Texture Process Diary

3D TEXTURING PROCESS DIARY

Texturing is the process of applying the surface qualities to objects such as colours and texture, assigning 2D images to 3D objects.

Notes from Lanier, L., 2011. Maya studio projects. Texturing and lighting. Sybex, Indianapolis, Ind. p123

UV snapshot and bitmap texture resolutions of 512, 1024, 2048 are resolutions commonly used in animation and stem from the historical use of binary-based computing.  1024 is 2 ^10 or 1K and is the number of bytes in a kilobyte, 512 is half and 2048 or 2K is double.  1K and 2K refer to the number of horizontal pixels carried by a render resolution, film scanner resolution or digital projection resolution.  The render resolution of 1024 x 768 is 1K with a 1.33 aspect ratio, the resolution wider is 1.33 times greater than the height.

FRUIT BOWL TEXTURES using MUDBOX, PHOTOSHOP AND MAYA

APPLE

I started by creating white shapes and marks in photoshop to use as stamps in Mudbox to make the stripe and dot colours on the apple.  The white area of these shapes allow the colour to come through on the surface of the apple model being painted.  When using these custom stamps in mudbox and adjusting the settings I was able to vary the size, strength, length, spacing and randomisation values of the shapes and strokes as I built up the diffuse colour of the surface.

some dot shapes created in photoshop

some dot shapes created in photoshop

dot and line shapes from photoshop in mudbox

dot and line shapes from photoshop in mudbox

stamps_001

stamp made in photoshop

stamps_002

stamp made in photoshop

stamps_003

stamp made in photoshop

testing shapes and settings on the stamp in mudbox

testing shapes and settings on the stamp in mudbox

appleStrip0esCompile_001

As in photoshop, mudbox uses layers, different layers were used to create the base colour, more refined colour, then the different coloured stripes and the dots.

appleDots_002

stripes and dots applied to the diffuse layer

I looked at creating darker areas of the diffuse, being the darker colour on the apple by using a black and white image.  Where the colour came through the white area and then by adjusting the weight on the shader I was able to use the black areas to darken the apple’s skin.  Later I included this in the colour file from mudbox and did not adjust the diffuse weight setting, setting the roughness of the apple across the whole of the surface and using this black and white image to work on my values for reflectivity and glossiness.

appleLayers_001

By using one or several of the layers created in mudbox to make these black and white images I’m able to adjust the attributes in the apple’s Mental Ray’s shader.  Adjusting the values of the file in colour balance of the shading network translates into how the surface qualities for reflection’s reflectivity and glossiness appear in the final render of the surface.

 

Sculpt layers can be set with different subdivisional levels for the numbers of mesh polygons. This allows more or less detail in the surface sculpting on that level for that layer.  The idea being to vary the surface relief of the apple making the shape for where there is variation and damage to the surface.  These different levels of detail can then be used to create different surface details and maps for the final render of the object.

To show these surface relief changes for the apple I sculpted the surface creating a normal map, the bluish coloured one above, to assign to the Maya shaders bump attribute to create the illusion for the changes in surface shapes. I was not doing all the sculpting on the appropieate subdivisional levels, creating differences in the sculpting of the surface is specific to that layers number of polygons, I was sculpting on the base layer as well.  Mudbox needs the difference in the surface shape to create the normal map and this reduced the amount of surface changes created by the normal map and changed how the surface looked in my render.  It was possible to adjust the normal map in photoshop by duplicating and adjusting the layers using overlay.

appleLayersCompile_001

redAppleDiffuseRoughness

When building the shading network a luminosity node is being used instead of ticking the Alpha is Luminance under colour balance. The luminance node converts the three channels of RGB to one alpha channel enamelling the values to be adjusted via the the colour balance’s Colour Gain.

refllectionGlossinessBalance_001

finding the balance between reflectivity and glossiness

refllectionGlossinessBalance_002

getting closer for the balance between reflectivity and glossiness

transferring the values from the reflectivity and glossiness to the alpha gain

transferring the values from the reflectivity and glossiness to the colour balance’s colour gain

I could not work out why I was getting a sharp line across the apple between the reflectivity and glossiness with two distinct areas and not being able to adjust them cleanly.  I was working out the values in reflectivity and glossiness with the diffuse value included.  Turing off the diffuse, putting the colour to black and then adjusting the values between reflectivity and glossiness independently made it clearer to see what was happening. Finding these values and then put them into the colour gain I had more control.

 

ColourGainApple

 

After more adjustments for the reflection for the apple shader and putting the values into the colour gain of the colour balance on the file node it was still not looking the same.  After trouble shooting by my lecturer it turned out I did not have a value of 1 for the whites in the images being used to control these attributes, this resulted in the wrong values controlling these attributes. The white in mudbox is one, I need to make sure that is what I select.

Another  issue I had overlooked was the glossy samples which was causing the graininess of the render – the number of samples (rays) for glossy reflections. Higher values create a smoother result and is less grainy.

appleColourBalance

I think I also painted on the wrong subdivisional layer.

finalApple

final red apple

 

BANANA

For the banana I use stencils to create the surface colour and textures. The stencils project the image onto the model’s surface, including the colour .

banana_011

reference

banana_010

mud box stencils

bananaLightDarkLayer

When using the projection option under the paint setting in mudbox for these stencils it was important to set the colour in the properties to white. Colour is a multiplier and when white was not selected I was not painting the colour selected in the stencil.

Again using layers to separate the areas of the banana, the base colour being flooded on the model and the darker layers painted in the areas where textures will be created.  These different layers for the damage and end parts gave options when selecting different layer combinations for the black and white images. This maps being used  for the shader’s smoothness, roughness, mate and reflective surfaces of the banana.

dotLayer

combined layers

 

bananaBW_002

black and white layer to use for the diffuse and reflection values where there is damage and darker areas

bananaBW_001

black and white layer to use for the diffuse and reflection values for the ends

This is an example of using painted layers from mudbox to create and control the realistic properties of the banana. The white areas will have a value of one and the black areas will have a value of zero when used to adjust the surface qualities. The roughness attribute of the banana is totally powdery and chalk at a value of one or white and complexity smooth when at a value of zero or black.

