- 1 Lightroom RAW Development
- 1.1 Lens Correction, Chromatic Aberration Correction, Camera Profiles
- 1.2 Broad White Balance Correction
- 1.3 Color Correction
- 1.4 Selective Adjustments
- 2 Photoshop Edit
- 3 Final Touches
- 4 Summary
Usually, when retouching portraits, quality is key.
Everything needs to be perfect in order to have the model presented in the most flattering way possible
However, when we are taking photos, everything is raw, the light can never be perfect, nor can the conditions.
The camera is set to RAW file shooting, which needs to be developed as well.
In this article I will demonstrate the way I do a portrait retouch; it doesn’t necessarily have to be the way you will do it, but you can learn many of the techniques and tailor them to your needs afterwards.
The goal is to achieve a colorful portrait since it will be used in the making of a poster for a party, and the client wants a brightly colored photo that will attract attention.
You need to be aware that for most of the edits I’m using a pen tablet, since it provides me with the accuracy I need for air brushing and masking. I’m not saying it can’t be done with a mouse, it is just that it will take you a tad more time due to the imprecision.
Lightroom RAW Development
Here I am going to do most of the work which includes the general development. That means contrast and color correction, correcting the light as much as possible, noise reduction, chromatic aberration correction, lens geometry and vignette correction, but not sharpening (I will sharpen the final image, just not before most of the editing).
Lens Correction, Chromatic Aberration Correction, Camera Profiles
Before anything is done to the image, you need to correct for the lens geometry distortion and the chromatic aberration. Usually, when you go to the effects box in the develop module and tick the “Enable Profile Corrections” and the “Remove Chromatic Aberration” boxes, it will do the job for you. These are both found in the “Lens Correction” section in Lightroom. However, the automatic profiles don’t work properly with certain lenses, meaning that if you are using a lens like the Tokina 11-16mm f/2.8 DX, the box won’t correct for it automatically, instead you’ll need to go to the “Profile” tab in the same box and from the drop down menu just choose your lens manufacturer, and from then on Lightroom should do the job just fine. This will correct the geometric distortion that the lens has, and the vignetting. You can do this manually as well, but the profiles are much more accurate - in fact, they should be perfect.
Chromatic Aberration Correction
Some lenses suffer from a great deal of chromatic aberration, and while ticking the “Remove Chromatic Aberration” box usually does the job, at times, you may need to do it manually. You can do this by going to the “Color” tab in the “Lens Correction” section in Lightroom, and then adjust the sliders manually.
The easiest way to do this is to zoom to 100% (done by clicking once over the image) and look for areas where the color is fringing (usually around high contrast edges). Now take the “Fringe Color Selector” eyedropper tool and carefully click over an area that has color fringing. Make sure that you are clicking the fringe itself. Do this to all the colors that are fringing. This will correct them generally - all you need to do is just expand the range of color corrected a tad since the eyedropper isn’t perfect. You can expand the range by widening the selections in the “Purple Hue” and “Green Hue” sliders.
This is something often overlooked by photographers when they are editing their photos. Lightroom treats all RAW files equally, which is not the best way to do it (not sure why they don’t have precise camera profiles yet, but it is what it is) thus you are required to pick one from the given profiles (or create your own).
Camera profiles are situated in the “Camera Calibration” box in Lightroom. It is wise to leave the “Process” box as is, “2012 (Current)”.
Broad White Balance Correction
While white balance is something really trivial for correcting, you still need to do it at least twice. First you’ll need to do a general correction, and then fine tune it at the end. This is due to the fact that when you do heavier edits, it tends to show the imperfections at the beginning. But you still need to correct it at the beginning in order to do the rest of the edits correctly.
Take the “White Balance Selector” (you can do this by pressing W) and click on an area of the image which you know is neutral (say a black shirt, gray shirt, white shirt, or anything that you are sure has no other color than monochrome). This will act as a gray card and Lightroom will correct for it.
Highlights And Shadows Recovery
Now we are getting to the cool part. My general workflow is to recover the shadows and highlights first, thus I’m reducing the “Highlights” slider to -85 and doing the exact opposite to the “Shadows” slider by setting it to +85. Note that these values apply to my image; yours may vary but it is still the general direction you should be heading. You’ll need to judge by eye about how far you’ll push the recovery.
Blacks And Whites Correction
After the recovery is performed to the shadows and highlights, we need to correct for the white and black points. This is done via the “Whites” and “Blacks” sliders. Since you need to push the sliders right before the image starts to clip, you need to know when that is occurring. Clipping is when the shadowed or highlighted parts of the image become pure black or pure white with no color detail whatsoever; this is something you want to avoid.
