To Texturize Water or Not?

When constructing a water composite, and trying to give it the look and feel of an old, classic painting by adding a texture, the question about when to texturize the composite will come up. For this genre, a model with dark hair in a dark swimsuit works best. Let’s begin with this studio portrait of Indonesian model Karina Cognomen:

Let’s use an image of some nice clouds against a blue sky, such as this HDR image taken in Singapore:

For the texture, let’s again use a Painterly texture from Photomorphis, in this case #7:

The question will invariably come about how to apply the texture. Should you texturize the clouds, then generate the water in Flood 2? Or should you generate the water first, then apply the texture to the entire clouds/water background? To save you the trouble, I have made both versions. Here is the first version, where the texture has been applied to the clouds only, then the water was generated:

Here is the version where to water was generated first, then the texture was applied to the entire background. Notice the clouds look different. That’s because texture had to be stretched to cover the entire background:

I prefer the second version, where the texture is applied after generating the water. It is much simpler. Because we are using the “paint over” technique, we only have to remember apply the color and luminosity of the texture to the model, to assist in unifying her with the background. But we want the texture’s color and luminosity only, not the texture’s detail painted on her. See my blog post of October 6, 2014 about adding textures to a boudoir portrait for information on how to do the “paint over” technique. Still, the texture on the above image looks a little too prominent for my taste. To remedy this, I applied a Gaussian blur of 8.0 pixels to the texture only. This is the final result:

Musings on the Square Format

Back in the days of film, photographers loved their 6 x 6 cm cameras. The square format was large, and provided plenty of detail. Furthermore, the photographer could orient the camera just one way, and did not have to worry about landscape or portrait orientations. 95+% of the time, I shoot portraits with the camera in portrait (vertical) orientation. Because I enjoy compositing, the composite often ends up with that same portrait orientation, as can be seen in this example of Indonesian model Karina Cognomen:

I like this image quite a bit. In the past I have included islands, the moon, planets, the sun, clouds, lightning, rainbows, flying birds and bats and even zombies in the background. Those added elements add complexity to the composite, and can easily draw the viewer’s eye away from the model, which is not good. But what about the format? What if the background was enlarged to a square format? Here is the same image in the square format. It is actually significantly larger than the image above, but has been compressed to fit into this blog post:

When looking at these two images side-by-side in their actual sizes, I like the larger square format better. I am hard pressed to say why. It has more of a stark and forlorn look to it, which I like. But there could be something deeper. The square format gives the composite a more environmental look, as though it was shot with a wide angle lens. We can more easily discern that she is alone. There is no smile on her face, and she is by herself in all that water, so perhaps we will subconsciously conclude the she is available. This subtle psychological nuance, contributed to by the square format, may be one of the reasons male viewers might like the image.

Here is a second square format image of Karina where a slight texture has been given to the water and sky. The texture contributes a dusty, gloomy appearance to the water and sky, and makes the model “pop” even more. This too contributes to the desolate look many find so compelling:

But if we do choose to extra elements to the composite, the square format allows plenty of room for those extra elements, and gives them breathing room. For example, here is the beautiful Melissa Stearnes imperiled by one of those pesky, relentless, approaching zombies:

The square format is perfect for showing a full-length model twice, as shown here in this silhouette image of Singapore’s Priscilla Martin:

Moreover, the square format allows for more than two versions of the model, as shown in these three images of Laurel Hervert:

Showing more of the environment can add mystery to the composite. Imagine if this image of Myka Frago was cropped so tightly that the viewer could not discern she was on the floor of an empty underground parking garage. It definitely would not be as compelling.

The same goes for this image of Natasha Loren. Also note that the square format allows you the place the model off center if you wish:

Like Natasha, model Loe Roe (seen below) is very much an Instagram person and rarely checks her Facebook page. Instagram will severely crop the top and bottom of an unmodified photo in portrait mode. It likes a 4:5 crop ratio, but has no problems with the square format.

I submit that the square format has many advantages and should be used more often.

Constructing an Evocative Fireworks Composite

For me, the best images are evocative in that they bring forth some emotional connection to something in our past that makes us feel good when we think about it. I loved setting off fireworks when I was a kid, so a fireworks composite seemed to be in order. Sometime ago I purchased a big box of fireworks with the intention of setting them off and photographing them. When model Loe Roe was visiting, I had the idea of putting the box on a hand truck and having her pose with it:

After masking out the white seamless background I began looking for a new background. I liked this image I had taken along the California coast between Monterrey and San Francisco:

The next step was to mask out the sky and ocean. Beneath this Photoshop layer I placed two images of fireworks I had taken in Paradise Valley, Arizona:

At this point, let’s show the final layer stack in Photoshop so we can follow it:

From here, you can see that I clipped a Curves Adjustment Layer to the Foreground layer of the grass to darken it, the added a Photo Filter set to Deep Emerald to make the grass a deeper green and give it more of a nighttime look. On a blank layer, a shadow was drawn where Loe was standing by using a soft black Brush set to 20% Opacity and 20% Flow:

Next I wanted to enhance the colors of the fireworks, so I added three Photo Filter Filters: Cooling (82), Red and Green. Black layer masks were used to hide these color enhancements on other parts of the image, and a white Brush was used on the black mask to reveal the color enhancement over the blue, red and green fireworks:

If an object is outdoors, its bottom will usually be darker than the top. A Curves layer clipped to the layer of Loe and the box of fireworks was used to achieve this effect by darkening them. A linear gradient, set to black to transparent, was used on the layer mask of this layer to gradually make the bottom darker. A second Curves layer, also clipped, was used to reduce those distracting highlights on the top of the box. Next a blank layer was used to paint on a thin shadow under the box and shoes, and slightly overlapping onto them, using the same Brush used to paint the shadows:

The Soft Pop action in the TK-8 panel was run. This slightly sharpened the image and warmed it, to give it that summertime look:

Finally, a Curves layer set to the Multiply blend mode was used to apply a custom vignette the image and draw our attention to the center of the image:

In my humble opinion, capturing an evocative image in camera is extremely rare. Composites are a way to transform an idea into an image that stirs your emotions. The look and feel of this image does that for me, although there was nobody like Loe around when I lit fireworks in my boyhood! Here is another composite from the same series:

The Modern, Revisionist Maxfield Parrish Look

My introduction the the influential American painter Maxfield Parrish (1870-1966) was from a premium video I watched called Channeling Maxfield Parrish by Doug Landreth of Photomorphis. Sadly, Photomorphis seems to be out of business. In this video, Doug showed the construction of a composite with a farm with a vineyard and billowing clouds.

There also exist videos on how to transform your own landscape image to give it the look of a Maxfield Parrish painting: https://youtu.be/ZBcISsGrG98. Here are some examples of Parrish’s paintings displaying the vibrant, saturated colors he is known for:

Daybreak, 1922

Grisalda, 1910.

Ecstasy, 1930.

Evening - Winterscape, 1953.

Enchanted Prince, 1934.

Morning, 1922.

My interest is not how to precisely reproduce the look of a Maxfield Parrish painting, but rather how to modernize or “revise” his look. His female models usually wore long, loose, white gowns. How would it look, for example, if I used an image of Loe Roe, a very popular model from Tempe, Arizona? This photograph was taken of her on the balustrade bordering my back porch.

Loe Roe, original portrait, Tucson, Arizona, 2022.

The Sonoran desert is not very Maxfield Parrish like, but what if an image of Vermillion Lakes, Banff National Park, Alberta, Canada, was substituted?

Vermillion Lakes, 2007.

Here is the resulting composite:

Loe Roe composite, 2022.

Although I really love this image, it could have been better if I had taken the original image of Loe in the shade and on an overcast day. The highlights on her dress are not motivated by the light on the background. The second thing I learned was that I could have made my job of masking the background away from her hair a lot easier if I had positioned a piece of gray cardboard behind her hair before I exposed the original portrait. Her body, dress, and the concrete were simple to select with Photoshop’s Pen Tool, although that was somewhat time consuming!

Here is a screenshot of the layers in Photoshop. I slightly blurred the top of the background, while leaving the bottom, closet to the model and balusters, sharp. The Soft Pop layer sharpens and warms the image slightly. I also color graded the image with Color Efex Pro 4 and the Dreamy preset in Intensity.

I chose not to take things further by giving the composite the look of an oil painting. Instead, opting for more of a photo-realistic modern look. That may come in a future post as I explore the possibilities.

Adding Grayscale to a Color Image for Added Interest

One of the primary goals of glamour and fashion photography is to make the viewer pause and focus on the model. One of my favorite ways to accomplish this is to mix black and white (grayscale) into the image. The viewer instantly pauses because something is profoundly different. But why don’t we see more images that are part grayscale and part color? The answer, I think, is that this effect requires some post processing, and traditionally involves selections and masks in Photoshop. Many photographers shy away from the prospect. However, this is changing as Photoshop makes it increasingly easier to make accurate selections with the click of one button. Let us consider this image of the magnificent Anne, who is not a model, taken in Tucson, Arizona:

No masks were used here, only the Select Subject command. The subject was put on her own layer, above the layer with the back and white conversion. With Anne being the only part of the image in color, there is less to distract the viewer. Consider another image of her, this one taken in Singapore:

This was processed in the same way as the first image. The similarity of these two images is that the walls behind Anne were both light in color. This is most helpful when using Select Subject, because if some of the wall behind her gets selected along with her hair, it is less noticeable. Select Subject has become quite good, but selecting curly hair remains a challenge.

In the above image of Tayler Vomacka, taken in Tucson, the white studio background was masked using a dedicated masking plugin. The background, from the botanical garden in Claremont, California, has a multitude of sharp, detailed foliage, which is a nice contrast to Tayler’s smooth, feminine curves.

Here again, the sharpness of the grayscale Joshua Tree leaves are a nice contrast to Tayler’s softness. Some colors, for some reason, work very well with grayscale backgrounds. Red is definitely one of those colors.

Grayscale backgrounds can also be applied to images shot in the studio, with the studio look maintained. Here, the gorgeous Chimi was photographed in Singapore on a white seamless background, which was masked out. In Photoshop, below the layer of Chimi, a solid black layer was placed, and below that, a solid white layer. A white layer mask was placed on the black layer, and a black Reflected Gradient was drawn on the mask to reveal the white layer below in the middle of the image. Again, red goes well with gray.

This image of Alexandra, from Singapore, is somewhat similar, but depicts a difference. Only the metallic catsuit and her eyes are in color. The rest of her, and the studio floor, are grayscale. Blue is another color that looks good when contrasted with grayscale. A white layer mask was attached to the black and white conversion. Select -> Color Range… was used to select the blue of the catsuit, then, on the mask, the selection was filled with black to reveal the color layer below.