Looking at this black and white image what will happen is that in the white areas it will be totally smooth and the black areas will be chalky. When adjusting the value between zero and one the range from smooth to chalky can be controlled.  This means where the banana is damaged the black area will appear rougher and the white area will appear smoother, so the slide set to black will show a totally smooth surface.

bananaBwhiteMap

 

I have also used these black and white values to control the banana’s surface qualities of reflection and glossiness as the same areas share these qualities. Because the black and white map is connected to each of these attributes I am able to vary the values for roughness, reflection and shininess separately.

 

building shading network and values

building shading network and values

 

 

final banana

final banana

 

PEAR

The adventure of texturing a pear.

Screen shot 2013-08-24 at 3.56.33 PM

Developing from the apple and banana I am looking at the pear differently, thinking about how to approach this while considering the surface qualities of smooth, rough, diffuse and reflectivity.  How painting these layers will be used for the pear’s surface qualities.

pear_011

reference

I created a variety of colours, shapes and types of dots for the pear on different layers using both custom made and stock stamps,  combining this with projections.

pear_012

mud box stencils

Again starting with layers, building and blocking out different areas of green and yellow that made up the main colours.

Another layer for the brownish stripe down the side and how it looks with the main colour.

The dotted marks on the pear were painted on the next layer, followed by a layer for the marks, damage and colour around the end area while thinking about how they are rougher than the rest of the surface, the dullness and the variations on the shape of the surface

The next layer was for the stalk area.

pearDots

 

pearSculptShape

Looking at the pear’s surface changes and the painted layers I was able to sculpt the surface variations and damage to go with the colour changes.

pearSculptPaintpearSculptaint

With these two images I started colouring for the marks and sculpting for the surface changes and damaged areas.

pearSculptStalk

I sculptured finer line details at the top and on the stalk, not sure how this will show in the reflectivity and highlights.  Thinking back on the tree stump we did in class it is probably where I could have used the ideas of a dirt map and ambient occlusion.

pearBlackWhite

colour layers for the dots and damage duplicated and converted to black and white

Working with the layers to make the black and white files to use in the maya shader for adjusting the defuse and reflective attribute qualities.

Then being able to adjust the differences between smooth, rough, reflective and dull surface values.

pearDiffuse

colour

In the maya surface shader the are:

Diffuse

– the colour of the object

– the weight, maximum amount of diffuse reflection scattered by the type of surface with zero being black and totally absorbing and one being white and totally reflecting the colour

– roughness which has black at zero and being smooth and one being white and the highest value for powdery or chalkie.

Reflection

– reflectivity with zero or black having no reflection and one being white and highly reflective

– glossiness with zero or black having a totally diffuse reflection and one being white and being a mirror reflection

It is how the paint information is used in these map through the object’s shaders attributes that creates the surface textures.  The values between 0 and 1 give the options to choose a texture quality between those values creating a range of surface values.  The idea being to use these textures to control the output of the shaders, with some characteristics requiring the use of several textures by combining these attributes.

 

pearNormal

normal map

 

The normal map shows the surface sculpt changes from the mudbox sculpt layers.

final pear

final pear

 

ORANGE

016

reference

017

mud box stencils

The image below shows the painted area of the orange as used for the colour attribute in the maya shader. The file is made by using the surface UV’s from a UV layout of the 3D surface. The map is a flattened 2D map of the model’s surface matching the same UV coordinates of the 3D orange where the colour has been painted.

 

orange_colour_sRGB_test copy

How do I go about the surface relief for the orange?  I wanted to see how using stencils to create a bump would work.

On the first attempt I had not realised how the stencil values would work with the black and white colour adjust when converted to black and white.  There were more raised areas from the white than indentations from the black for the orange pores.  After painting more versions with the stencils and with various sizes, shapes and colours it was improving, less white areas for the pores and more control over how the pores would be on the surface.

orangeBumpValue

adjusting bump value

I had sculpted a few damaged line areas on the surface to include in the extracted normal map, how was this going to work with the bump?  Thanks Eric for explaining how to do this. I converted the bump to a normal map, set to overlay with it on top of the normal map I had extracted then was able to output the file into my Maya projects source images folder to be used on the oranges shader.

Normal maps create relief detail on the surface by changing the angle of the surface normals to reflect the light in different ways.

orangeNormal

normal map that started out as a painted stencil that was made into a bump map

tangent space normals not selected

tangent space normals not selected

tangent space normals selected

tangent space normals selected

 

final orange

final orange

GREEN APPLE

For the green apple’s painting and sculpting I used a mix of the stencils I created in photoshop for the dots and lines with stencil images.

appleReferenceCompile_001

My green apple had a very strong and blown out difference between the reflection and glossiness attributes of the shader which I was not able to control.  I could work out how to solve the problem. It turned out to be the lights, too bright and realising they cannot be excluded from the equation. They were too large and too bright and after adjusting the lights I had more control over the shaders reflectivity and glossy values.
greenAppleLights_001

Screen shot 2014-06-15 at 5.13.31 PM

UV layout in photoshop

A problem I had was with the painting of the black areas for the reflection and glossiness. The areas were too large, not matching with the areas on the green apple.  It was interfering too much with the overall surface qualities, showing up as large blocks of dull and diffuse surface instead of with the selected damaged areas.

