Compositing doll photos

The PNW BJD Expo is in Seattle in April and this year the theme is Paris.
Well we've never been to France at all, much less Paris so I was sort
of stuck about what to do for an entry. I planned to do some sort of
composite but I didn't want to infringe on someone's copyright. Enter
Wikimedia Commons; a wonderful comprehensive collection of imagery
that is freely used as long as you attribute it.

I also acquired a new toy recently; an 18" ring light. Ring lights are great
for photographing smallish items with no discernible shadows. They are
widely used by bloggers to photograph their craft or food or.... so are
widely available and at low cost. You shoot through the open center of
the light so the light comes from all around the camera. The doll in this
composite was photographed entirely with the ring light. No supplemental
lights at all.

Anyway as it turned out they moved up the entry date for the photo contest
and I didn't get the memo so when I went to upload my entry it was too
late so I went ahead and uploaded it to Flickr. My girl is shown leaving an
actual Parisienne boulangerie with baguettes in her shopping bag.

Shopper Star by Tom Beach, on Flickr

 

Hmmm, point well taken!

That image was based upon the painting "The Kiss" by Hayez.

The background in that painting is tack sharp so from that standpoint
what I did wasn't incorrect -but- It also has much less detail. Mine
is too busy so you're right, in my rendition the background fights
with the subject for attention.

I agree that when the background is well distant from the subject
it makes sense to soften it a bit and I've surely done that on occasion
usually I drop my subject into an existing image and if the image plane where
the subject will fit is sharp then I've left the background sharp as well.
Else I'd need to use a progressive mask to soften the far background
while leaving the near background sharp....

It's an interesting thought and I'll assuredly keep in mind for future and
current work. Thanks for chiming in. I've always wanted
this thread to be a conversation rather than a lecture....

 

Please please don't hang your head!

Everything I do has layers and masks so I'm reasonably fluent in using them.
I have no quarrel with your use of my image for illustrative purposes and in
fact it pointed out that when I created that work I did in fact create such a
busy background that it compromised the impact of the resulting image.

A question for you. Here is another of my Paris composites. It recreates the
scene in Funny Face where Audrey Hepburn runs down the Daru staircase in
the Louvre museum. The background image, from Wikimedia Commons, is
sharp front to back. I left it as I acquired it. In your opinion ought I to soften
the more distant elements?

Paris-2 by Tom Beach, on Flickr

Part of the sharpness of modern photos is that sensors are just so
spectacularly good anymore. Many professional grade cameras capture
40+ MPx and really good photos at 49,600 ISO (not mine though) It does
make sense to soften those images if you want to convey a moodier ambience.

 

Here's a suggestion for you. That image looks dark to me.
I'd pull it into curves and brighten the background a bit.
It's also somewhat confusing to me. The multiple figures
just create a jumble, not a cohesive story. I also wonder if it
would be enhanced if the bunny bounced around a bit. He
seems too static for the action surrounding him...

Have you tried the Quick Selection tool rather than Magic Wand.
You might also experiment with Color Range though I don't usually
find that as useful as the others...

That brought this image to mind...
Multiple layers and no selection required.
Camera on tripod and remote shutter actuator.
Layered all three images and adjusted opacity to suit.

I really appreciate your insights! My local photo art club has quite a
negative attitude toward doll photos so I don't have a lot of opportunity
to discuss artistic doll photography.


Megan by Tom Beach, on Flickr

 

We are in the process of dramatically reducing our book collection.
Among the casualties are several of my older photography books.
One of them described a technique for using a double exposure
to show his model emerging from a pool of water. That image inspired
this composite. It is so much easier to implement with digital images
and Photoshop than it was in the days of film...

River Sprite by Tom Beach, on Flickr

 

The image discussed today is not a composite.

It does require selecting the subject so from a practical view it could just as
easily have been composited.

I call the image "Survivor" and it is intended to depict a young lady who
has participated in and survived some sort of apocalyptic calamity.
The challenge was to make her grimy and gritty. I first tried applying a
texture layer and perhaps if I had found the perfect texture that might
have worked but alas, none of my textures seemed appropriate. I finally
chose a brush that had a sort of mottled appearance. Then I enlarged it
so that it was large with respect to the subject. I used black ink at low
opacity and stared dabbing all over her and her clothing. I did mask her
weapon so it is not dirtied as she is. I'm not sure it made any difference
but I believed it to be consistent with my character.


Survivor by Tom Beach, on Flickr

 

A new selection tool.

My laptop's battery charging tcircuitry failed so my laptop
worked while connected to the mains which sorts of negates
the who;e idea of having it. So I bought a replacement so
I've been busily loading software and migrating data. Last
evening I loaded my photo editing tools. While on the Topaz
Labs web page I found their new selection tool AI Mask so
of course I had to download it and try it out.

At $99 it's not cheap but it's head and shoulders better than any
other selection tool I've used. Note the lace at the edge of the sleeve
on this image and I didn't need to do anything to assist it. The hairs
by the edge of her wig aren't as well selected as I'd like but again
I didn't do anything speciall there and few tools do anywhere nearly
this well in one default pass I got a discount because I already
own their ReMask product which is the predecessor. I'll be using
this tool a lot.

Composite by Tom Beach, on Flickr

 

I've almost always composited my girls into another photograph.
Well I did here also but I first converted that photograph to a virtual pencil sketch
so I have my photo realistic girl in a sketched environment.
It's a very different look. One of a series of images harking back to mid century
pin up art.

Hitch hiker by Tom Beach, on Flickr

 

In my first post to this thread I talked about lighting with the intent to composite.
At that time I was using a reflective umbrella to create a soft main light. Recently
I acquired a ring light and I've been using that for my main light when the intent is
to use the image to composite. I still use a pair of side lights to create good definition
to the edges of the subject to make selection easier. This is my current lighting
arrangement.

