A Guide to Create Eye Catching Composite Images - Digital Photography School

A Guide to Create Eye Catching Composite Images

A Guest Post by Tom Di Maggio from Tom Di Maggio Photography.

Compo_StreetFight.jpg

My use of composite pictures has been drastically increasing in the last few months. It started with locations that got cancelled due to unavailability, then bad weather made it impossible to shoot at certain location and as of lately people came to me asking specifically for composite photographs. And to be honest I really enjoy doing them, it gives my imagination a lot more freedom, and there is certain flexibility that I really enjoy.

As with the preparation of any shoot, the most difficult part is finding the right setting. In this case the right background. I try to shoot them myself whenever I can but I tend to rely on stock photography as well. Basically I usually split a composite into three main parts: the subject or model, the background and the post processing.

The Subject

These are the images as they were captured.

The base for a good composite is obviously a correctly exposed subject. I adapt that light to the theme or mood that I want to convey in the final picture. It is really important that you have your final image in mind at all stages of the creation of a composite. It will dictate the lighting scheme you will have to use and at which angle you will need to photograph your model.

Which lighting setup to use usually varies from one to four lights? It really depends on the expected result. No matter which setup you choose lighting position and exposure are key factors here, they need to be spot on or you will spend a great amount of time in Photoshop trying to correct this. At this stage I try not to build too much contrast, so I like the light to be on the soft side.

I can only stress the fact that even though I will be using Photoshop to a certain extent, I will always want to work with a correctly exposed subject. Exposition and position are crucial in every kind of photography. Something you can solve in a few seconds during the shoot might take hours to correct in Photoshop. I don’t have any preference as to the background from which I will extract the subject from. As long as there is enough contrast between the subject and the background it won’t be a problem.

The Background

Original Background

Even when I shoot on location I really put a lot of effort into the choice of the background. The background is what gives the picture a context; it is an integral part of the story the picture is supposed to tell. You can photograph a very beautiful model in front of a dumpster; it will have a lot less impact as if you photograph the same model on a beach or in a nice hotel room. It’s the same when working with composites. If I have the chance to photograph the background for my composite myself I have a lot more flexibility in terms of angles.

A bought background gives you very little leeway as to perspective in Photoshop. When I photograph the background myself I usually have the idea for the subject shoot already in mind. Should I find myself in a situation where I see a nice background and I have my camera with me, I shoot it three times: ground height, belly height and chest height. I resort to stock photography when a shoot is booked and I don’t have the possibility to photograph the background myself.

As I mentioned earlier in the article angles is one of the key elements here. There’s no point in choosing a background that was shot at a dramatically different angle that your subject. It will look fake. I give my backgrounds a light HDR touch but I actually don’t use any HDR technique. I apply tonal adjustments and basically the same treatment as for the subject. It will nicely blend the subject and the background together.

The Post Processing

Like most of us out there I didn’t invent the wheel when it comes to post processing. All of my techniques are a mixture of things I picked up in books, on the web, trainings and so on.

To start with I shoot everything in RAW; the techniques I will explain below will suppose that you do the same. To be honest, and this cannot be said enough, with the pricing on memory cards nowadays there is absolutely no reason not to shoot in RAW. The advantages of doing so are just to enormous to ignore.

I usually start in Lightroom by taking out as much contrast from the picture as I can. The way I work I will build up the needed contrast during the post processing in Photoshop. Why do I remove contrast only to add some back later? It’s all about control.

I basically use a combination of two techniques that include Blur, High Pass filters and the Apply Image function. This allows me to add contrast and sharpen my images to the pixel and thus I can use it to create a clean smooth skin or a grungy style image with one workflow only. (see the portrait of Bora) I spend as much time here as I need to until the image is 100% the way I want it to.

Any imperfection here will be amplified once I put the Color Efex Pro layer on top of it. I created a few Color Efex Pro recipes that I use as a starting point on my images. I will then tweak them slightly to fit the image. I will intentionally push this process way over the top in order to give me more flexibility on the final image. Once I am happy with the result I save it as a layer on top of my image and then adjust the opacity of that layer accordingly.