Selecting a new layer, flooding with white and using the multiply blend option leaves the layers underneath coming through. I could then see the area in these layers where I wanted to paint the black.

first reflective and glossy map

first reflective and glossy map

second reflective and glossy map

second reflective and glossy map

stronger, sharper highlight

stronger, sharper highlight

flatter highlight

flatter highlight

apple_001

a more diffuse highlight

apple_002

brighter highlight

 

looking at the reflectivity and specular highlights

looking at the reflectivity and specular highlights

greenAppleReflectiviyt_002

shinier highlights

 

greenAppleReflectivity

duller highlights

 

I’m not sure about the level of sculpt detail required and the depth of the changes on the sculpt surface to display the changes via the normal map on the green apples surface.

apple_003

some apple sculpt detail

A way around this when the sculpt detail is not enough is to use a multiply node in the shading network and adjust the level of detail there. The out colour of the normal map is connected to the input one of the multiply-divide node and below this are the R, G, B boxes to change the values. This is then connected to the bump in the shading network through a luminance node which converts the RGB to a single value to drive the bump value.

multiply value set to 2.0

multiply value set to 2.0

multiply set to 1.1

multiply value set to 1.1

final green apple

final green apple

I’d like to develop:

– more sculpting details

– painting with the projection of real textures created from the real world environment

– to break down the surface into more layers enabling a great range and flexibility to mix them for different surface textures

– a better understanding of light, dark, black and white in building surface detail when making the texture maps

– to experiment more with the colour adjust, converting to black and white values seeing what a larger range of values does

– to understand more how the mia shader texture network works by compositing the different attributes after working out the values separately

Procedural Network for an Orange

Procedural textures are created through mathematical calculations and have no limitation in size (resolution).  3D procedurals ignore UVs and have the ability to texture non-uniform geometry.

DIGITAL PROCESSES REPORT

This tutorial is looking at the procedural bump network for an orange using Maya’s hypershade.

To create realistic images shading networks need to consider qualities such as colour, specularity, reflectivity and surface detail. They are the building blocks describing what happens, how objects respond to light, creating the surface appearance and sending this information to the render. The shading networks are build as nodes that define how various colour and texture attributes work together with associated lights and surfaces. The hypershade window lets you connect the nodes attributes and view the connections.

one

Shader types handle specularity in different ways and needs to be considered for the orange’s surface colour and texture.  The phoneE gives control over the specular highlight in regards to the amount of light that is reflected, it has a roughness attribute which can make the material appear rougher and a whiteness control to blend colour highlights between the base colour of the material and the white.

Firstly considering the colour of the surface, mapping a ramp texture shader node into the colour attribute of the phongE creates options for the graduations of colours. The ramp controls the direction on the surface and its interpolation or blending. For the orange giving it a range of colour including the green around the stalk area.

two

When mapping a noise 2d texture into selected ramp positions the surface can be   manipulating by the nodes attributes. This node breaks down the surface into different qualities creating small circular shapes on the orange surface, like dots. The noise shades has attributes such as type, density, spottyness and size rand that are not available in other nodes. Adjusting these values controls the black and white values of the noise.  The attached place 2d texture node’s gives control of the position, size and rotation of the UV’s attributes that are created and placed on the surface.

Using the remap colour node, in utilities tab of the hypershader, I’m able to choose colours for the black and white values of the noise shader by converts RGB to HSV giving accurate colour control that relates to the black (zero) and white (one) values set in the noise attributes.

three

The relief surface detail of the orange can be set up using a fractal texture node. The black and white values are adjusted creating size, shape and depth values for the surface relief. The attached 2d texture node controls the UV space of the geometry, which is inherent in nurbs surfaces. The origin of the surfaces UV coordinate are (0,0) and the surface extends to the left and up where the coordinate is (1,1) on a graph. This is mapped directly on the surface with a join.and the coordinates are used for positioning.

four

To create the relief appearance the noise texture node is duplicated and mapped into the shader bump attribute.  Adjusting the surface normals based on the texture map’s alpha values where white is the height value there is control of the relief detail. The bump maps are grey scale textures and a negative value in the bump notes value creates the illusion of depth.

five

Creating a stacked bump map allows adding the fractal values to the bump attributes, giving height detail of the orange by adjusting the positive numbers.

Eric I have included this for you,

“ You can also “stack” different bump maps on top of one another. To do this, connect the Out Normal attribute of one bump map to the Normal Camera attribute of another.

Checker       Bump3d        Cloth         Bump3d           Blinn

——-       ——        —–         ——           —–

Out Alpha –> Bump Value    Out Alpha –> Bump Value”

Mapping a second fractal into a third stacked bump and using a colour remap node allows the addition of surface relief and colouring for damage and dirt on the skin.

six

Considering the shininess of the orange surface adjusting the reflective of the shader under specular shading will reduce how much the surface reflects this light. Setting to zero will not reflect at all. The orange has low reflectivity which could be as low as .05.

seven

Some helpful definitions:

diffuse : white = 1, position of surface and it’s angle relative to the light. How much light is absorbed and how much is scattered in all directions by surface imperfections. Rougher surfaces tend to have higher diffuse values.

specular : calculated from a specific camera angle and is based on the angle between the light, the surface and camera. View-dependent, shading. Give appropriate size. more highlights

References:

Maya online help files

Birn, Jeremy, and Jeremy Birn. Digital Lighting and Rendering. Indianapolis, Ind.; London: New Riders ; Pearson Education [distributor], 2006.

Brown, Tim H, and Alias (Firm). The Art of Maya: An Introduction to 3D Computer Graphics. Toronto: Alias Systems Corp. Education, 2005.

Derakhshani, Dariush J. Introducing Autodesk Maya 2013. 1st ed. Indianapolis, IN: Wiley Pub., Inc, 2012.

UV LAYOUT DIARY

Texture Maps attach themselves to geometry by texture coordinates called UV’s.

Nurbs come complete with UVs and UV coordinates are always locked, they cannot be adjusted without adjusting the geometry.

Polygons need UV’s assigned to them, it is done by using projections and the UV positions can be edited without affecting the geometry.  The projections arranges the pattern of the UVs, it can be distorted, stretched, pinched or overlapping.  To get more uniform checkered patterns the UV’s need to be adjusted.

Pelt Mapping unfolds the surface so that it becomes a single large shell with a minimal number of seams.

Ptex, per-face texture mapping, does not require UV mapping and is able to store separate texture maps for each polygon face in a specialised file format.

 

model_002UV mapping uses two-dimensional layouts of the 3D surface for applying textures and using the software headus UVLayout the character’smodel_001 surface has become a 2D layout map.  The UV map is a guide for the placement of images.

In this exercise I was aiming for minimal stretching, tearing and distortion in the layouts so when surface textures are applied to the model they are as clean as possible.  The brief was to do the UV layout from the front camera’s point of view.

