For a
final render image at 4000 pixel, we change the setting so that it is
higher than before to bring more quality to the image.
For GI
method, we used Irradiance map for Primary bounces and Light cache for
Secondary bounces at a high setting.
We chose
Adaptive QMC for Image sampler and Michell - Netravali for
Antialiasing filter.
For post
work, we also rendered out the VrayReflection pass, VraySpecular pass
and VrayZDepth pass (you can add these passes by going to Render
Elements and click the Add button).
6. Post Work
We use Photoshop for post work. Level adjustments and
Color Balance are used to improve the contrast and color for the
image.
A real sky background is added for more realism.
Reflection
and Specular passes are used to add more reflection and highlight for
the image to provide more realistic effects.
We also render out a
Vray Dirtpass render so we can use it in Photoshop to boost the
shadow.
Some
close-up details:
And
there it is. We get the lighting as we wanted. It is quite simple as
you can see, but it sure is full of challenges to get the right look
for the final image. I hope you will get other ideas in lighting an
exterior scene as well.
Enjoy!
Rendering and Post 1. Introduction
Before
we began work on the L5 project, we spent quite a bit of time trying
to find the right look for the imagery we were going to produce. The
art direction was influenced heavily by the work of photographer Julius
Shulman, the American architectural photographer best known for his
work featuring icons of modernism like Pierre Koenig's Case Study #22
in Los Angeles. We also wanted to accentuate the location of the
project by really playing up the desert setting of Las Vegas,
particularly the way that light interacts with spaces. We settled on
something akin to the bleach bypass look of film for the stylized look,
pushing the contrast and tonality in the image.
2. Getting Started
When
it came time to render this scene we had to make some decisions:
a.
File format - a few versions back Photoshop introduced the
ability to work with high dynamic range images (HDR or EXR). 3D Studio
MAX allows you to save images as EXR. The VRay rendering engine also
allows you to save VRay Image Files (vrimg) which can be converted to
EXR's. High dynamic range files give you the ability to work in a
higher bit depth and with better color precision. Having this ability
gives you a much broader range of possibilities when post-processing
your image. We used MAX's EXR format for this rendering.
b.
Render Elements - Rendering to elements lets you separate
information in the rendering into individual image files. These are
very useful when doing post-processing in Photoshop. Render elements
render "for free", meaning they don't add any additional time to a
rendering, so we always opt to include a number of them with our
renders - even if we don't end up using them. The usual suspects are
below. For more info on using render elements with VRay go here:
I. VRayDiffuseFilter
II.
VRayGlobalIllumination
III. VRayLighting
IV. VRayReflection
V.
VRayRefraction
VI. VRayShadow
VII. VRaySpecular
VIII.
VRayZDepth
3. Occlusion
Pass / VRay Dirt
the occlusion pass has become
ubiquitous in CG over the last few years. We use it to achieve a number
of different effects, but mostly it helps to enhance shading around
the edges of your scene. We use both the Occlusion shader in MentalRay,
as well as the VRayDirt shader in VRay. For this scene we went with
VrayDirt. For an in-depth explanation of how this shader works go here: VrayDirt
explanation
See fig 01 for our settings used
in this particular instance. Note the "Subdivs" value, which was
purposely kept low to produce a noisier result resembling actual dirt.
Because occlusion passes have little color information we typically
save them as either Targas or TIFF's.
[Fig 1]
4. Mattes
mattes quickly
become your best friend when doing a lot of post or "paint" work in
Photoshop. They render quickly (for the most part) so here again we
usually create a list of objects that we're planning on affecting in
one way or another during the post process. Usually the most prevalent
surfaces/objects get their own mattes automatically: glass, walls,
floor surfaces, etc. For this process we create a copy of our working
file so that we can apply the appropriate materials to render our
mattes. There are always two materials we create:
a.
Solid self-illuminated white - see fig 02
[Fig 2]
b. Matte -
using the VRayMtlWrapper material - see fig 03
[Fig 3]
The solid
self-illuminated white material is applied to the object(s) you want to
isolate, while the Matte material is applied to the objects you want
to (you guessed it) matte out. We typically save matte render files as
16-bit Targa files (TGA) - see fig 04 for the Targa Image Control
settings to use.
[Fig 4]
5. Assembly in Photoshop
a.
Base render i. Because our image rendered
as a 32-bit EXR there are a limited amount of adjustments that can be
made - one of them being tweaking the Exposure. See fig 05 for
accessing the Exposure control in Photoshop. We adjust only the Gamma
to 1.60 - see fig 06.
[Fig 5]
[Fig 6]
ii. Before
we can do anything else we have to convert our image to either a
16-bit or 8-bit file - we chose 16-bit so that we could continue
working with a maximum amount of color info. See fig 07 for this
process. Note that here we can also adjust the Exposure and Gamma like
we did in the prior step.