Below the two Layers after my Color Efex treatment. You can see that the effect is way to strong here, but a slight adjustment on the opacity solves this. And it’s always easier than to go back into Color Fx and redo the effect until it fits.

At this stage I proceed with the extraction of the subject from the background. Personally I think that there is no perfect solution for this. I use whatever tool fits the need. It ranges from the refine edge tool to the smudge tool and even a normal brush.When I am done with the extract I drag the subject onto the prepared background and tweak both again so that the color tones and temperature match each other.

This portrait of Bora was processed with the same technique. All I did differently, was to apply different opacities to the correct layers.

Feel free to get in touch if you have any further questions regarding the subject. Below the two Layers after my Color Efex treatment. You can see that the effect is way to strong here, but a slight adjustment on the opacity solves this. And it’s always easier than to go back into Color Fx and redo the effect until it fits.

See more of Tom Di Maggio’s work at Tom Di Maggio Photography, InFocus Photography and on his Flickr Account.

Read more from our Post Production category.

Guest Contributor This post was written by a guest contributor to DPS. Please see their details in the post above.

Become a Contributor: Check out Write for DPS page for details about how YOU can share your photography tips with the DPS community.

  • Salomanuel

    awesome shot!
    but it would be awesome if you could make a step by step tutorial, so I finally can understand how you end with that kind of images!
    I’m not a photoshop newbie, but I’m not at that level

  • constance lowery

    You did not tell me how you extracted your model or how you got them onto your background.

  • http://digitalcameraadventures.blogspot.com/ Leopoldo “Leo” García

    TOM: Great work, thinking ahead makes the difference, and in a digital world, this is the key to separete oneself from the crowd. Congratutalions. I am saving this article. Way to go, NEED TO LEARN.

    “IF YOU DARE TO TEACH, YOU BETTER NEVER STOP LEARNING”

    Articles like this, are inspiring. Today, one needs to learn all tools of the trade as if one knows the pos-production tools and what can be done with them, you will be able to take your photography a notch higher.

    Great reads for a starting weekend.

    Leopoldo García
    Venezuela, South America

  • http://www.Kasparphotos.com Judy Kaspar

    Agreed! A step by step would have been appreciated instead of hinting at the techniques used. There’s no way to know how any of this works as is. Plus using Color Efex Pro, which is an awesome tool, we have no idea which filters were used and how.

  • http://www.tomdimaggio.com Tom

    @Leopoldo: Thank you, .. my thoughts exactly :)

    @Constance: I usually start with a very crude selection using the wand tool. I then use the refine edge option but only for the hair. I don’t do anything else there. Once I have my mask in place I start doing the fine tuning of the mask using the Brush tool and the much underestimate Smudge tool. I work with those two tools until I am happy with the result. I might add that I check my mask against white and black backgrounds before I put them onto the final background. Once there I fine tune the mask so that it fits perfectly.

    @Judy, Salomanuel: The details of the post-processing would be another tutorial, but here are some quick details as to the post processing I use.

    The Lightroom part:

    I decrease the contrast as much as I can without loosing detail in the picture. It will obviousely look flat but fear not, contrast will be build during the Photoshop and the NIK part of the processing. I have attached a picture of the model I’ve used in this composite so you can see how flat the image actually is. It doesn’t look natural but once the B&W is on top of it, it will look MUCH better ;)

    [eimg url='http://www.tomdimaggio.com/pics/Composite_Flat.png' title='Composite_Flat.png']

    For the Photoshop part:

    I create two layers in photoshop with the base image. The top one is converted to B&W using channels. The B&W conversion is different for each picture, I adapt it so that I get a maximum of detail out of it. I then change the blending mode to Overlay, depending on the end result I want, smooth or grungie I will increase or decrease the contrast on the B&W Layer. I then merge the two Layers to a new one and duplicate that one straight away. I use a Gaussian Blur Filter on the bottom layer with a value of 22 and then use “Apply Image” on the top Layer. In the apply image menu you’ll have to choose the blurred layer as a base and a blending mode of Subtract with a value of 2 and 256. Click ok and change the blending mode of the top layer to Linear Light. Group the two last Layers together. No I start using the Healing Tool on the top Layer to clean the skin of the subject. This method will not affect the color tones but only the imperfections in the skin. If I want a grunge looking picture I don’t use this too much. For a smoother skin I spend a lot of time on this.
    For the color tone I simply use the Clone Stamp tool with different blending modes (Lighten or Darken) at a approx. 20% Opacity. These are the steps I use on every single picture I take. They are quite similar, they only differ in the amount used and opacity.

    For the NIK software part:

    This is a bit trickier. I use only three Filters, Detail Extractor, Tonal Adjustments and Bleach Bypass. But I apply these three Filters differently on each picture. It depends on how much contrast I have in the picture, how many dark areas and/or how many bright ones I have. I am basically playing around with the settings until I am happy. I know it’s quite tough in the beginning but with time you will know what works and what not, and get to the wanted result faster. I usually end up on the heavy side of things here, but decreasing the opacity of the NIK Layer is a clean and simply solution to that.

    Once this is done, the last step is to match the subject with the background in terms of color and depth of field. Once again there is no fixed recepi for this as it is different for each and every picture. I usually start this process with color filters in Photoshop and finish them with Tone Curve, Color and Split Tone adjustments in Lightroom.

  • Salomanuel

    Thanks Tom.
    Thanks for listening to the comments feedback and for spending some time to answer properly!
    THIS is what being professional is all about!
    as soon as I will come back from holidays I will try to do some on my own!

  • http://penelopesoasis.com Penelope (NYC Blogger)

    This is awesome, I wasn’t even aware of this technique with layers (I only recently got my first dslr, but this opens up a lot of options and ideas for me!)

  • John

    hi, fantastic response. I was following the steps with the image you provide but when I reach the “apply image” part I just can’t get it. Where does the values “2″ and “256″ go. In opacity, scale or offset? Also, the gaussian blur is 22 pixels right?

    thank you!

  • http://www.tomdimaggio.com Tom

    @John: Here are the settings I use in the apply image. It’s important to keep those them same, otherwhise both layers won’t match each other.

    [eimg url='http://www.tomdimaggio.com/pics/Screen-Shot-2012-08-10-at-07.20.19.jpg' title='Screen-Shot-2012-08-10-at-07.20.19.jpg']

  • http://www.tomdimaggio.com Tom

    @John: I was to quick hitting enter :)

    The Layer that you used “Apply Image” on should be set to the “Linear Light” Blend mode. Opacitiy on both stays at 100% and yes the Blur is at 22.

  • ccting

    Merged photo.. !

  • John

    Hi it’s me again, fist of all thanks for the reply. But here I am again I just can’t pass the apply image step. It doesn’t leave me to put “256″ it change it to “255″ and when I put the blending mode in “Linear Light” the layer go totally white. What I’m doing wrong?

  • http://www.tomdimaggio.com Tom

    Hey John,

    Are you working on a 8 bit image or 16 bit image? Try entering 128 instead of 256. If you still have questions contact me on my email straight away, it might be easier ;) tom@tomdimaggio.com

    Cheers,
    Tom

  • chuck from Virginia USA

    The “right-handed” puncher should have been used in order for the head position of the “punchee” to be correct for the result of the blow shown.

    Then you could have simply over-lapped the two fighters for an even better composition and more realistic
    picture.

    As shown, the result of the blow thrown would have turned the “punchee’s” head to the left not to the right.

Some older comments

  • chuck from Virginia USA

    August 26, 2012 02:28 am

    The "right-handed" puncher should have been used in order for the head position of the "punchee" to be correct for the result of the blow shown.