021

Looking at the character from the front the first consideration was where to start making the seams.  The obvious ones were between the cloths and the body, then how to break up the cloths considering the shirt and pants for unseen seams and clean painting and texturing.  For the top I decided to go along the underarm and across the shoulder and for the pants along the side and across the inside.  I did have a look at doing the join down the back of the top and pants, it was more difficult to unfold evenly as a single piece.  I found the compromise between this and doing two pieces gave me a more even unfold.

022

The idea is to move the UV’S in the layout to achieve minimal distortion to the surface of the model. It was not so easy to decide on where to put the seams for the arms, hands, legs and feet.  After cutting a few different ways between the arms and the hands and around the finger nails I ended up cutting the hands and arms as two pieces. After flattening both pieces then working out a place to join them to put the seam away from the camera.

023

There were several overlapping UV, particularly around the finger nails and this was another area where I compromised on getting rid of the overlaps and keeping minimal distortion of the numbers on the 3D model in not so obvious areas.

Screen Shot 2014-05-11 at 3.39.31 pm

In the 3D view numbers can be displayed in each square on the surface and they display the size, distortion and blurriness created when working on the 2D texture map. The face was not so easy around the nose and eyebrows to get an even unfold on the model.  It was a constant toggle between moving the UV’s on the map and looking at the distortion on the model to find the best possible compromise between the green colour on the map and the undistorted numbers on the model.

Screen Shot 2014-05-11 at 3.39.43 pm

For the final layout it is efficient use of the layout’s space to be able to stack the same parts that have been UV mapped, they can be  painted together as one piece.

allGreen_001

UV TEXTURE MAPPING

“There are two types of textures:  procedural and bitmapped.  Procedural textures use the Maya nodes attribute to generate an effect, such as ramp, checkerboard or fractal noise textures. You can adjust each of these procedural textures by changing their attribute values. A map, on the other hand, is a saved image file that is imported into the scene through a file texture node.” Introducing Autodesk Maya 2013,  page 286.

UV mapping places the texture directly on the surface, giving the point on the surface by using the surfaces 2D coordinating system for width and height.  UV’s are not often laid automatically for immediate use, often being bunched up and needing to be adjusted. Seams can be chosen to work the better ways, such as away from the camera’s view and focusing on areas of less detailed texturing.  Also needing to eliminating overlapping for a clean layout and choosing the best options for the surface texture layout required.

Maya and other software is able to lay out these 3D UV’s as a flat surface connect the textures to the 3D surface making it possible to choose options for how this would look best. Where do I put the edges, where will they appear on the surface of the model?  How do I map out the 3D surface as a 2D map with cutting, unfolding and stitching together the pieces while considering how the texture will look while minimising the visible distortion. The aim being to create a flat piece that will wrap around the shape with as few seams and with as little distortion of UV’s as possible.

The model’s surface shader have attributes that give the surface definition and each attriubte has different effects on how the model looks. Texture maps can be connected to these shader attributes or the whole object.

OTHER OPTIONS FOR TEXTURING

We continued from painting in photoshop to painting the surface mesh of 3D models. Polygons and Nurbs have UV coordinates that are part of the surface mesh enabling the use of a 2D UV maps. These maps can be exported to photoshop and used as references to the surface geometry for the painting of the texture.

We touched on the pipeline connecting these photoshop texture files being saved in Maya’s source images and connected to the shaders attributes via the file nodes.

The files can be made through different software programs, include digital pictures and scanned photos.

Maya also has a built in paint effects tool, a rendering effect using brushes that creates strokes to paint textures on the surface of 3D meshs.

PHOTOSHOP DIARY

What creates the texture, what do we respond too? It is manufactured, soft, flexible, light, strong, regular, uniform, matt, warm colour, slightly rough? What is it? My first experiment was adding an adjustment layer in photoshop to this image of Suzie Idiens work taken at the SNO Gallery which focuses on non-objective art. Making colour corrections to the layer or layers below with hue/saturation and seeing if a different mood or feeling to the surface is created.  The original photography is on the left.

compile_001

Texture is something we see and feel. It can have an emotional relationship and/or response that conveys messages and memories. It is tactile, with an appearance and orientation of a physical composition and structure with colour and intensity. In painting and 3D it can be real or an illusion, creating the surface of a visible object with those surface characteristics and appearance.  These can include aspects of its physical composition and structure including size, shape, density, arrangement and proportion.  Textures are a variety of rough and smooth and can be described in many ways such as is it living or manufactured, is it hard, soft, light, heavy, fragile, strong, rigid, flexible, is it smooth, shiny, matt, reflective, dull, rough, sharp, dimpled, cracked, chalky, flaky, what is the colour or pattern. Starting with photoshop brushes we experimented with how they are able to create various size lines, textures and shapes with options for  colour, flow (how quickly the paint is applied), tip pressure and opacity variations.  With the brush tool there are an amazing range of options available for each brush choice. This can be the diameter, spacing, scatter, tilt, angle, roundness, direction of the stroke, the number and position of the brush stroke marks, texture, dual brushes using two tips, using adjustments with foreground and background colours. If this is not enough noise (random pixels), wet edges (like watercolour), airbursh (soft airbrush look), smoothing (smooths arcs) are brush tip characteristics that can be adjusted.

 brushCompile_001Making our own brushes from captured shapes of textured rubbings and scanned images opened up a whole new range of options. This included rubbings we did of the floor, ground, walls, plants, leaves, bark, furniture and objects.  I created this sample of brush shapes from one of the rubbings and made adjustments with shape dynamics, texture, scattering and lots of other sliders, not so technical and lots of fun.

texturingCompile_001 copyTo make our custom brush we made a selection from the scanned files, went into Edit > define brush preset and named the new brush, it is saved at the bottom of the stack in the brushes preset window, no 795.  To manage the brush, such as saving the setting as a new brush select New Brush Preset from the brushes preset window drop down options. The brushes can also be saved by selecting the Preset Manager in the same drop down menu with options to Load, Save Set, Rename or Delete, there is also a reset option to reset the defaults in the same drop down menu or select particular groups.

brushesCompile_002

The same as for the brush, we created patterns from the textured rubbings. By selecting an area of  the pattern, saving it – edit > define pattern a pattern can be created for  use when painting surfaces in photoshop, the selection cannot  be feathered. The pattern can be used for Pattern Stamp tool or as a basis for the Healing Brush (which blends), Paint Bucket and Patch tools. The patterns are selected from the option bar and manage using the Preset Manager.  Options can also be accessed via Edit > Fill, choose pattern and options from the pop-up menu.  There are also other fill options such as Mode, Opacity or Preserve Transparency and it is probably better to do this from the Layers panel giving more flexibility later. The seam lines can be offset with Filter > Other > Offset and offset the edges of the frame in the image area, remembering not to lose the essence of the texture.