[Fig 7]
iii. Next
we matte out the main render using the main alpha channel.
b.
Occlusion/Dirt Map
i. We add the
occlusion pass to our scene above the main render pass and use the
"Multiply" transfer mode.
ii. You'll notice that
the image turns very dark. This is because the Multiply transfer mode
in Photoshop uses the color values to affect the image - the darker the
Multiplied layer is the darker the overall image will become. In order
to control which parts of the occlusion pass affect our main image we
do two things.
1. Levels adjustment layer linked
only to the occlusion pass. In the levels control we bring up the
brights so that the occlusion image goes almost completely white except
for the corners and edges of our scene where the "dirt" lives.
2. We also add a Hue/Saturation adjustment layer linked to the
occlusion pass and use the Colorize feature to add color to it - in
this case a pale blue.
[Fig 8]
iii. As
you can see in fig 08 our image now has more definition in and around
all the nooks and crannies, as well as a bluish cast to begin
implementing the bleach bypass look we were striving for.
c.
Render Elements - we usually use these elements in an
additive fashion so for this the Screen transfer mode works best - in
most other compositing applications Add is a blending mode, but not in
Photoshop. We usually use the elements below to accentuate things like
reflections, specular highlights and control overall lighting.
I.
VRayGlobalIllumination
II. VRayReflection
III. VRaySpecular
d.
Paint - paintwork is usually reserved to specific elements
within a scene that we need to adjust. In this particular case we
tweaked the corrugated metal and glass by slightly augmenting the blue
cast on both of these objects being that they are both reflective and
as such would be affected by the blue environment. We also had to
change to color of the metal railings after the fact due to a client
request. These changes are often times easily managed in post with the
proper preparation.
I. Corrugated metal
1.Using the proper matte for this object we added a Color
Balance adjustment layer to enhance the blue reflection on the metal.
2.
We also used some dark yellow brush strokes on a Color Dodge transfer
mode to add some highlights where the sun would be hitting the metal.
II.
Glass
1. Again using the proper mattes
we added a Color Balance to give all the glass a bluish cast. 2. We also separately affected the glass facing the sun by
increasing the brightness to enhance the effect on the parts that were
being hit with direct sunlight.
e. Global Adjustments
- in the case of images that are meant to be heavily stylized, we
usually do a good amount of global adjustments once we have all the
elements reading the way we want with relation to one another. As
mentioned before, for this particular project we were going for a
"bleach bypass" film look - for more on this visit:
The
"bleach bypass" process renders images that have reduced saturation and
a high level of contrast. To achieve a similar look in post we did the
following:
I. Color balance - first we
introduced a good amount of warm tones in the shadow and midtone areas
of the image using yet another Color Balance adjustment layer - see fig
09.
[Fig 9]
II.
Hue/Saturation - using the Colorize feature in the
Hue/Saturation adjustment layer we create a duotone version of the image
(blues and blacks). By using the Soft Light transfer mode we are able
to blend the blue tinted image with the warmer version. Note that Soft
Light will tend to not only brighten bright areas, but also darken
shadow areas, so by using the Lightness control in the Hue/Saturation
settings you can manage this blending - see fig 10.
[Fig 10]
III.
Overall levels - the image is still a bit dark in fig 11 so we
add a Levels adjustment layer to very subtly bring up the light areas
of the scene.
[Fig 11]
IV.
Vignetting - this effect is in real life an optical phenomenon
attributed to the physical properties of a camera lens as well as the
aperture settings used to shoot photographs. In a lot of cases it is an
undesired effect, but sometimes it can be used to draw attention by
framing the center of the image - this is what we wanted to achieve
with this rendering. You can create this effect a couple of different
ways.
1. Using a VRay Physical Camera you can
check the Vignetting checkbox and achieve the effect in a realistic way
by using the right combination of lenses and f-stop settings - see fig
12.
[Fig 12]
2. In
Photoshop you can control this effect in a much more fluid way by
simply painting a halo of any given color (usually a dark gray) around
the outer edges of your image. In this case we used the Multiply
transfer mode to darken the existing colors of the rendering. We then
accentuated the effect by using a Curves adjustment layer with a mask
affecting the same outer edges of the image.
And voila, we're
done! One of the things you'll notice is that most of the post work on
this image was done using adjustment layers. This method is completely
non-destructive allowing you to make adjustments every step of the way
while also leaving you with the option to go back to original imagery
if necessary. It's not quite as well implemented as it is in
AfterEffects but with some practice it can be almost as efficient. See
fig 13 for the before and after.
About the author
I'm a Web-Graphic Designer, Freelance and Webmaster, blogger and more. I like to push the boundaries. I love creating unique, clean, usable design for the web and other digital sources.