While were are on the subject metallic catsuits photographed in Singapore, this image of Myka also demonstrates a conversion of the color of the catsuit from green to blue. See the blog post of December 30, 2021 to see how it was accomplished.

As shown by this image of Celestine of Singapore, this concept works nicely with headshots as well. Fortunately, she was wearing a nice dress with blue sequins that look great in contrast to to the gray tones.

This image of Natasha, taken in Tucson, shows a solid black background with results in a dramatic look. Using Select -> Color Range… makes quick work out of selecting the lacy part of her outfit so it can be masked from the black and white layer.

Jewelry, such as model Loe Roe’s earrings, can be included too. Another consideration is file size. Instead of creating a black and white pixel-bearing layer, a Black & White adjustment layer will result in a smaller file, thus saving disc space. Be sure to check the presets contained in its pull-down menu.

This final image, again of Anne, shows the variation where the model is totally grayscale, while the rest of the image remains in full color. Select Subject was once again used, and a Black & White adjustment layer, with the preset set to Green Filter, was clipped to the layer of Anne, resulting in her absence of color. The plain wall behind her, the sunglasses adding a little mystery by hiding her eyes and her strong, and the graphic pose all contribute to a nice look.

It is hoped the examples presented here will inspire you to try some of these techniques, and propel your creativity and photography in a new direction.

The TK Panels and Ideal Workspace

I mentioned the TK Panels in my previous blog post, created by Tony Kuyper here in Tucson, Arizona. Again, here is the link if you are interested in buying them: https://goodlight.us/panels-and-videos.html. Sean Bagshaw has excellent videos demonstrating the use of these that can be purchased here as well. Beyond that, Dave Kelly posts superb YouTube videos about the TK panels every Friday as part of his The Joy of Editing series. He has a discount code (DK15) for a nice discount when ordering from the above web page.

To a certain extent, these panels can be modified by the user. Their layout has been extremely well thought out. While Tony started out creating an easy way to generate luminosity masks, this latest version (TK8) has evolved to do so much more! I have incorporated TK8 into my everyday workspace, which looks like this:

The Multi-Mask panel on the right is where you generate various types of masks and implement their use. The Cx Panel to its left also comes in a shorter, wider version which has the same buttons, called the Combo Panel. On my 27” iMac, I have plenty of screen real estate on the left for the image.

These panels are priced very reasonably at only $29 and are highly recommended!

Soft Pop to Sharpen (and Warm) Portraits

The Soft Pop action is built into the TK8 panels, written by Tucson’s own Tony Kuyper. ($29, https://goodlight.us/panels-and-videos.html). It generates two smart objects: a Curves adjustment and a Gaussian Blur, so the action can be adjusted if desired. So far I have found no need to make any custom adjustments, so I rasterize the Soft Pop layer to diminish the file size (Layer -> Rasterize -> Smart Object). Here is a “before” version of an image of model Ziva Fey:

And the “after” version, with the Soft Pop action added:

I quite like the effect. It is sharper without being “crunchy” and definitely warmer.

Here is a “before” studio-style image of Laurel:

After the Soft Pop action is run, the image is both sharper and warmer, in a pleasing, subtle way:

If a cooler look is desired, you can mitigate the warmth by adding a Deep Blue Photo Filter set at its default 25% Density. The blue color cast can then be adjusted by lowering the opacity of the Photo Filter Layer. Here is how it looks at 100% opacity:

I rather like the light blue background, so I often leave the layer opacity of the Photo Filter layer at 100%.

Re-Coloring Metallic Fabric

Lots of interesting apparel is now available in shiny metallic fabric. A favorite of mine are metallic catsuits. The model must have a stunning figure, and the lighting must be controlled to avoid blown highlights, but the results are well worth the trouble. What follows is one method to change the color of a metallic catsuit. I always begin in the development module of Lightroom by taking both the Highlights and Whites sliders all the way to the left. This gives the metallic sheen of the fabric a nice glow. Here is how I changed a green catsuit to blue. Here is a beginning image of the gorgeous Shawna:

One easy way to effect the change would be to select the green color of the catsuit then use a Hue/Saturation adjustment layer to effect the change. But I wanted something more, so I will convert this to a black and white image, then re-color only the fabric. After making a selection of Shawna and putting her on her own layer, I added a 50% gray layer, in the Multiply blend mode, beneath her. A darker background “pops” the model more:

Next I did the black and white conversion. There are numerous ways to accomplish this, such as a Black & White adjustment layer in Photoshop. I used a legacy plugin from MacPhun called Tonality, specifically the Fashion Model 2 preset:

To begin the conversion to blue, a Deep Blue Photo Filter is used at its default 25% density setting, then a stamped layer created and another Tonality conversion made, this time using the bluish Cold Heart Preset. A mask of just the catsuit was added to this layer:

A second Photo Filter, Cooling (82), also set at its default 25% Density, was clipped to this black and white layer. Its mask constricts the blue color to just the catsuit:

Finally, a vignette was applied using a Curves adjustment layer set to the Multiply blend mode and the layer opacity reduced to 40%. Notice the mask which allows darkening across the entire image except on Shawna’s face:

Here is a screenshot of the layer stack:

Of course other colors are possible as well. Here is a conversion from the same green catsuit to red. Notice the Soft Pop action from the TK-8 panel (discussed in the previous blog post) was used:

Here is a screenshot of the layer stack:

Endless variations are possible. Here is an image of Nikki where a different blue was achieved by use of the Cooling LBB Photo Filter:

The layer stack for this image:

Adding "Pop" Via Blurring the Background

At the urging of my friend on Instagram, fashion photographer Douglas Bassett of Milan, Italy, I have been thinking about how blurred backgrounds can enhance glamour portraits. It has long been known that photographing a model at a wide open aperture while on location gives you a blurred background. This blur helps to eliminate distracting background elements and focus the viewer’s attention on the model. But there are concerns.