I can select my pattern in the texture options of the brush and use this to create brush tip textures with adjustments for the scale and depth. The scattering option varies the values in scatter, count and jitter with the other options available as for all brush tip shapes.  As well as starting to collect textures for a brush and pattern library with these rubbings I was thinking about how they could be applied to surfaces.

MID  TONE  BUILD

Painting timber and wood textures using the mid tone build technique and using new brush shapes.

midToneBuild

Selected Queensland Maple, being a Queenslander, I started with a flat mid-tone colour flooding the surface with the colour. Creating a new layer using the new brush shapes we had created and bigger brush sizes I started to build the colour with darker tones. Also varying the brush, its size and adjusting the mid roundness of shape dynamics to help take away the straight edge. Then doing the same for the lighter layer and varying the colour, pressure, opacity and flow bringing in more of the lighter and darker detail. To create the textures I used  smaller brushes on a new layer, painting in the very dark and light fine details. I was then able to select and distort the marks with the transform option, adjusting the position and shape over the surface. With only part of the surface covered I then duplicated, repeating and repositioning the layer on a new layer to cover the area. To lighten and vary the colour of the marks I used a low opacity and the eraser tool adding to the surface texture.

SOFT  TO  SHARP (Pull Focus)

This weeks exercise was painting smooth and rough stone,                                                                                                          learning about pull focus in our work as we built up the layers from soft to sharp.

softToSharp copy

polished rock

 

For the smooth stone we started with wet general colours, a mid tone, using big soft brushes building soft pale colours using around 30% opacity and maybe flow around 75%. Putting down random shapes suggesting the colour and building the out of focus base.  Increasing the opacity and decreasing the size of the brush in the next layer to sharpen the image.  Varying the thick and thin lines and starting to include some texture , while adding lighter and darker tones to build the mid layer. To develop the point of focus I continuing to work with brush sizes and settings while varying the brush’s opacity and flow to increasing the light and dark fine details showing deeper textures. Continuing with the detail in Shape Dynamics adjusting the Minimum Diameter, Angle Jitter, Roundness Jitter and Minimum Roundness I was able to get closer to a brush stroke that gave the highlighted colour edges and texture to the image. At the end I used the erasor, adjusting the opacity and varying the size to soften the colours and add depth to the textures, giving another quality into the surface.

rockRough

rough rock

 

The rough stone was built in a similar way to the smooth stone using different strokes and blocks of colour.  Again starting with a general base layer of broad colour then working in a more detailed middle layers, adding variation while keeping the stokes soft and adding a little texture without saturating the colours. Starting to include the outline of some of the rough surface shapes with more variations of colour, brush size and the brushes options to introducing more variety than in the mid tone layer. I think there was possibly not enough areas of contrast with the solid colour in the mid tone layers to creating the detailed surface, like cracks and edges and to building up the focus areas. I used scattering more and started to develop areas leading to finer details, looking more at the textures the brushes were creating. I needed to look more at the areas of stronger contrasts, the opacity and the direction of the light with lighter and darker toned areas of the stone that were catching the light and shadows.

BLEND  MODES

Blend modes are how the colour pixels on one layer blend, combine or interacts with pixels in the layer or layers underneath or sometimes above. They are grouped in categories with the general blend modes of darken, lighten, lighting or contrast, invert or comparative and HSL or composite colour. The layer opacity value can also be adjusted in the layers panel, allowing the mixing of the active layer with the layers below in varying percentages with 100% being completely opaque and 0% being completely transparent.  Above are some learning experiments I did with colour, black and white for different mode options.

PAINTING  REALISTIC  FUR

The underpainting textured layer of the fur was done with more of a blunt, stiff brush as opposed to the softer brushes used for the base layer of the timber and stone surfaces. Building the mid layers with a dual brush, the contrasting lines for the fur helped define the surface texture. By adjusting the brush tip and using variations of black and white values with the foreground and background colours in Colour Dynamics it started giving the illusion of depth.  Adjusting the opacity, flow direction, size, shape dynamics and angle jitter direction for a row of hairs and having the brush stroke follow the direction of the fur suggested shadows and deeper textures.

fur

To make singe hair strokes on the finer detailed layer I worked with brush no 60.  Adjusting the size and angle to improve the stroke edges and spacing to give a continuous line. In Shape Dynamics adjusting the minimum diameter worked to have the ends of the stroke pointing, not a blunt square and I left the angle and roundness jitter alone as this took away from the straight edge. Now I am able to continuing building depth and focus with the single stroke, varying the colour , having control of the brush direction and pen pressure to help with the stroke.  For the detailed layer there was more focus on the random hair detail, to look more natural as nothing in nature is repetitive.

fur_07

 PAINTING  REALISTIC  SKIN

skinStart

For the skin we continued as we have been doing by starting with a base tissue colour layer or fill layer.  This is a solid colour and the next layer builds the varying colours by picking different values and tones to create the random contrasting colours. Lowering the opacity and pressure on both the brush and eraser tool to varied and soften the surface.  On the third layer the stronger variations of light and dark on smaller areas were done to be able to use bevel and emboss and give the visual impact of depth and height to the surface. Layer > Layer Style > Bevel and Emboss. This gives options for soft or hard edges, how raised or sunken the edge of the bevel or pattern appears, the direction for adjusting the position of the highlight either along the light source or the shadow near the light source, size and soften for bluring the shading of the effect and contours can be adjusted and textures added. Expanding on this we painted the veins in a couple of colours with different brush sizes and applied the bevel and emboss to show the veins on the skin.