    Then you could have simply over-lapped the two fighters for an even better composition and more realistic
    picture.

    As shown, the result of the blow thrown would have turned the "punchee's" head to the left not to the right.

  • Tom

    August 16, 2012 04:52 pm

    Hey John,

    Are you working on a 8 bit image or 16 bit image? Try entering 128 instead of 256. If you still have questions contact me on my email straight away, it might be easier ;) tom@tomdimaggio.com

    Cheers,
    Tom

  • John

    August 15, 2012 07:38 am

    Hi it's me again, fist of all thanks for the reply. But here I am again I just can't pass the apply image step. It doesn't leave me to put "256" it change it to "255" and when I put the blending mode in "Linear Light" the layer go totally white. What I'm doing wrong?

  • ccting

    August 14, 2012 02:55 pm

    Merged photo.. !

  • Tom

    August 10, 2012 04:05 pm

    @John: I was to quick hitting enter :)

    The Layer that you used "Apply Image" on should be set to the "Linear Light" Blend mode. Opacitiy on both stays at 100% and yes the Blur is at 22.

  • Tom

    August 10, 2012 04:04 pm

    @John: Here are the settings I use in the apply image. It's important to keep those them same, otherwhise both layers won't match each other.

    [eimg url='http://www.tomdimaggio.com/pics/Screen-Shot-2012-08-10-at-07.20.19.jpg' title='Screen-Shot-2012-08-10-at-07.20.19.jpg']

  • John

    August 8, 2012 03:49 pm

    hi, fantastic response. I was following the steps with the image you provide but when I reach the "apply image" part I just can't get it. Where does the values "2" and "256" go. In opacity, scale or offset? Also, the gaussian blur is 22 pixels right?

    thank you!

  • Penelope (NYC Blogger)

    August 7, 2012 11:39 am

    This is awesome, I wasn't even aware of this technique with layers (I only recently got my first dslr, but this opens up a lot of options and ideas for me!)

  • Salomanuel

    August 6, 2012 08:36 pm

    Thanks Tom.
    Thanks for listening to the comments feedback and for spending some time to answer properly!
    THIS is what being professional is all about!
    as soon as I will come back from holidays I will try to do some on my own!

  • Tom

    August 6, 2012 03:21 am

    @Leopoldo: Thank you, .. my thoughts exactly :)

    @Constance: I usually start with a very crude selection using the wand tool. I then use the refine edge option but only for the hair. I don't do anything else there. Once I have my mask in place I start doing the fine tuning of the mask using the Brush tool and the much underestimate Smudge tool. I work with those two tools until I am happy with the result. I might add that I check my mask against white and black backgrounds before I put them onto the final background. Once there I fine tune the mask so that it fits perfectly.

    @Judy, Salomanuel: The details of the post-processing would be another tutorial, but here are some quick details as to the post processing I use.

    The Lightroom part:

    I decrease the contrast as much as I can without loosing detail in the picture. It will obviousely look flat but fear not, contrast will be build during the Photoshop and the NIK part of the processing. I have attached a picture of the model I've used in this composite so you can see how flat the image actually is. It doesn't look natural but once the B&W is on top of it, it will look MUCH better ;)

    [eimg url='http://www.tomdimaggio.com/pics/Composite_Flat.png' title='Composite_Flat.png']

    For the Photoshop part:

    I create two layers in photoshop with the base image. The top one is converted to B&W using channels. The B&W conversion is different for each picture, I adapt it so that I get a maximum of detail out of it. I then change the blending mode to Overlay, depending on the end result I want, smooth or grungie I will increase or decrease the contrast on the B&W Layer. I then merge the two Layers to a new one and duplicate that one straight away. I use a Gaussian Blur Filter on the bottom layer with a value of 22 and then use "Apply Image" on the top Layer. In the apply image menu you'll have to choose the blurred layer as a base and a blending mode of Subtract with a value of 2 and 256. Click ok and change the blending mode of the top layer to Linear Light. Group the two last Layers together. No I start using the Healing Tool on the top Layer to clean the skin of the subject. This method will not affect the color tones but only the imperfections in the skin. If I want a grunge looking picture I don't use this too much. For a smoother skin I spend a lot of time on this.
    For the color tone I simply use the Clone Stamp tool with different blending modes (Lighten or Darken) at a approx. 20% Opacity. These are the steps I use on every single picture I take. They are quite similar, they only differ in the amount used and opacity.