Lighting for compositing by Tom Beach, on Flickr

This is my newest composite.

Forest Trail by Tom Beach, on Flickr

 

Selection tools are getting better and easier to use all the time...

Earlier this week I got a notice from Adobe that there was an upgrade
to Photoshop. Well, more than an upgrade it is a whole new release;
Photoshop 2021. This release brings a complete rework of Adobe Camera
Raw and a new Sky Replacement tool but today's topic is the Select Object
tool. This tool was introduced in Photoshop 2020 and like many of their tools
the initial release was underwhelming.

The new iteration is a whole different beast. When you make your base
selection you get two options; select object and select and mask. The
advanced tools are in the select and mask option. Here you can iteratively
add to and subtract from the selection and invoke the Refine Hair tool which
does the best hair selection I've ever had the pleasure of working with.
Refine hair is also available within the quick select tool.

If you use Photoshop you will love the new ACR and you'll also like the new
selection tool improvements.

Notice Diedre's hair on the left side of her wig *your right)

City Girl by Tom Beach, on Flickr

 

Thank you! Practice helps a lot. I'm far better at this than I was all those years ago when I began this journey.
Also the tools are immeasurably better. I just looked and this thread is 5 years old.

 

It's been a while since I've visited this thread so I's time for an update. I noticed that the
images used in two recent posts aren't showing. Fortunately the links work OK because
I'm completely unable to get those image to show in the posts. If I edit the posts the images
are fine within the editing fields but are gone when I apply the edit. Oh well, such is life on
DoA I guess. Today's topic is background images. I try to use my own when Possible but
sometimes it just isn't possible. Recently I got an Indian belly dance costume for my SmDs.
Well I've never visited India so nothing in my personal library to use. As a photographer
myself I'm really sensitive to usage rights and copyright infringement so I just don't lift
images from an arbitrary web site. Fortunately there is an amazing source of public domain
imagery including image from almost any location you can imagine; Wikimedia Commons.
I've mentioned it before but it is worth mentioning again.

Anyway, here is my girl Bonita in her Indian belly dance outfit performing in an India setting.

Performance - 1 by Tom Beach, on Flickr

 

Thanks for that. Even when the preview is missing I find that I can usually
get to the images by clicking the link which still usually remains.
I went to try your method but tonight the photos which were missing earlier
are all back. So frustrating....

I try never to composite my dolls into mortal danger; just in case you're right.

 

This is something of a departure for me in my personal creative journey. It's been a week or more
since I've picked up my camera and when I thought about which dolls I wanted to photograph and
which props and backgrounds to use I realized that I've hit something of a creative wall so I did
what I sometimes do to refresh my creativity and revisited a CreativeLive class that I've found
inspiring.

This is the result. There is nothing conceptually new here. No new selection tools or exciting blending
modes but me rethinking my creative objectives. I love exploring fantasy worlds and here I've composited
Bonita into a formal garden. I manipulated the photograph of the garden to create an unworldly setting
for her and chose an image of her slipping out of her dress. The whole idea was to create something of
an enigmatic image. Discussion welcomed....

Enigmatic by Tom Beach, on Flickr

 

I've not yet made the effort to install Photoshop on my iPad so that's all new to me.
It appears to work just fine and it's certainly a different technique. Thanks for sharing!

 

There are fundamentally 2 models for selecting a subject from a photograph.
You can look for edges or you can look at color. If you work with Photoshop
then "quick select" and "select object" look for edges. "Magic Wand" selects
a color. Akvis SmartMask and Topaz Mask AI also allow you to create a selection
using both techniques.

This issue arises because I got a new doll last week and she has a lace dress.
Edge selection algorithm just don't work for lace. You need to drop all of the
background color visible through the lace. I did two variations of this selection;
once with Photoshop and once with Akvis SmartMask. I'm satisfied that both
tools are capable of selecting a lacy subject.

As an aside though Akvis is based in Russia and their web site is currently
inaccessible.

Anyway, here is the composite.

Gothic by Tom Beach, on Flickr

 

This is not a Photoshop tool I use very often but when you need it it's incredibly useful.
It's called "blend if" and you'll find it in the layers menu. (I use Photoshop)
This is the composite:

Stranded - 2 by Tom Beach, on Flickr

There are three layers here; the background, her reflection in the water, the doll herself.
The reflection layer is above the background and underneath the doll.
I didn't want the reflection to show on her little island nor did I want it to cover
the ripples around the stones. This is where blend if comes into its own.
With the reflection layer selected right click that layer in the layers panel and
choose blending options from the pull down.

There are two blend if sliders.
The top one allows you to protect the selected layer while
the lower one allows you to protect the underlying layer which
is what I did here:
Pulling the left hand slider to the right protects the darker tones
of the island from being effected by the reflection.
Pulling the right hand slider to the left protects the brightest ripples
from being effected by the reflection.
Yes, not something you'll use very often but when you need it then
you need it.

 

Adobe recently released new "beta" versions for some of their applications including
Photoshop. This new tool is called "Generative Fill" and it creates an image within a
selection based upon a short text description of what you want there. For anyone who
loves to composit, this is as close to heaven as one can get.

I asked it for an abandoned factory and it gave me this background for a fashion shoot.

Haute Couture 1 by Tom Beach, on Flickr

I asked for a medieval market and it gave me this background for my fantasy girls.

Going to market by Tom Beach, on Flickr

It's definitely not perfect yet. If you create a figure within a selection it merges it with
the background within your selection so you can't edit the new object without doing a new
selection and it's only notion of trains is from the UK but it's still in beta and it you enjoy
creaing new backgrounds for you kids this is flat out amazing!