skinMid

The next layer introduces a lighter colour using a screen mode, lowering the opacity and taking into consideration the race and skin colour. When this is applied to the layer it lightens the image by multiplying the inverse of the blend and base colours, screening with black leaves the colour unchanged and with white produces white, the opposite to multiply.  It also reduces the contrast of the overall colour tending to flatten the image. Using the spattery brush with emboss selected and depth down, skin pores are created on a new, more detailed layer.  This layer can be duplicated to create sun damage with softer and darker areas and then adding freckles for more detail. skinLayer_05 skinDetail_06 I am yet to work on Barry Deans secret of painting skin with a number of thin, transparent layers to created the subtle variation of surface texture and to see how this looks giving the skin the translucent and alive look he refers to in his notes.

STILL  LIFE  STUDY  –  APPLE

apple

I starting with the sketch layer while thinking about the apple’s placement with reference to the objects point of view, the setting, any interesting points of reference and the lighting. Applying the layered working structure we covered over the last few weeks there are the sketch, under paint, mid-pass, refine, detail, highlight and shadow layers to create the apple and surface texture.  For the underpainting we chose two or three colours from the colours on the apple surface, the brushes do not matter much and slosh down the colour breaking up the surface with solid colour.  The mid-pass breaks up the surface a bit more, bridging colours and starts to work the contour of the surface, called cross contouring. Follow the direction of the form from the stalk to where it disappears at the bottom and using slightly lower opacity and soft edges on the brush to continue building up the apple’s texture.  The colour is also stronger in the middle with softer edges, blend by picking soft colours with bridging colours, a couple of variations of yellow and red with some apple green mixed in.  Moving to a smaller brush, modifying it by adjusting bristles, thickness, softness and stiffness to start building more detail. appleDetail_04 To add strokes to the surface we used the brush a bit like a rake, keeping the pen vertical and varying the opacity and pressure. Then working with a finer brush to build and refined the detail, using colours and shapes while looking carefully at the surface to see what it looks like. Now I am starting to put in the shapes that I see on the apple, becoming more random and spending more time on these layers.  Keep looking carefully at the highlights and break the shape up a bit, thing about the shadow areas, the stalk,  imperfections, spots and shadows with grey to give volume and bring the opacity back up. appleFine_05 Select the white background and paint in with graduation and shadow under the apple. For the grey background put the layer mode to multiply and opacity to 50%.  To clean up the edges, shift select all the layers while hiding the layers that are not wanted. Select the top empty new layer, above what is to be cleaned and flatten and apply what is visible – Image > Apply Image. Thanks Barry for opening another door, now I am looking forward to learning more about photoshop and including it in as part of my knowledge.

RESEARCH

“Architectural Material,” n.d. http://docs.autodesk.com/MENTALRAY/2014/ENU/mental-ray-help/files/shaders/architectural/arch_mtl.html#mi_shader__mia_material_x.

Autodesk Maya Techniques: Hyper-realistic Creature Creation. San Rafael, Calif.: Autodesk, Inc, 2007.

Birn, Jeremy, and Jeremy Birn. Digital Lighting and Rendering. Indianapolis, Ind.; London: New Riders ; Pearson Education [distributor], 2006.

Brown, Blain. Motion Picture and Video Lighting. 2nd ed., New ed. Amsterdam ; Boston: Elsevier/Focal Press, 2008.

“Common Techniques to Improve Shadow Depth Maps,” n.d. http://msdn.microsoft.com/en-us/library/windows/desktop/ee416324(v=vs.85).aspx.

Demers, Owen. Digital Texturing & Painting. Indianapolis, IN: New Riders, 2002.

Derakhshani, Dariush J. Introducing Autodesk Maya 2013. 1st ed. Indianapolis, IN: Wiley Pub., Inc, 2012.

Gyncild, Brie. Adobe Photoshop CS6: Classroom in a Book. Classroom in a Book. Berkeley, Calif: Adobe Press/Peachpit, 2012.

Kermanikian, Ara. Introducing Mudbox. Sybex Serious Skills. Indianapolis, Ind: Wiley, 2010.

“Mudbox Online Help Files,” n.d. http://docs.autodesk.com/MUD/2015/ENU//index.html#!/url=./files/GUID-0DF909C0-A02A-4729-A285-F7E5F40BD71A.htm.

Obermeier, Barbara. Photoshop CS5 All-in-one for Dummies. For Dummies. Hoboken, N.J: Wiley Publishing, 2010.

Palamar, Todd. Mastering Autodesk Maya 2014: Autodesk Official Press. 1st ed. Indianapolis, IN: John Wiley & Sons, 2013.

Präkel, David. Lighting: N. Light of a Particular Quality or the Equipment That Produces It. Lausanne; London; New York: AVA Academia, 2007.

<www.sophia.org/tutorials/elements-of-art-texture> http://msdn.microsoft.com/en-us/library/bb976075(v=snagamestudio.31).aspx NVIDI Corporation, Mental Ray Architectural and Design Visualization Shader Library, 1986, 2011

Elliminate Texture Confusion: Bump, Normal and Displacement Maps

THE BEGINNING of EXPLORING PROCEDURAL PROCESSES

Screen shot 2013-03-31 at 4.39.18 PM – ramp and ramp shader I discovered are not the same thing

– with the colour shader selected in AE, clicking the map button the book asked that the Normal is checked – where and what is that?

