your screen; the image looks softer and may need some extra sharpening and contrast. The reason for that is that when printed on paper part of the sharpness and contrast gets lost due to the printing process. In Adobe software you can choose the amount of sharpening applied at the time of print by choosing one of the available options (high, standard, low). There are also plug-ins available that let you choose the amount of sharpening depending on the type and dimensions of the paper used, see for example NIK Software Sharpener Pro 3 - Getting the most out of Output Sharpener (video).

The need for sharpening will vary: nearly all images, when converted from RAW, will need some capture sharpening – images shot with a camera without a AA-filter (like the A7R) being the exception. All images also will need some extra sharpening for printing. In many converters and editing programs both is taken care of more or less automatically. If you want to additionally (image-)sharpen (part of) a image, you will need to decide yourself on the settings of the various sliders. You may also want to reduce sharpening occasionally, especially with portraits of females.

3.4.2 Noise reduction
When shooting in low light conditions with a higher ISO setting, present day sensors will not always be capable to capture the correct luminance or colour with every photosite. The result is that in certain areas of the image you may see specks of the wrong colour and/or luminance. Especially in areas where a uniform colour should be visible without a structured pattern this can be very noticeable and annoying. The amount of noise will vary from camera to camera, the ISO setting chosen, in the relative amount of chroma (colour) and luminance(lightness) noise and in the way it looks. The a900 for example has a rather coarse noise structure, whereas the a99 with its later generation sensor shows a much finer and less irritating noise structure. At higher ISO settings the noise also may be accompanied with smearing - a distinct loss of sharpness that may be even more objectionable then the noise itself. In image #15* the original image is shown together with 2 images with different amounts of noise reduction. The image was shot with the a900/24-70 CZ set to 24 mm, ISO 6400, 1/45s, f/ 2.8 and can be regarded as a worst case scenario (both noisy and unsharp due to the high ISO setting).

Image #15*



Because of that you better avoid shooting at the highest ISO settings if you can, usually a setting two stops lower then the maximum setting possible will keep noise within acceptable and correctable limits. Thus if the maximum ISO setting available is ISO 6400 it is wise to restrict yourself to use ISO 1600 and lower, when the maximum setting is ISO 25600 you better not go beyond ISO 6400 - less is better when you want to avoid noise. Of course the high settings are there to be used, but only do so when really needed. If you have the choice between no shot or a shot that will show noise, just take the shot and deal with the noise later. In image #16* the same image as in image #15* is shown uncropped in original and in adjusted format. When you are not a pixel peeper by nature, even noisy high ISO images can be acceptable!

Image #16*:



Noise often is the subject of heated debate where people try to convince you that some manufacturers are dealing with it a whole lot better then others. There are indeed visible differences. But I doubt whether they are as important as some people might like you to believe. Most shooting is done under decent lighting where noise is no problem at all. And if you shoot in the dark and the noise is clearly visible on your computer screen set to 100%, you should realise that when you print the image on a paper of 60x90 cm you will look at it from a far greater distance then when working on a computer - so the noise will be far less visible and objectionable then you envisaged. In real life noise is much less of a problem then all kinds of so-called experts tell you…

Now that I have got that of my chest, what can we do about noise? The first thing to do is take a closer look at the kind of noise. Is the noise generating the wrong colour or the wrong luminance or both? If the colour is wrong you should start with adjusting the colour or chroma sliders. That will remove a large number of the specks of the "wrong colour" - the result may be fine tuned with additional sliders for detail and smoothness. If after that there still is visible and objectionable noise with a grainy structure caused by varying luminance that may also be reduced by using the luminance slider. Additionally you may also find sliders for detail and contrast to fine tune the result.

Noise reduction always comes at a price. Reducing (luminance) noise may result in loss of sharpness. That was to be expected - noise reduction essentially comes down to reducing contrast in certain areas and that unfortunately also brings about a loss of (perceived) sharpness. You can not fully compensate for that by adding some additional sharpening - when you increase the sharpening you will also reintroduce some of the noise that you managed to get rid of! Therefore you should always start out with reducing chroma noise first and only try to reduce luminance noise after that and only when really necessary. Finding the right balance between reducing noise and retaining sufficient sharpness will not always be possible. If you are unable to get the result you want, there is only one way out: change the image to black and white. The noise will not disappear completely but it will be much less visible and objectionable!

In recent years quite a number of improvements have been made in sensor design, so noise now is far less of a problem then it used to be. Also the way conversion software tries to reduce noise has been improved quite a bit, so additional software to reduce noise will be far less required then a few years ago. Recently (2013) DXO has introduced a new method of noise reduction in their software. The result is quite spectacular. Hardly any loss of sharpness and a far better reduction of noise then previously available. Nothing comes for free though: due to the very calculus intensive way of finding out what changes need to be made, the process of reducing noise is very time consuming. For a a99 image it can take up to six or more minutes, but with important images that may well be worth the wait. The program also includes a standard noise reduction component that works within seconds and in most cases is quite adequate. If interested you can find a relatively recent comparison between Lightroom, DXO and Capture One here (original French text) or here (Google translation into English).

What you learn from the comparison is that all converters nowadays do a good job, and apart from the DXO Prime variety, there is no clear winner. The other contestants do a quite good job as well, but do so in different ways - depending on the actual contents of the image. You may prefer one method of doing it above the other. The tests also show that reducing noise in all cases leads to a decrease in sharpness.


3.5 Optical corrections (distortion and chromatic aberration)

No lens is perfect. Especially wide angles and zoom lenses are not. Some of those shortcomings can be minimised with appropriate software. Shorter focal lengths often tend to exhibit some barrel type distortion, longer focal lengths tend to exhibit pincushion distortion. In image #17 you can see what happens when the two types of distortion are present and how the different kinds of distortion got their respective names.

image #17:



Lenses also may show chromatic aberration, especially near the edges of the image. It is visible in structured objects with fine detail - what should be rendered as a more or less evenly coloured line may have a green edge on one side and a magenta on the other. Chromatic aberration can easily be brought down to acceptable levels with appropriate software. Image #18 shows a example of chromatic aberration.

image #18:



Correction of distortion has only more recently been introduced in most image editing applications. "Purists" have long been ignorant to the fact that their fabulous high end lenses from the best supplier in the world still can exhibit visible shortcomings. Lens correction got some more appreciation after some clever camera manufacturers started introducing software based in-camera corrections - admitting that their lenses were not so perfect as advertising in the past would have liked you to believe. In-camera lens correction makes it possible to obtain good image quality with relatively simple and cheaper lenses. Nowadays many camera manufacturers include software based in-camera lens corrections on camera-generated JPG's and inclusion is no longer limited to typical consumer models. Unfortunately in-camera corrections can only be used when shooting JPG's, another reason to shoot RAW!

When shooting RAW you can add optical corrections yourself during processing. DXO has offered lens correction for particular lens/body combinations for quite some time, more recently Adobe and other software vendors have followed. The corrections can be chosen manually or carried out automatically based on a particular lens/body combination (DXO) or a specific lens (Adobe). For Adobe software you can also make your own correction profiles and use of a specific profile is not limited to the lens it was made for. You thus can use a correction designed for a particular Canon lens to "improve" the rendering of a Pentax lens if you want to...Who on earth would want that however, is totally unclear to me. DXO follows a far more strict approach: a correction module can only be used for a particular lens/body combination. That particular correction is the result of extensive optical measurements, a approach that can not be replicated by a user because he will lack the costly measuring equipment needed. DXO also uses those measurements as a basis for "location aware" sharpening (where the edges are sharpened more then the centre to bring down corner softness), to reduce vignetting that may be visible with wide angle lenses and to correct chromatic aberration. The Adobe approach is far more simple - just a few shots can yield a profile. The beauty of the Adobe system is that it is fairly easy to make a correction profile yourself (as was already longer the case with additional software like PTLens). There is of course a trade off between simplicity and accuracy. As Adobe itself puts it:

"Keep in mind that a single iteration for a camera/lens combination will optimally correct for the behaviour of a single given lens setup (focal length, aperture, and focus distance)."




3.5.1 Lens correction
Lens correction may be well something you never thought to be necessary. Once you have seen results, you most likely will no longer do without. Seeing is believing! The easiest way to apply corrections is automatically - if you have access to the right correction module. If not, you can set corrections manually. That is somewhat more work and it is not always clear what the best setting is. It depends a lot on the actual contents of the image. If the image contains straight lines parallel and close to the edges of the frame, barrel or pincushion will be clearly visible - straight lines are not rendered straight but curved - concave or convex. They can be straightened automatically if a suitable profile is available or manually. For manual adjustment just move the slider back and forth until you like what you see.

Image #19*:



In image #19* a example is shown for the a99/24-70 CZ combination shot at 24 mm. It is difficult to notice the differences when the images are shown next to each other - the 24-70 CZ is not a bad lens! When you click on the image you get a far better view. Some more examples:


• Sony a900/20-2.8 combination - barrel distortion correction
• Sony a900/70-200 SSM set to 200 mm - pincushion distortion correction
• Sony a900/135 CZ - slight pincushion correction


3.5.2 Chromatic aberration
The green and magenta striping along the edges of fine structures near the edges and corners of the frame can easily be corrected. Additionally you can correct purple fringing. Purple fringing can occur as a ghost like magenta glow around high contrast edges in the image. It differs from chromatic aberration perse, because the fringe will only be magenta (and not green) and can be visible anywhere in the frame, whereas chromatic aberration usually is seen only near the edges. In image #20 a before and after shot is shown: in the top half the chromatic aberration is visible, in the second half it has been removed by just one mouseclick.



Some more examples of chromatic aberration can be seen here.



3.6 Special effects - adding vignetting and grain
Most editing software is able to add a vignette to a image and a form of grain that more or less makes the image look as it was shot with film. Both effects may add a special look to the image and both have nothing to do with trying to get the image as close to nature as possible - exactly the contrary! Introducing a slight vignette may be appropriate to add more emphasis to the centre of the frame, in a portrait for example. When adding a vignette the corners and edges of the image are made darker or lighter - whatever you prefer or think necessary. In image #21* a example of a negative and a positive vignette is shown alongside the original image.

Image #21*:



Adding grain will be visible all over the image, a kind of nostalgia I personally can do without. But to each his or her own...

3.7 Cropping and straightening
Cropping is simply cutting off parts of the image - usually because that improves the composition or when you want to us it where a specific height/width ratio is required. At what time in the chain of making adjustments the image should be cropped is up to you - you can start with it, or do it at any other time. Personally I consider it the last step in the adjustment process, you may prefer to do it much earlier. Straightening means setting the horizon horizontally or vertical lines vertically or purposely under a angle to add drama to the image. The process is quite straightforward - you draw a line over the image that you want to be horizontal or vertical and that's it. If you want a certain angle, you can type in a value in degrees, use a slider or draw a non-horizontal line across the image. Both cropping and straightening are quite simple - WhatYouSeeIsWhatYouGet. And if the result does not suit you, you can do it over again. In image #22* several examples are shown.

Image #22*:



In Lightoom there seems to be yet another way to "straighten" a image. Under the panel "Lens Corrections"/"Basic" you can click on a number of buttons named "Level", "Vertical" and "Full" respectively. ACR contains a similar set of adjustment options. The fact that those buttons reside under the "Lens Correction" panel, should be taken as a hint that this tool is something completely different from what the crop and straightening tools do. What you can do with the buttons in the Lens Correction/Basic panel is let Lightroom adjust the perspective of the image. That may result in vertical lines and horizontal lines that have the position where you want them. In most other editing software this kind of adjustment is known as perspective correction. You more or less can mimic some of the adjustments possible with a view camera - after the shot that is. "Lens Corrections" is a strange place to harbour these tools imho - they are not meant to correct anything, but to alter something...If you use this tool, be aware of the fact that some parts of the image may be stretched - loosing some sharpness and resolution in the process.

4 Local adjustments

All the above adjustments were carried out globally - the whole of the image is effected, all be it not always to the same amount. Various software vendors offer the possibility to also carry out local adjustments when converting a RAW file into a image file.

The way Adobe handles local adjustments differs depending on whether you use Lightroom or Photoshop. In Lightroom both global and local adjustments can be made when working with a RAW file - the adjustments are not brought into the file but stored separately in the catalog (and in a XMP file when chosen) and are used for showing the adjustments on screen. When exporting the file they are incorporated in the file generated based on the the list of virtual adjustments carried out when editing. The resulting file no longer is a RAW file (with the extension ARW for Sony cameras) but is a real image file in one of the available output file formats. When working with Photoshop the processing of the RAW file is done by the ACR plug-in. After conversion the contents of the file is transferred into Photoshop and unless you choose otherwise will be saved in the proprietary PSD format from that moment on.

Carrying out local adjustments in Lightroom is simple. You select a certain area of the image that you want to adjust with either the gradient filter, radial filter or adjustment brush tool. When one of those tools is selected, you will see a number of sliders that can be used for various adjustments, changing white balance, exposure, contrast etc. It works exactly the same as with the global adjustments, but only locally on the area you selected. In Photoshop and similar programs the number of options for local adjustments is sheer endless and for the novice rather overwhelming. That has something to do with the origins of the software: Photoshop and other similar software was originally designed for the digital artist who started out with a raster image from a scanner and subsequently wanted to add all kinds of effects and objects to the scan - creating a whole new image that not necessarily looked anything alike the original scan. There is a reason why GIMP is a acronym for GNU Image Manipulation Program….

In Photoshop you have 2 choices to introduce local (and global) adjustments - direct in the image you started with (which subsequently is irreversibly changed when you save it) or you can make the adjustments on separate layers (the standard when you work on a file in the native Photoshop PSD format). A layer can be regarded as a kind of transparency placed over the original image on which additional imaging effects or images are applied. Using a separate layer to host each adjustment is the best way to carry out adjustments, because the original file remains unchanged until the combination of the file and the layers is saved in a output format like a JPG or a TIF file. When you decide to save the image in a typical output format the layers will be "flattened" (combined with the original image) and the changes applied become final and irreversible.

Discussing all the options for local adjustments goes beyond the scope of this article (and also far beyond my knowledge of the subject). I will restrain myself to only three examples of the use of local adjustments in Lightroom - use of the graduated filter, the adjustment brush and the spot removal tool. All are easy to apply and can be very useful.

4.1 Graduated filter
The graduated filter can be very useful when dealing with landscape images. Often the sky is very light and the foreground is much darker. You could have shot the image with a graduated neutral density filter as is often advised, but unfortunately you did not bring one with you. Thus you are stuck with a image where either the sky is far to light and without any detail or with a foreground that is much darker then you want it to be. The graduated filter can help out. Use of the graduated filter is not restricted to landscapes. In image #23* a example is shown - the lady on the left and her white handbag have been darkened a bit. If you click on the image a new window opens showing a more detailed sequence of events on how to apply the graduated filter.

Image #23*:



4.2 Adjustment brush
With the adjustment brush you can paint over part of the image to apply a specific effect to that area. One of its uses is to lighten or darken certain parts of the image, just like the burning and dodging in the classic darkroom. It works exactly the same way, but the area of the image that is effected can be chosen much more precise and adjustments can be far better controlled. Use of the adjustment brush is not limited to changing local exposure. You may also change saturation, sharpness, contrast etc. or use a combination of various adjustments. In image #24* a example is shown where the face of a man is lightened somewhat - "fill flash" after the shot!

Image #24*:



Both the graduated filter and the local adjustment brush may seem rather crude instruments when compared to the sophistication that full fledged image editing programs can offer. No doubt that is true. The advantage of these more simple tools is that learning to use them only takes a few minutes and from that moment on you can get very rewarding results. Learning to use the myriad of local adjustments possible in Photoshop, GIMP and the like can keep you occupied for the rest of your life.

4.3 Spot removal
Occasionally you may note blobs of slightly different colour, especially in sections with a bright blue sky. Those blobs can be the result of dust on the sensor. That sensor should be carefully cleaned before shooting again, but that is not much help for images already shot. The spot removal tool (available in nearly all except the most basic editing programs) can help out. You open up the image concerned, open the spot removal tool, set the size to just a little bigger then the spot you want to get rid of en then click on or paint over that spot. Most programs will automatically find a similar area in the neighbourhood to replace the area you want to clean up. If the automatically selected area does not suit you, you can select a different area. In image #25* you can see a example, image #26 shows how it is done.

Image #25*:



Image #26:



Use of the spot removal tool is not limited to the removal of "dirt" in the image, it can also be used to remove skin blemishes, freckles, warts and the like in portraits as shown in image #27*. Whether that is a acceptable practice is for you to decide. Personally I restrain myself to remove imperfections only if they are of a nonpermanent nature - removing imperfections everybody knows about gives the image a very processed and manipulated look.

Image #27*:

5 Some additional remarks

In this article the focus was on how to make simple software based adjustments that can visually improve images after the shot was taken. It is not meant to be a substitute for the diligent studying of a software user guide, nor does it substitute practice. Both (the user guide and practice) can help you improve your skills far beyond what was shown here. It is neither useful nor possible to discuss all details here and it is also not necessary. The adjustments described can be carried out in nearly all editing software in very similar ways. Just starting with simple experiments will teach you rapidly and will inspire you to new ideas, because not only you will learn by doing, you will also at the same time develop your skills to preview in your mind what kind of adjustment would be most appropriate. From that moment on you will continuously discover more and perhaps unusual ways to arrive at the image as you want it - which is essentially the goal of any form of editing!

The software vendors referred to in this article are by no means the only suppliers of capable software. There is a multitude of products available, ranging from the very basic to extremely sophisticated - both free and commercial. That a product can be obtained free does not mean restricted in options or of lesser quality. GIMP and RawTherapee are prime examples of high quality software available at no cost at all. Camera manufacturers also supply software to be used with RAW files from their cameras. Most camera manufacturers tend to be better at camera design then at development of user-friendly software though.

6 What is the best software for you?

The most appropriate answer ever given to that question was in a statement from one of the Administrators of this forum, a statement I occasionally refer to as Pegelli’s Law :

pegelli wrote: I believe there is no "best" raw converter. The best one is the one you feel comfortable with and has an interface you like to work with and becomes intuitive for you.

Indeed, I could not have put it any better myself....now start using the software you had lying around for quite some time!


Edited by romke - 12 January 2014 at 17:21

ditto Mipr´s comment.

I´ll be reading this again, and looking more closely at the picture comparisons you made for sure.

Thanks for the time and effort you spent on it.

Thank you for your time and considerable effort in doing this! As my editing has been pretty much, by the seat of my pants type, I enjoy these articles very much. Lots of info here sure to help. Thanks again!

Thanks so much for this primer! This should be included in the box of every new camera. It takes so long to learn how to process raw files when you are using the trial and error method, that the in camera jpeg is most times better. This really gives me a much better understanding of some of the basic tools.

Excellent article! Really helps an editing noob like me get an idea what to look for!

Great post, I just got LR for Christmas and am very new to the whole PP thing and this is a nice summery of the important bits.

Thanks for this thoughtful article. Very helpful.   

Mike