    For the NIK software part:

    This is a bit trickier. I use only three Filters, Detail Extractor, Tonal Adjustments and Bleach Bypass. But I apply these three Filters differently on each picture. It depends on how much contrast I have in the picture, how many dark areas and/or how many bright ones I have. I am basically playing around with the settings until I am happy. I know it's quite tough in the beginning but with time you will know what works and what not, and get to the wanted result faster. I usually end up on the heavy side of things here, but decreasing the opacity of the NIK Layer is a clean and simply solution to that.

    Once this is done, the last step is to match the subject with the background in terms of color and depth of field. Once again there is no fixed recepi for this as it is different for each and every picture. I usually start this process with color filters in Photoshop and finish them with Tone Curve, Color and Split Tone adjustments in Lightroom.

  • Judy Kaspar

    August 5, 2012 02:25 am

    Agreed! A step by step would have been appreciated instead of hinting at the techniques used. There's no way to know how any of this works as is. Plus using Color Efex Pro, which is an awesome tool, we have no idea which filters were used and how.

  • Leopoldo "Leo" García

    August 4, 2012 10:48 pm

    TOM: Great work, thinking ahead makes the difference, and in a digital world, this is the key to separete oneself from the crowd. Congratutalions. I am saving this article. Way to go, NEED TO LEARN.

    "IF YOU DARE TO TEACH, YOU BETTER NEVER STOP LEARNING"

    Articles like this, are inspiring. Today, one needs to learn all tools of the trade as if one knows the pos-production tools and what can be done with them, you will be able to take your photography a notch higher.

    Great reads for a starting weekend.

    Leopoldo García
    Venezuela, South America

  • constance lowery

    August 4, 2012 07:42 am

    You did not tell me how you extracted your model or how you got them onto your background.

  • Salomanuel

    August 4, 2012 07:13 am

    awesome shot!
    but it would be awesome if you could make a step by step tutorial, so I finally can understand how you end with that kind of images!
    I'm not a photoshop newbie, but I'm not at that level

Receive a FREE SAMPLE of our Portrait Photography Ebook

  • Guaranteed for 2 full months
  • Pay by PayPal or CreditCard
  • Instant Digital Download

Receive a FREE SAMPLE of our Portrait Photography Ebook

  • Guaranteed for 2 full months
  • Pay by PayPal or CreditCard
  • Instant Digital Download

Receive a FREE SAMPLE of our Portrait Photography Ebook

  • Guaranteed for 2 full months
  • Pay by PayPal or CreditCard
  • Instant Digital Download

Sign up to the free DPS PHOTOGRAPHY COURSE

  • Guaranteed for 2 full months
  • Pay by PayPal or CreditCard
  • Instant Digital Download

GET DAILY free tips, news and reviews via our RSS Feed

Sign up to the free

DPS PHOTOGRAPHY COURSE

  • Guaranteed for 2 full months
  • Pay by PayPal or CreditCard
  • Instant Digital Download

GET DAILY free tips, news and reviews via our RSS Feed

Sign up to the free

DPS PHOTOGRAPHY COURSE

  • Guaranteed for 2 full months
  • Pay by PayPal or CreditCard
  • Instant Digital Download
DPS NEWSLETTER
DPS NEWSLETTER
DPS NEWSLETTER

DPS offers a free weekly newsletter with: 
1. new photography tutorials and tips
2. latest photography assignments
3. photo competitions and prizes

Enter your email below to subscribe.
Email:
 
 
Get DAILY free tips, news and reviews via our RSS feed