– still getting my heard around projects, scenes, new and set when setting up and working in Screen shot 2013-03-31 at 4.44.30 PMscenes and projects

– 3 hours to do the wheels on the wagon from the book, the hypergraph and AE, it is becoming more familiar in my brain

Screen shot 2013-04-01 at 12.28.35 AM– doing the wagon strips with the UV texture editor and photoshop is a totally new experience, going slowly and taking time over a couple of days to read through it

– do not understand how the selected to laid out in the UVTE. I was not able to work out the stripes and logo for the side of the wagon

– Photoshop, now to remember and learn more, do not understand what setting to use

Texturing, from the book using a photo converted to iff in photoshop and creating a photoshop file from texturing>create psd network.Screen shot 2013-03-24 at 9.53.48 PM

The first box is a polygon, second is a nurb, third is using projection, next one I was attempting to access the Place2dTesture node and did not work it out and the last one is the photoshop file form create psd network and did not work out about the transparency or bump nodes.

Texturing with layered shaders, from the book – 2 hours

Screen shot 2013-03-23 at 8.54.30 PMI was introduced to the ‘Fit to bBox’ and placement nodes for the wooden handle, then the layered shader for the metal spike. I’m still working out what nodes do what, how and why they are ordered as they are, how I access and order them in the hypershade and the basics of what they do in the attribute editor. Not clear on the order and process of the layered shader. Could not work out the baking a texture, kept coming up with ‘wood2 is connected to more than one shading group’.

 

HYPERSHADE and LIGHTS

Questions – straight edge on ramp when interpolation is none, when creating a hole and joining the edges does this impact on the ramp lines for colour

– can there be a u and v ramp on the same object

scene has 3 default shaders loaded,  textures and shaders to give objects their visual appearance, shader networks are made Screen shot 2013-04-24 at 11.29.16 PMup of render nodes, node attributes in the AE, shaders dealing with specular and diffused Screen shot 2013-04-24 at 11.07.43 PMparameters together with lighting

– lambert evenly diffuses and scatters light evenly for dull or matte surfaces eg paper

– phong specular highlights and reflectivity creating sharp hot spots that drop off sharply eg plastics, glass, most metals

– blinn accurate specular lighting that reflects light and diffuses more gradually than a phong eg shiny and metallic surfaces

Screen shot 2013-04-28 at 9.01.34 PM– phong E more control over the specular highlight to adjust the glossiness of the surface. The specular drops off more gradually and yet remains sharper than a -Blinn giving more options for metallic reflections

– anisotropic reflects lift unevenly creating irregular-shaped specular highlights deformed surfacesScreen shot 2013-04-28 at 9.02.54 PM eg foil or plastic wrap, directional grooves

– layered stacking of complex effects, use transparency maps to define areas eg labels, dirt to aged surfaces

– ramp smooth transitions between colours and to control paths with ramp texture nodes attached to the attributes colour, transparency, incandescence, special curve ramp

– lights, the light coming from the lamp was a challenge, not sure about the white on the wheels and the highlights and shadows could be more subtle.

Texturing for 3D Short Answer Exam

Question 1: Colour 

Explain the following words in terms of how they define colour:

Hue (2 marks) or pure colour, what we commonly refer to as colour, the names we are familiar with and thehexacone-5 characteristic that distinguishes one colour from another, red from yellow. The colours on the colour wheel are three primary, three secondary and six tertiary. These colours are additive,  created by light, pure RGB = white light. There are subtractive colours made by pigments which are RYB. The reflected light from a surface creates colour with all colours being absorbed you get black.

Saturation (2 marks) Also known as intensity, saturation describes the strength or purity of a colour. The ratio of a colour to white is a measure of the colourʼs purity with pure colour being fully saturated. Colour can be taken out leaving a black and white version of an image.

Value (2 marks) or brightness of a colour, how light or dark it is from black to white, adds values for depth relative to lightness and darkness. Light colours are high value and dark colours are low value and can lead to complete black or white based on the amount of light emanating from the colour. It can separate object in space.

HSV: Value 0 to 1 in the HSV slider goes from black (0) to white (1) When the value is black there is no colour even when changing saturation and/or colour. When it is white colour can be added with saturation from white to pure colour.

Question 2: Colour 

Colour – Discuss colour temperature, by giving a brief description of what is meant by ʻcoolʼ or ʻwarmʼ colours.

Colour temperatures are measured in degrees Kelvin, the unit of absolute temperature which originated from the heat glow when a block of carbon was heated and this is refereed to as colour temperature. With black absorbing all the wavelengths, it produced a range of different colours at different temperatures and is describing the colour characteristics of the light. The Kelvin scale starts at ʻabsolute zeroʼ.

Kelvin scale can range from the flame of a candle at around 1800K and the deep blue sky at around 10,000K with average noon daylight being 5000K or white light.

The colours start to become bluer at higher temperatures, over 5,000K and the cool colours are the lower temperatures increase the orange and reds.

We interpret colours using them to describe our world with emotion, feelings and our senses.

The cool colours of green, blue and purple recede, often being used for shadows and helps to place and anchor objects in the environment. They are also the colours we associate with night, water, nature, calming and relaxing.  The warm colours, reds, yellows and oranges advance in our perception. Warm colours can be cooled down by adding blue or made warmer by adding yellow.

Question 3: Photoshop 

Describe how to create an original brush in Photoshop and how to save and manage that brush.

An original brush can be created from a selection of images such as photographs, art work and texture rubbings that can be scanned, created or imported to create the brush. The image can be colour though when saving the selection in photoshop as a brush photoshop converts it to grey scale. The white pixels become transparent, 0% opacity and the black pixels become opaque, 100% opacity. The image can be adjusted or cleaned  before saving using a selection of options with adjustments under the image menu.  This include options such as levels, curves, exposure, shadows/highlights and brightness/contrast.

One way to make the brush is by making a selection using one of the selection tools from the tools panel.  With the right amount of contrast make the brush,  edit > define brush preset and a box appears, name the brush and the selection is saved in the brushes preset palette at the end of the list.

Brushes are saved using the preset manager, under the edit menu, which is used to load, rename, delete, organise, and save custom content and presets for brushes, swatches, gradients, styles, patterns, contours, custom shapes and tool settings.  Select the brushes, go to the preset manager and save (set) brushes, choose the location to save the brushes and they will be save as an .abr file. They can be saved in libraries then reloaded through the preset manager for later use.