Lenses capable of wide apertures such as f/1.2 or f/1.4, also known as “fast” lenses, are considerably larger, heavier and more expensive than “slower” lenses, such as f/2.8. Moreover, when shooting at f/1.2 or f/1.4, it is likely only one eye of the model will be in sharp focus; the remainder will be soft. If the model moves slightly after focus is established, everything may be soft. Setting the aperture to something like f/2.8 will get more of the model in focus, but the background will be less blurred. What follows is one way of dealing with this compromise by blurring the background in post. Here is a retouched image of model Low Roe, taken with an f/1.4 85mm lens set at f/2.8:

Of course we cannot simply blur the image without blurring our model, so we must mask out the background and put her on her own layer in Photoshop. Here is the result if we give the original layer a 100-pixel Gaussian blur:

It can readily be seen that the blur smears the underlying image of Loe onto the background, resulting in a halo around her. So let’s delete her from the background and try blurring once more:

And now with a 100-pixel Gaussian blur applied, with the masked layer of the model overlaid:

The blank area in the middle of the image has been stretched onto the background, resulting in a light halo. So, instead let us try a Content Aware fill of that blank central area, clean up the edges with the Patch, Spot Healing and Clone Stamp tools, then applying the blur. Here is the result with the masked layer of our model not overlaid:

Finally, success! Next, add a Curves adjustment layer over the background layer, and pull the midpoint of the curve down to darken:

Next, let’s add two of Photoshop’s Photo Filters, both at their default settings. First, Deep Blue, then over that Cooling (82). These two filters work well together. I really like backgrounds rich in deep blues and greens. Here is where we are at with the masked layer of Loe overlaid:

There is one final touch we can apply: Soft Pop. This is an action contained in the TK-8 panels written by the amazing Tony Kuyper of Tucson, mentioned in my previous blog post. This set of Photoshop panels is truly extraordinary, and a bargain for only $29. They can be purchased here: https://goodlight.us/panels-and-videos.html. Dave Kelly has a 15% off code of DK15. Be sure to check out his videos on YouTube. He publishes a TK-8 video every Friday. These panels are written primarily for landscape photography, but I am having a lot of fun learning how to apply them to the type of work I do. Here is the above image with Soft Pop applied at its default settings:

Soft Pop results in a warmer image. If it is too warm, you can apply a Deep Blue Photo Filter over the top. This was not done here. Soft Pop creates two Smart Objects. I typically do not adjust them, but instead rasterize the Soft Pop layer (Layer -> Rasterize -> Smart Object) to keep the file size smaller.

I hope you will give this technique a try. Here is a second image of Loe that was processed in similar fashion:

Processing Blown-Out Backgrounds

In glamor photography, we typically try to darken the background to drive attention to the model. This can be done by a number of ways, most often by vignetting at the very end of the post-processing workflow. However, how can we deal with a background that is so bright that various darkening techniques are ineffective? This is exactly what I had to deal with here, in this image of the remarkable Tiffany Felisha:

The sky is totally gone and the blooming palo verde foliage is pretty ugly. So why not straighten the image in Lightroom, add ⅓ stop of Exposure to her face with a Brush, then click the Mask button, choose Select Subject, then click on the Invert button, and use the Exposure slider to lighten the background even further?

Much better, I think. I wanted to leave just a little detail in the background for added interest, or else it might start to look like a studio shot. Minor retouching in Photoshop brought the image to this point:

After this, I only did one thing: I ran the Soft Pop action from Tony Kuyper’s TK-8 CX panel. This indeed does make the image “pop” a bit, and warms it also. Here is the final image:

This action creates its own layer and two smart filters. I always rasterize the Soft Pop layer to save disk space (Layer -> Rasterize -> Smart Obect). Tony Kuyper is a resident of Tucson, Arizona, and the TK-8 panels are a bargain at $29 for the set. Go to https://goodlight.us for details. In addition, Sean Bagshaw has a set of superb videos detailing their use. On YouTube, Dave Kelly releases a video every Friday, as part of his The Joy of Editing with Dave Kelly series, on the implementation of these panels. These videos are excellent. Moreover, he offers a 15% discount code when you purchase these panels.

Concepts from Classic Horror Films

Since I had already done a gypsy fortune teller shoot, I decided to do a shoot based on a famous scene from the 1941 classic horror film The Wolfman, starring Lon Chaney Jr. and Bela Lugosi. If a person is bitten by a werewolf and survives, and if the wolfsbane is in bloom, and if there is a bright autumn moon, then that person will transform into a murderous half human-half wolf, returning to human form at dawn.

In the memorable scene I wanted to recreate, the great Bela Lugosi plays a gypsy fortune teller, but he is also a werewolf. A young lady enters his tent to have her fortune read. As she extends her hand, the fortune teller recoils in horror, for he sees a pentagram, the sign of the werewolf, on her palm. The pentagram can only be seen by a werewolf (and the audience), and it designates the werewolf’s next victim. Sure enough, he murders the young lady that very night.

We changed things around a bit by having our fortune teller/werewolf female, only because we had costumes for lady fortune tellers. The next victim in our scene would be male.

Photograph1.jpg

It was a fun takeoff from the typical fortune teller theme discussed in a previous blog post. Better yet, it was a true group effort. My friend Patricia again played the fortune teller, and Engel was her victim. Anne assisted in Patricia’s wardrobe styling, and photographers Benjamin and Serena assisted with the lighting and composition. Serena drew that excellent pentagram on Engel’s right hand.

The Dreamy with Detail Background

Portrait photographers know the benefits of an expensive lens with a wide aperture: it allows them to blur the distracting background behind the subject, thereby giving more emphasis to the subject. This is because the human eye is naturally drawn to sharpness and contrast. A similar effect can easily be achieved in composite work by simply blurring the background with a Gaussian Blur in Photoshop. What is presented here is a very simple way to blur the background while preserving a lot of its detail. This image of Melissa is a good example of this look:

Photograph01.jpg

Here is the original background, from the Raffles Hotel, Singapore, where the Singapore Sling originated. It is an HDR image taken with a 45mm Nikon PC-E (tilt-shift) lens to keep the vertical architectural lines true.

Photograph02.jpg

I typically like to add a Curves adjustment layer to darken the image. As you can see, I pulled both the right highlight end of the curve and the midpoint down a little.

Screen-Shot01.png
Photograph03.jpg

I added to the dreamy look by adding two Photo Filters. First was a Deep Blue Photo Filter with the Density set at 38%. This was followed with a Violet Photo Filter with its Density left at the default 25%.

Photograph04.jpg

A separate stamped layer of the background is then created that looks like the above image. It is then given a somewhat over-blurred Gaussian blue of roughly 160 pixels.

Photograph05.jpg

The layer opacity of this blurred layer is simply reduced to 75% to achieve the Dreamy with Detail look.

Photograph06.jpg

When an image of Melissa with the studio background masked out is added, we can readily see she is too warm for this cool background.

Photograph07.jpg

To fix this, I first want to darken her a little. Since she is outside, her legs should be a little darker than her face. A Curves adjustment layer, clipped to her layer, with the midpoint of the curve pulled down slightly, was created. A black-to-transparent linear gradient was drawn on the mask of this Curves layer to block the darkening from the upper part of her. A second Curves layer, also clipped to her and with the midpoint pulled down a little, was used to darken Melissa overall.

Photograph08.jpg

The final blending of Melissa with the background was done with Nik Color Efex Pro 4. Darken/Lighten Center, Brilliance/Warmth, White Neutralizer and Pro Contrast filters were used to cool down the color on her. They also cooled the background further while blending the subject and background together nicely, as can be seen in the first image of this series. Here is the layer stack.

Screen-Shot02.png

Here is another example with Lauren at Keppel Bay, Singapore. It was made in like fashion to the preceding image. A Deep Blue Photo Filter, set at 25%, plus a Cooling (82) Photo Filer set at 25% were used. These two filers work very well together with render a blue color cast on an image.The Blur layer’s opacity was set at 80%..

Photograph09.jpg

The final example is of Krystal. In this example, only a single Photo Filter was used, Magenta set at the default 25%. The Blur layer opacity was reduced to 75%.

Photograph10.jpg

I hope you will find some use for this simple technique. It enables the model to “pop” while at the same time it gives the viewer a good idea of the environment where the model is located.

The Woes of Cropping

Cropping a portrait where the model is not photographed in front of a seamless studio background is usually a compromise. To understand what I mean, let us begin with this uncropped image of model Lauren, who goes by the nickname of Loe, but whom I call The Temptress of Tempe:

Uncropped.

Uncropped.

It is easy to see where she is: in a bathroom, sitting on the edge of a bathtub with two windows behind her, most likely at sunset. However, there are a couple of concerns. First, she is somewhat small when compared to the size of the entire image. Second, the dark marble tile at the bottom of the image is not interesting, and seems unnecessary. So, if we crop the bottom of the image, we get this:

Starting image with the bottom cropped away.

Starting image with the bottom cropped away.

This crop rids the image of that dark bottom, but now her head is near the center of the frame, yielding a somewhat unpleasing composition. Let’s crop it once more, this time using a standard 4:5 ratio:

A crop to the ratio 4:5.

A crop to the ratio 4:5.

Better… in fact, this is one of my favorite compromises. Loe looms fairly dominant in the frame, but the blue and orange in the windows can still be seen. If we crop to a square format, we get this:

A square crop.

A square crop.

The square crop looks pretty good to me as well. Loe is again dominant in the frame, and we can still see the color in the windows. An even tighter square crops takes us to this:

A tight square crop.

A tight square crop.

Now we can really appreciate Loe’s beauty, but the blue window light on the right has been totally lost.

In my old age, I am learning to NOT attempt to fill the frame with the model in camera. It is best to leave some room around the model so that we can have more cropping options later. This is what many photographers might call an “environmental portrait” because it gives the viewer an idea of where the model is. In so doing, try to resist the urge of using a wide angle lens for fear of getting the dreaded “dog nose distortion.” Instead, use a telephoto lens whenever possible and position the camera as far away from your model as is practical.

For those interested in working with Loe, her email is Bookingloeroe@gmail.com. She is based in Tempe, Arizona (she will travel), and is reliable, easy to work with, very creative, and gorgeous!

Ruminations on a Gypsy Fortune Teller Shoot

I have long wanted to do a photo of a gypsy fortune teller, at night in her tent behind a glowing crystal ball. Years ago I had ordered a crystal ball from Amazon, and this shoot would provide yet another opportunity to use it. Three gypsy costumes were ordered from Aliexpress. My friend Benjamin offered to assist with the lighting, and my friend Anne offered to help with the wardrobe styling. Patricia had the look of a gypsy, and was anxious to shoot. Here is the final result:

Final image.

Final image.

The primary question is how to light such a scene. I wanted a mysterious, dark look to the image, but I also wanted the items on the table to have enough light cast on them to reveal color and detail. Is it better to light the set normally and then darken parts of it, or underexpose the set and then brighten parts of it? It is far easier, I think, to underexpose the set. Here is the starting point:

The RAW file with no adjustments.

The RAW file with no adjustments.

Three flashes were used. Two strip lights, one on either side of the model, plus an octabox in front. A 70-200mm lens was used, set at 135mm. The exposure was 1/200th second, f/14 at ISO 100. Several adjustments were made in Lightroom to bring the image to this point:

The result of a few adjustments in Lightroom.

The result of a few adjustments in Lightroom.

Notice the blue look? That was not achieved by changing the Color Temperature slider because that would change the entire image. Instead, I wanted blue primarily in the shadows, so I went to the Split Toning adjustments and adjusted only the Shadows sliders to get the bluish look, and left the Highlights sliders alone:

Screen Shot01.png

I knew that the crystal ball and lantern on the right would be casting light on the table, so the Elliptical adjustment tool was used to draw a couple of ovals on the table, and the Exposure slider adjusted to lighten those areas. Likewise, Patricia’s face, chin and hands were lightened with the same tool. I first tried using the Brush to lighten her palms, but the Elliptical tool proved far superior.

Once inside Photoshop, the image was cleaned up a bit. The big fold on the blanket above the model was eliminated and her eyes were brightened. Now, let’s go through the items on the table, and how they were retouched. Don’t let the look of the final layer stack scare you:

The final layer stack in Photoshop.

The final layer stack in Photoshop.

The flames for the four candles came from separate files. They were positioned and re-sized as needed. The glow for each lantern was created by using the Radial Gradient tool and making a white ellipse, which was then colorized with a Hue/Saturation adjustment layer with the Colorize box checked. Another smaller white radial gradient was placed on top. The layer blending model for both radial gradient layers was set to Linear Dodge. The glow on the crystal ball was achieved in a similar fashion.

Aside from two Curves adjustment layers for the knife, the objects on the table were handled by Hue/Saturation adjustment layers, again with the Colorize box checked. A selection was made of an item, usually with the Pen Tool, and the adjustment layer added. Because the Colorize box was checked, I could change the original color to something altogether different if I wanted.

A stamped layer was sent back to the Camera Raw Filter. A tiny amount of Vignette was added, plus a little Grain, Clarity and Texture. Another stamped layer was created and sent to Nik Color Efex Pro 4, where Darken/Lighten Center (to further add a subtle vignette), Brilliance/Warmth (to slightly warm), Skylight Filter (to again warm) and Pro Contrast were used to further refine the image.

Sharpen AI from Topaz was used on another stamped layer. I really like this plugin, but when I let it automatically choose the method to sharpen it chose Stabilize. Since my camera was on a tripod, I chose Sharpen instead, and the result looked much better.

A Curves adjustment layer was used to darken the entire image ever so lightly. The final adjustment was a stamped layer sent to Luminar 4. The Glow and Orton Effect were used, and nothing else.

In summary, this was a fun project. Yes, all the selections took some time. But because the camera was on a tripod, those selections can easily be used on other images in the series. Here is another version with Patricia in a different costume:

Patricia as a Celtic gypsy.

Patricia as a Celtic gypsy.

Popping the Model from the Background

As an example, I selected the image of Janellie, shot against a white seamless background:

Original image of Janellie with white seamless.

Original image of Janellie with white seamless.

In choosing a background for the composite, I often like to use an image which has similar color theme. Since her dress is predominately yellow, I choose am image with a lot of yellow. This gives the viewer of the final result the notion that a professional art director was involved in the shoot, and much thought was given to choosing the ideal shooting location. This image was taken in the beautiful Phoenix Hotel in Yogyakarta, Indonesia:

Phoenix Hotel, Yogyakarta, Indonesia,

Phoenix Hotel, Yogyakarta, Indonesia,

I positioned the masked version of the model over the above image, which was enlarged and positioned, then the entire image was cropped to a square format:

Janellie looking as though she is in Indonesia.

Janellie looking as though she is in Indonesia.

As you can see, the background is quite busy. There are simply too many places for the viewer’s eye to visit. The solution was to convert the background layer into a smart object, the give it a 24.7 pixel Gaussian blur. This creates the same look that would be obtained by capturing the model in the hotel with a lens set to a very wide aperture. Now the model really “pops!”

Final image.

Final image.

A few comments about the layer stack in Photoshop are in order. Here are the layers on the completed image:

Screen Shot 2020-06-24 at 9.27.39 AM.jpg

Notice that the original, but retouched image of the model is at the bottom of the layer stack, and the image of the hotel directly above it is set to the Multiply blend mode. Doing this will often recover some of the model’s hair that was lost in the masking process. Also note that the background part of the image was darkened somewhat by the use of a Curves adjustment layer. This makes the model (in the top layer) look brighter than the background, thereby separating her even further.

Instagram: Please Fix the Problem of the Followers Game

If you have an Instagram account and are like Scott Kelby, you too will have over 119,700 followers while you yourself are following only 157 people. All this would be accomplished with only 510 posts. He can achieve these numbers because he is a prolific and best-selling author, a fine photographer, and gives great seminars throughout North America and Europe. He's a famous, friendly and engaging guy who travels a lot, and posts beautiful images from his travels.

Unfortunately, many Instagram users try to achieve a similar high followers to following ratio in an attempt to make themselves look good. This is accomplished by following everyone they can, and they may even leave a nice comment about your work. If you do not follow them back they will unfollow you. If you do follow them back they will still unfollow you.

If a person follows me, I politely follow them back, but a high percentage will then unfollow me. To track this, there is a free app called Followers. I installed it and check it once a day for people who have unfollowed me. These are nearly always people who have a lot of followers, yet they themselves follow very few people. With a single click you can unfollow them back. This is a somewhat ridiculous waste of time, but is the only way for me to be sure I am following those who are likewise genuinely interested in my work.

Back when Instagram was in its infancy, having a lot of followers surely was a sign that you were important and/or great at your craft. Today, however, it is absolutely meaningless. LOTS of people are playing the Followers Game on Instagram. Plus, you can even buy followers! How big a problem is this? BIG! Since I reset the Followers app, I have gained 75 new followers, but 79 have unfollowed me.

So, if you get a follow from a guy named Joe Doaks, he has 8,200 followers, is following only 200 people, and you have never heard of him, you can be certain he is playing the Followers Game. It is a silly waste of time. Why? Because the followers/follow ratio is meaningless. What is meaningful is how many likes and comments you get. In other words, how well you can engage viewers, which is a reflection of quality content.

Hopefully Instagram will remedy this problem by simply deleting the Followers and Following counts from everyone's home page. Those numbers are now meaningless, and signify nothing. We should be interested in content - not followers/following ratios. Such an upgrade to their app would put a stop to this game, and thereby improve Instagram for everybody.

A Unique Semi-Black and White Glamour Composite

Presented here is a unique method for creating a black and white glamour composite that contains just a little color. Three MacPhun (Mac only - sorry!) Photoshop plugins are used. For this look, I chose this image of Andrelica, photographed in a studio environment on grey seamless paper. Please note the shadows around her feet:

The background is an HDR image taken at Clarke Quay, Singapore. Notice that the vertical lines are vertical, This is because it was photographed with a Nikon PC-E (perspective control - otherwise known as a tilt-shift) lens:

This technique uses two versions of this background, so duplicate it to get two identical copies. On one copy, run Intensify CK's Gloomy Day preset at full strength to achieve a "punchier" version:

Andrelica03.jpg

On the second copy, run Tonality CK and use your favorite preset. I used Barcelona from the Architecture presets:

Andrelica04.jpg

Naturally, the grey studio background must be masked away from the model. After creating this mask, I save the file as version one. On version two, combine the layers used to retouch the model by selecting them and hitting Command-E, resulting in two layers: the layer just created containing all the retouching, and the layer of the model with the background masked out. Insert the black and white version of the background in between these two layers:

Andrelica05.jpg

The bottom, unmasked layer of Andrelica, with the shadows, is completely hidden. Simply change the layer blending mode of the black and white background layer to Overlay to get back those shadows around her feet:

Andrelica06.jpg

Here is how the layers stack should look:

Screen Shot1.png

Shift drag the color version of the background over the black and white version. In Normal blending mode it will completely hide the black and white layer and the shadows. Set the layer opacity to 30% to drain most of the color, and also set the layer blending mode to Overlay. It should look something like this:

Andrelica07.jpg

At this point, you should be able to see where we are going with this technique. We now have a punchy, background drained of some of its color, whereas the model is still in full color. Next, create a stamped layer at the top of the layer stack (Option-Merge Visible from the Layer Panel's flyout menu). I used the Luminar plugin and its Noble Beauty preset to achieve this nice black and white look:

Andrelica09.jpg

To get that desired semi-black and white effect, simply reduce the layer opacity of the top layer. I went to 70% (= 30% color) but this is simply a matter of taste. Realize that when doing this, you are adding much less color to the background than to the model, because a lot of the color has already been drained from the background. Here is how the layer stack should look when done:

Screen Shot2.png

The final image:

Andrelica10.jpg

Masking Semi-Transparent Fabric

Here is another image of Laurel. I want to mask out the white seamless studio background so I can composite her into another scene. The tricky part will be to mask the background showing through the fabric at the bottom, middle part of her dress:

Topaz ReMask 5 has the proper tool to accomplish this task. Start by creating a "trimap" where you paint a blue line over the edges where you want ReMask to compute the edges of the mask, and fill the areas to be masked out with red. Notice that I have done nothing to the area of semi-transparent fabric in her dress:

Screen Shot 2017-02-15 at 1.06.09 PM.png

After I pressed the Compute button, ReMask created a nice mask. To mask the area of transparent fabric, we need to see it better. I choose a split screen view, and changed the version on the right to "Keep" and set the background color to red to better show any unwanted white background:

Screen Shot 2017-02-15 at 1.07.43 PM.png

Select the Transparency Brush. It used to be called the Dual Color Selection Tool because you choose two colors: what to keep and what to mask out. You are presented with a green eyedropper. What you do is on the right version of the image you click on a solid black area of the dress, in this case the fabric over her leg. This represents the color in the transparent area you want to keep. The eyedropper changes to red. Click on an area of solid white background. I clicked on the area of white seamless just under her dress, between the legs. This represents the color you want masked from the transparent area. The red eyedropper changes to a brush. Paint on the area of transparent fabric. You will now see a little black showing through on the left (Mask) version, and a little red showing through on the right (Keep) version:

Screen Shot 2017-02-15 at 1.12.22 PM.png

Back in Photoshop, Option (Alt)-click on the layer mask to see it in black and white. Paint over the hard edges (with no hair) of the mask with a white brush set to Overlay. Stay away from the transparent fabric:

Screen Shot 2017-02-15 at 1.16.05 PM.png

Here is a closeup of the area of transparent fabric created by ReMask:

Screen Shot 2017-02-15 at 1.19.25 PM.png

It is masked, but the mask is somewhat weak. Set the foreground color to black, and with the brush still set to Overlay, brush over the area of transparent fabric to get this:

Screen Shot 2017-02-15 at 1.20.08 PM.png

Brush over the same area again:

Screen Shot 2017-02-15 at 1.20.35 PM.png

Brush over the area a third time:

Screen Shot 2017-02-15 at 1.20.55 PM.png

Now that the transparent area of the mask is stronger, we should be able to actually see through the fabric a little. Here is the final image, using a scene from Singapore's Little India district:

Laurel2.jpg

Two Easy Tricks to Regain Masked Hair in Photoshop

Here is a photograph of Laurel, taken against a white seamless background in a studio environment:

I wanted to replace the white background with this spotlight background:

GraphicLoot Spotlight Background.jpg

To do so, I of course needed to mask out the white background which I did with ReMask from Topaz. The Layers Panel in Photoshop looks like this:

Screen Shot 2017-01-03 at 7.27.05 AM.png

Next I placed the spotlight background under the Extract layer, transformed it to fit the frame, then I combined the bottom three layers into one layer. For trick #2 to work, this Background layer of the retouched model at the bottom of the Layers Panel is needed. The Layers Panel now looks like this:

Screen Shot 2017-01-03 at 7.33.36 AM.png

Here is what the image looks like at this point:

Laurel2.jpg

Some of Laurel's beautiful hair is definitely missing. Here is trick #1 to get it back: Make sure the top layer is active, and click on the layer thumbnail to select it. Go to Layer -> Matting -> Color Decontaminate... and drag the Amount: slider to 100%. Here is the result of this one move:

Laurel3.jpg

This looks better already. But wait, there is more - trick #2: Select the background layer (in this case the Spotlight Background layer) and set its layer blend mode to Multiply. Now there is even more hair:

Laurel4.jpg

I hope these two simple tricks will make you look like a masking guru!

How to Mix Glamour and Water - Part 10

In this final instalment of the series, some ideas for different effects will be shown. First, you can texturize the background to get the effect of an old Maxfield Parrish painting. A texture from Photomorphis was used here:

Model: Andrelica.

Model: Andrelica.

Splashes in the water can also be added. The Image Manipulation Store sells a series of these on a black background, which disappears when you set the layer blend mode to Screen:

Model: Laurel Hervert.

Model: Laurel Hervert.

Here's is another example of a splash with rougher water:

Model: Marketa Havlinova.

Model: Marketa Havlinova.

The "flinging hair" pose, though commonly seen, really shows off the model's hair:

Model: Jatoya Gallo-Stubblefield.

Model: Jatoya Gallo-Stubblefield.

The model can pose with lanterns, crystal balls and such, with can be "lit up" in post processing. Note that I painted in some mountains on the horizon using a black Brush. This should be done before generating the water:

Model: Marcela Zuniga.

Model: Marcela Zuniga.

Model: Marcela Zuniga.

Model: Marcela Zuniga.

Another possibility is putting a second version of the model inside a jar. See my blog posts dated March 16 and April 2, 2014, Putting the Model Inside A Jar - Parts 1 & 2, for information on how to do this:

Model: Sara Bartfai.

Model: Sara Bartfai.

Model: Sara Bartfai.

Model: Sara Bartfai.

If lightning is added, add it before the water is generated so you can get a reflection in the water:

Model: Kaela Isa.

Model: Kaela Isa.

Water can be added to scenes such as this:

Rancho Santa Ana Botanical Gardens, Claremont, California.

Rancho Santa Ana Botanical Gardens, Claremont, California.

Simply add water as high in the scene as you want, then mask out the parts where you do not want water:

Rancho Santa Ana Botanical Gardens, Claremont, California.

Rancho Santa Ana Botanical Gardens, Claremont, California.

Model: Andrelica.

Model: Andrelica.

There is no rule that says the color of the water must match the color of the sky:

Model: Simone Knight.

Model: Simone Knight.

Try experimenting with rougher, more turbulent water. Do this by hitting the Complex button in Flood 2 and playing with the sliders:

Model: Tiffany Felisha.

Model: Tiffany Felisha.

Simulating a full moon backlighting the model at night is another interesting look:

Model: Kaela Isa.

Model: Kaela Isa.

There is no rule that says the model must wear a swimsuit, and there is no rule that says she must be in or near water. Experiment and have fun!

Model: Sara Bartfai.

Model: Sara Bartfai.

I hope these examples stimulate your creativity, and you have a lot of fun exploring this most interesting technique.