While moving the “Whites” slider to the positive side, and “Blacks” slider to the negative side, hold the Alt key. This will show a mask instead of the image, and once the highlights or shadows start to clip, they will appear as the opposite color. When you see the first dots of clipping, stop pushing the slider. Do that for the both the “Blacks” and “Whites”.
“Clarity” is a contrast slider. But instead of adding contrast as the regular slider, this adds it just to the midtones, thus acting as a sort of contrast and sharpening fusion. I usually tend to add a setting of around 20 to 40 depending on the image. Don’t overdo it however, or it will make the image look blotchy.
Vibrance and Saturation
In order to know which one you should use, you need to know the difference between them. Vibrance boosts muted colors, meaning that before it saturates certain colors it analyzes the picture and boosts the colors which are less present first, only afterward that will it saturate everything else equally. Saturation on the other hand just multiplies the colors as they are, meaning that there will still be colors which are less saturated than others.
That is why I almost never use the “Saturation” slider, “Vibrance” does it perfectly. Usually I don’t use values larger than 40 with the “Vibrance” slider.
This is the part where most of the magic happens. HSL stands for Hue, Saturation, and Luminance. In this section you can affect the hue, saturation, and luminance to each color channel separately. This way you can do quite powerful corrections with just this section alone.
On this particular image, there is a lack of fill light, meaning that the model is a tad darker than intended. Now I can brighten the model by using an adjustment brush with increased exposure, but that will generate heaps of noise which I like to avoid. However, by increasing the luminosity of the orange channel, I will boost only the skin color, while the yellow channel will bring out the hair and highlights as well. The red channel is boosted to match the exposure on the skin with the reddish parts of the model, like the lips for example.
Since I want to restore the color on the lips, I will boost the saturation on the red slider, and I’ll do the same on the green slider in order to bring out the trees in the background.
I will reduce the luminance on the blue and aqua sliders in order to bring down the sky a bit (will fix it later selectively).
Since the hues are to my liking, I will not alter them.
So, only a few sliders moved and yet so much has changed in the image, in just one section.
Since the sky has the sun in the background, but it is a bit blown out, I will need to repair that, and simulate the look that the client is after.
The first graduated filter I’m going to use is a top down with lowered exposure and highlights in order to bring back the details in the sky. As you may notice, there is a slight whitish cast on the left hand side, that is due to the sun being just outside the frame - I’ll use that to my advantage later. The filter also has some cooling effect, meaning that I’ve adjusted the white balance to it in order to make the sky blue.
I’ve brushed out (new feature in LR6) the model from the mask since I didn’t wanted her to be affected by it.
The second graduated filter will go from right to left in order to cool down the side behind the model to add a complementary color to the picture. The exposure stays the same.
Usually I use radial filters in order to enhance the sun or to simulate a light source. For this photo I am going to do both.
I’ve set the radial filter centre point right next to the sun location since I want to simulate the sun being brighter. I’ve inverted the mask in order for it to affect the inside of the selection circle and reduced the feather so it is as smooth as possible. By making the filter warmer and brighter I’m effectively simulating the sun being in the background. I’m going to brush out the model again, the sun is behind her and it shouldn’t affect her that much.
Now since everything is in order, it is time to do the heavy lifting in Photoshop. Here we will smooth out the skin, correct the contours, correct the light even further, as well as simulate some effects which are required by the client.
Frequency Separation Preparation
To prepare the image for frequency separation we need to create two more layers which are the exact copy of the original one. The first copy would be the Tone layer (where you can do the tonal changes on the image), while the second will be the Texture layer (where you can affect only the texture of the photo).
Tone layer is created by using “Gaussian Blur” with settings from around 8-9 pixels. And the Texture layer is created by subtracting the Tone layer from the Texture layer. This is done by using the “Apply Image” function which is located in the “Image” menu. Of course before you activate the “Apply Image” function have the Texture layer enabled and selected.
In the “Apply Image” dialog box find the “Layer” drop down menu, and pick the Tone layer from there. Next, as “Blending” pick “Subtract” from the drop down menu, and have the “Opacity” at 100%, set “Scale” at a value of 2, and “Offset” at a value of 128. These values have to be exact. The “Preserve Transparency” and “Mask” boxes should be unticked. Now hit “OK”. The result should be a gray image with some details on it. From the Layers panel, set the blending mode for the Texture layer to “Linear Light”. Now you should be able to see the original image.
Select both Tone and Texture layers, and press CTRL+G. This will group the layers for better workflow. Turn that group on and off several times, if there is no visible change on screen, you’ve done a good job.
Frequency Separation Editing
The first thing I do is to broadly correct the tones, meaning that I want to smooth out harsh shadows, and take care of the uneven gradients.
I do this by brushing over the Tone layer. Now, since I want to be able to edit everything later (non destructive workflow) I will create a new layer right above the tone layer and use that instead of the tone layer. I will still sample the color from the tone layer.
You can sample the color you are going to use by having the “Brush Tool” selected and pressing the ALT key while clicking over the color you want to sample.
The brush settings I use are “Opacity” at 100% and “Flow” at 1 to 5%. I usually use pressure to control this but for some reason Photoshop is not recognizing it lately. This would work well for this tutorial as it would apply for mouse users as well.
Sampling the correct color is crucial. The whole idea is to sample the color every time you change location, in order to have a seamless transition.
Once you start painting the new tone you’ll realize that the changes are minimal, fear not, that is due to the fact that there is little tone information left inside the texture layer, and you’ll fix that later.
Another way to do this (and one over which you don’t have much control, especially around edges) is to use a selection with decent amount of feathering and Gaussian blur. If you overdo it, it looks flat, but if you don’t blur enough the blur is noticeable, and if you do it around the edges it creates a weird glow. It is best to do this on areas that are away from the edges to be safe.
This is the tricky part. You will not be able to use the “Healing Brush” because it doesn’t work properly due to the lack of information. Clone stamping also doesn’t work properly because it ruins the texture.
This leaves us with limited options. Either you copy and paste in the texture layer (you can’t use another layer since they will blend and make a mess) or you can use the “Patch Tool” -this works perfectly in this kind of scenario.
Remember that you need to use textures which are general, for example: if you are trying to replace a blemish - don’t replace it with the texture of hair. On the other hand if you feel like the texture in a certain area is too strong, you can replace it with a texture from another area which isn’t as strong. The good thing about this is that if there aren’t extreme contrast differences you don’t have to worry about the tonality or the light, because it will work out perfectly 95% of the time. Just select and replace until you are satisfied with the result.
This is the part where you can restore or create more contours. You can usually do this by dodging and burning, but I don’t like how those tools affect the color (burning increases saturation since it is doing tonal compression). That is why I use two layers: one set to “Lighten”, the other set to “Overlay”.
The layer which is set to “Lighten” is used for adding highlights, and thus a white brush is used. Set the “Brush Tool” “Opacity” value to around 30% and the “Flow” to 1 or 2%, while the “Hardness” setting should be at 0%. Now paint away, making sure that you paint over areas which already have highlights, but don’t overdo anything. This is used to enhance the highlights a bit, for giving more depth to the image if the light is a tad flat.
The same applies for the layer that is set to “Overlay”, but in this case you will be using a black brush. The other settings are the same. Paint over areas where you want to enhance the shadows, and of course, as previously mentioned, don’t overdo it - be gentle and careful. Paint over areas which already have shadows, and remember that you are only enhancing here.
Since the image is supposed to be for a poster, it needs some exaggerated colors. That is why I am going to simulate more orange and blue colors generated from the sun and the sky.
The layer for the sun will be of orange color, and set to “Screen” blending mode. Make sure that if you simulate this very same effect you are careful with the saturation, because you don’t want to overdo it and make it look fake. I will let some of the color spill over the model in order to make it look more natural, even though it is far from it.
The blue layer is set to “Vivid Light” and the opacity is very low since there is already a good amount of blue color in the picture itself. The opacity of that layer is just 5%.
And now, we can save the image in the same location as the .psd file in order for Lightroom to pick it up right away.
After I have the final image loaded into Lightroom I do the final touches on it. First of all I need to sharpen the photo, and I’ll do that with the settings of 120% “Amount” and 95% “Masking”, while the “Radius” is set to 0.5. That will apply quite a strong sharpening effect but it will be limited to the small details and it won’t generate halos since the radius is just half a pixel. Meaning, it will enhance the tiny details with longer edges (won’t affect noise, but it will affect hair and eyelashes) without messing up the texture.
Additionally, here is the part where you can apply any film looks and other presets you want used. I tend to create my own film look since the presets never get it right.
This is just one portrait editing process from the beginning to the end, for a portrait made for a commercial purpose. It was ordered by a client and done with a previous set of requirements. Even though there is a bit too much color or effects for certain people (including me) the general concept is quite good since it will be used as the background for a poster. It has to attract attention while being flattering at the same time, and it has enough negative space for some text to be written on top of it.