To modify brushes use the brush presets under the brushes tab where there is access to the options for altering the appearance of brush strokes. There are adjustments for options such as spacing which can create dots, a dotted line or several strokes rather than the seamless line. Shape Dynamics can control the size, angle and roundness of the tip. The bottom panel displays a preview of what the brush stroke currently looks like.

Question 4: Photoshop 

Name two layer blending modes and describe their blending effect.

Blend modes are how the colour pixels on one layer blend, combine or interacts with pixels in the layer or layers underneath. They are grouped in categories with the general blend modes of darken, lighten, lighting or contrast, invert or comparative and HSL or composite colour. The layer opacity value can also be adjusted in the layers panel, allowing the mixing of the active layer with the layers below in varying percentages with 100% being completely opaque and 0% being completely transparent.

The multiply shading mode when applied to a layer the areas of pure white disappear from view, everything else becomes darker and areas of pure black remain black as you cannot make it any darker than it is. The colour of the blend layer blends with the layer below and the colour becomes darker as it multiplies the base colour with the blend colour. This results in a darker colour unless the blend colour or sections of the image are pure white resulting in completely transparent and the layer below becomes visible through the top layer or 100% opaque black multiplied with any colour will result in black, grey scale has partial opacity. The multiply mode is looking at the colour information in each channel and multiplying the base colour by the blend colour. Adjusting the opacity option is also calculated and taking it to zero for the blended layer leaves nothing to blend with and the image for that layer disappears showing the layer below as normal. You can duplicate an image and if they are Multiply Blended they will become darker, if it is too much slide the opacity slider back to 50%.

Overlay shading mode is part of the contrast group, the tonal range of an image shifting the pixel values. Adjusting contrast intensifies the values of the saturation of the image. The mode multiplies dark areas making everything darker and screening light areas at the same time making them lighter and losing the gradations of colour. The layer with 50% grey completely disappears from view thus boosting the image contrast and losing the mid tones. The black does not stay black and the white does not stay white from the top layer with the Overlay mode selected as more importance is given to the Background or underlying layer. When I put the black, white and grey on the top layer and a colour on the bottom the colour comes through losing the black and white. When I switch the layers with the colour on the top and the black, grey and white gradation on the bottom or underlying layer then the black and white stay black and white, the colour is lost and the mid grey becomes the colour.

FROM  DIGITAL  MEDIA  COURSE

DIGITAL  IMAGES

colourCorrestion_compile_001

Colour Correction: Curves under Image > Adjustments > Curves, adjusting the elephant’s original photo on the right by adjusting black, white, red, green and blue individually. Using Hue and Saturation to change the chair colours from red to pink, yellow and blue using Image > Adjustments > Hue/Saturation.

colourCorrect_compile_002

Colour Correction: Image > Adjustments > Hue/Saturation and change the selected colour which is the red to green.

colourCorrect_compile_003

Colour Correction: Image > Adjustments > Hue/Saturation and desaturate to create the black and white with some colour kept in the image.

colourCorrec_compile_004

Colour Correction: creating a stronger image under Image > Adjustments > Brightness/Contrast.

colourCorrection_levens_hue_saturation-003

Colour Correcting using levels to clip the colours and adjusting red, green and blue individually.

Take a black and white photos and colour it.

This was great fun. I had a problem with the copy and paste as the image was not directly in the same place. This was when the selection was made, using the layer mask option and reselecting the standard mode then needing to copy and paste the selected image into a new layer  Before using the magic wand to select and then inverse the image before colouring. The last image in the about sequence is an example, I moved the image to position it correctly.

SHOOT  MY  OWN  SHUTTER  SPEEDS

EXPERIMENTS  WITH  SHUTTER  PRIORITY  MODE

Nikon Electronic Format (NEF)

Exclusive to Nikon cameras, the NEF is Nikon’s RAW file format. RAW image files, sometimes referred to as digital negatives, contain all the image information captured by the camera’s sensor, along with the image’s metadata (the camera’s identification and its settings, the lens used and other information). The NEF file is written to the memory card in either an uncompressed or “lossless” compressed form. 

The primary benefit of writing images to the memory card in NEF format rather than TIFF or JPEG is that no in-camera processing for white balance, hue, tone and sharpening are applied to the NEF file; rather, those values are retained as instruction sets included in the file. You can change the instruction set as many times as you like without ever disturbing the original image’s RAW data. Another benefit of the NEF file is that depending on the camera, it retains 12-bit or 14-bit data, resulting in an image with a far greater tonal range than an eight-bit JPEG or TIFF file.

After-capture processing of the NEF file by Nikon’s Capture NX2 software, or other imaging programs, offers greater control over the final image than the processing of a JPEG or a TIFF. After processing, the NEF file can be saved as a TIFF, JPEG or again as a NEF with the addition of any applied Capture NX2 processing saved inside the file as a second or alternate instruction set. As long as the original NEF file is preserved, the “digital negative” remains untouched; processing a NEF file does not alter the original instruction set.

What are the differences between: RAW, NEF, Compressed-NEF, TIFF, and JPG file formats?

Nikon calls images saved in the RAW format “NEF” files. Nikon RAW NEF files can be edited in Nikon View Editor, PictureProject, Nikon Capture Editor and Nikon Adobe Photoshop plug-in.

This unique format consists of the RAW data of an image, along with an instruction set that provides extensive image editing capability not available with other file formats.  With a NEF file, the original RAW data of an image is never changed.  All corrections and adjustments that you make are preserved in the file’s instruction set. You can change the instruction set as many times as you like without ever disturbing the original image’s RAW data. Using the software listed above you can change the shooting White Balance, adjust Exposure Compensation as well as basic color, sharpening and levels controls.

Current Nikon DSLR cameras, including the D3-series, D2-series, D700, D300(S),  D200, D100, D7000, D5000, D3100, D3000, D90, D80, D70s, D70, D60, D50, D40X, D40, all support the NEF RAW file format. All future Nikon Digital SLR cameras will support the NEF RAW file, and some Coolpix cameras will as well.

Many users think of their NEF files as their original digital “negative” which they then make changes to and save the changed files as TIFF (or JPEG) for printing.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: