Thursday, 29 October 2009

(Group Project) Week 2 - Presentation week

*This is an individual post as a response to the progress up to and after the Presentation on the Thursday*

Aaron and Richard began their roles by creating the wooden pillar stands and beginnings of the pump's floorplate and the modelling of the 24" diameter flywheel, essentially in preparation to display them on the Sildeshow to give an example of early progress and design intentions


I decided to create a crucial part of the pump, albeit smaller in scale than the other members sections - by creating a simple nut-and-threaded-bolt that can be edited for future use, it is also useful for 'Cloning' and 'Replicating' it throughout the model's various stages



 To design the nut, it was a case of creating a 'Cylinder' with a radius of 2.5cm (remembering to double-check the parameters of the Max file before I created it), height and cap segments at 5 and reducing the sides down to 7, 'Morphing' the cylinder to this form here

 

Under 'Extended Primitives', there is the object classed as 'Hose', which I knew would be ideal from which to create the threaded bolt
Various parameters were altered including the percentage start and ending falloff of the spiral-hosing and I needed to modify the width and sides of the hose shape so that it would scale up against the nut



As I had kept all scaling close, all I needed to do next was to place the threaded-bolt against the nut as displayed here - in the top right window, both are shown with the same amount of sides, which was intentional in order to maintain the form of the objects when the bolt is needed to 'Boolean-Out' the nut

The Presentation

As a whole group, we all had a part in describing and justifying our progress and design decisions since we visited the Museum

One particular slide we included was to list the possible problems that we may come across during the modelling process: items such as keeping to the set scale, motion capturing, as well as the general quality and timing of the animation in keeping with setting created

One aspect that we focused on afterwards was further research into the past uses of the steam pump and its assigned locations throughout its working life

Some sites of interest include:  

http://www.steampumps.net/JosephEvansReliable.html


http://www.bedwascolliery.co.uk/contents.html


http://www.welshcoalmines.co.uk/forum/read.php?4,3832

Saturday, 24 October 2009

(Group Project) Week 1 - Our Plans and Decisions

Having been assigned to a group with three other MM students (Aaron Cracknell, Richard King and Stuart Norton), I and the others went to the Museum of Power at Langford, Maldon

During our guided trip around the place, we came across a particular steam pump which was of interest to us straight away - it was bright red and make a fantastically-steam-age antique racket...



The Joseph Evans Reliable Steam Pump

^The information board in front of the pump - a pump bought for £10 in 1987!^


 
 
 

Critical Measurements

- The pump is 4 feet long, 2 feet wide and 4 feet high

- It has a 4" steam cylinder, 2" pumping cylinder w/ 4" stroke and a 24" diameter flywheel (double acting)


Our first group meeting will decide each of our roles and most importantly, choose a scale which all four of us will work from - be it US/UK Metric, Imperial or ISO (doing this will then narrow down the chances of mismatched measurements and improper scaling!)

A Powerpoint presentation will be undertaken in the 2nd week which will present our choice of machinary, roles posed and some early design decisions

Wednesday, 21 October 2009

Week 5 Part 2 - A Reflection (no pun intended...)

 This is a post describing the trails and tribulations (and occasional moments of honest achievement!) throughout these hectic 5 weeks, also including those final screen-shots describing the mapping of materials for the skull and neck areas and a small environment adaptation before rendering...




(Just a quick update of how the front-view of the face looked mid-way this week)
I was pleased that I was able to get to this point before carrying onto the next section and not spot any pinching or discordance with the material-mapping itself



I had to adapt the 'Multi-Sub Object' modifier's default settings when using it on the skull and neck
As they were considered, by Max anyway, to be one single object, I needed to find a way in which the 'Multi-Sub' could distinguish its 'ID Tag' setup with the one given to the skull/neck object (this was classed under 'Polygon Properties' as 'Material' 'Set ID' 1)
And so, by way of going down into 'Polygon' level amongst the Modifier List, I could highlight the sections of the skull I wanted - up until where the next was extruded where I could use another material...



When the skull was highlighted, I specified that it be classed as 'Material ID' 2 and assigning this comparative to the settings on the 'Multi-Sub Object' modifier
When highlighting the material needed for the neck, this was classed as 'Material ID' 3 and again, this was distinguished within the parameters of the 'Multi-Sub Object' modifier

(The materials themselves were just small thumbnails grabbed from the original reference PNG that is used on the front-view of the face - a section of hair just above the fringe and a skin tone taken from the neck area!)


 

Once the 'Multi-Sub Object' was primed and ready to be dropped into the scene, all I needed to do was to advance back up the 'Modifier List' to remove the 'Polygon Selection' red-highlight and make sure the 'Material IDs' were correct
Once the material was applied to the 'Skull_Object' as a whole, the requeset areas went to the correct materials - the benefit of the 'Multi-Sub Object' modifier without needing to 'UVW Unwrap' for just two simple, continously-tiled textures
 The final application of a 'UVW Map' meant that both material would need to comply to the settings but this wasn't as much of a problem as I first thought (the hair itself when under the mapping of 'Box' actually creates a semi-realistic feeling of natural flow of a hairstyle, I feel)

The Conclusion


The whole process over the past 5 weeks had been a very rapid learning process. Of course, Max itself has a fairly steep learning curve (what 3D modelling program doesn't!), but it rewards users who devote time and attention at every step a enhanced end result


Of course, from the very beginning, my original reference photographs were not as 'straight' as they could have been - the front-view certainly- and the resulting topology was rushed (even though, by my own addmission, I felt it would suffice according to possiblities of modifications later on!)


The next step of extruding the face itself didn't confuse me as much as it should have, but I certainly became aware of my inaccuracy of the scaling of the reference images when trying to gauge how far out the material would be extruded and discovering the gridlines for the nose meant I couldn't rely on either 'plate' too much without measurements going awol quickly...


It was this first stage which was to cause me to run into many concurrent difficulties, many of which have been expressed to a fair degree over the half-dozen or so posts over the weeks - it is such a dissapointment that due to the problems, the model itself isn't as I hoped it would have been (I intended to animate it and get it to speak subtle praises about its designer!)


The one aspect I'm glad worked out much as expected (what with the scenario I was experiencing where I had limited resources to work with) was the facial details I captured with the material - the UVW Unwrap tool was invaluable and, although it perhaps wasn't used to its fullest extent, the tool still meant that I could retain that extra nuance of personality into the model (especially with the inclusion of details that I spent a good deal of time on: the eyeball textures and faux-set of teeth visible behind the mouth itself!)


Given more time, it would have been entirely plausible for me to create an entire UVW Unwrap of the complete skull and features from reference images that I would spend more time on getting as accurately-scaled and well-defined as possible and, ultimately, creating a 3D model that would have been such an accurate and life-like representation of myself, that it would be a real worthwhile achievement in a timescale of just over a month!



Oh yes, here is the final rendered image - the background colour tint relates to improving definition aroudn the edges and my personal colour preference that is evident in the initial reference images...






Week 5 - Wow, its like looking in the mirror(!)

 The final week and I plan to sort out the UVW UnWrap/Mapping to give my facial detailing that final expression of realism and, comprehensively, the final step in creating a 3-D Splined Model of my head
  
This post will describe the complex (and occasional backtracked) method which I used to achieve this 

*As described in the first paragraph of the 2nd Part of Week 4's post, you will notice that the first half-dozen or so screen-shots taken of the modelled head in Max will contain no rear skull as I had begun to experiement with the UVW Mapping a couple of days before the skull/neck was constructed*




I wanted to see how just using an edited version of my original JPEG reference-image would work when applied to the model itself and so I opened it up in Photoshop CS4 and began to get rid of the parts of the image that I didn't need
Using the 'Quick_Mask' layer-type, I was able to 'brush' in the the areas of the image that would be removed and use the 'rubber' to retain areas that I wanted left alone - as I would be applying the image to just my face and not a skull/neck, I only needed this amount shown here



 Once PS had removed the material (done after exiting from the 'Quick_Mask' layer-type, visualising the 'Lasso' tool which shows the highlighted area and then hitting the 'CTRL+X' keys and then pasting with the 'CTRL+V' keys - PS will automatically add this 'pasted' material onto a new layer
I also needed to remove the interior eye detailing so the the map could be placed onto the model and allow space for the present 'modelled' eyeballs  - this was achieved within the 'Quick_Mask' but I had to use the hotkeys of 'CTRL+SHIFT+I' to inverse the selection and make sure I deleted only the interior material - the file was saved as a PSD format to retain the information and layers for potential re-editing

 

Having finished with Photoshop for the moment, I went back into Max and opened up the 'Material Editor'
From there I assigned, under the 'Diffuse' channel, the 'PSD' that I had saved of the edited reference image imported as a 'BitMap' file and then applied this to the 'Splined_Face' object which presented it as shown
The 'UVW Mapping' modifier, as a Default setting, incorporated it as a 'Planar_Mapping'-type which allowed me to gauge how the material would interact with the model without any alterations from its original reference



 I was now ready to assign the 'Unwrap UVW' modifier which allows the user greater flexability when modifying their 'UVW Mapped' materials'
The image shows how, after applying the modifier to appear at the very top if the 'Modifier_List', it is possible to select the vertex, edge or face of the object the modifier is assigned to - here I have highighted how its possible for the 'Select' tool to capture all faces of the model all at once (ease of re-scaling is a primary use for this feature)
The image in the window on the left is the 'Un-Wrapped' version of the 'Splined Model' and displays it as if all faces are aligned along the 'XY' co-ords

 

Now that I had the 'Unwrap' modifier together with the 'Mapping' modifier, I was able to re-positioning more effectively and there were customisation possibilities by the combination of the 'Mapping_Gizmo' and the various options in the 'Parameters' menu-tab - here I'm able to move the material along all 3 co-ords independantly with the 'Gizmo'
It was at this stage where I was able to see how the BitMap would fit alongside the features and forms around the face itself



The final aspect I was to experiment with, before the rest of the model was completed (my skull/neck attachment primarily!), was combining the selection of the vertices under the 'Unwrap' modifier and the 'Gridline_UVW_Unwrap' window's modifying abilities which allowed myself a chance to re-scale, rotate and manipulate the 'UnWrapped' material
The 'Gridline' Window itself is set to display in a 'checkerpattern by default so to bring up the material the user wants to work with, its a case of selecting the drop-down menu in the top-right of the window and choosing the 'Pick_Texture' which brings up the folder option to pick the 'BitMap', which is then imported directly onto the background of the gridlines



Try as I might, the quota of success from trying the above method wasn't getting the material where I wanted it to be, no matter how I adpated the 'Unwrap'
So, I decided to return to Photoshop and edit the BitMap file one more to get another chance at 'Mapping' it with greater accuracy and definition
Basically, all that was needed was to divide the one image up into three seperate sections, saving them as threee seperate PNG files so that I could incorprate them into Material Editor using another techique I picked up over the course of learning Max



Returning to Max, I went into 'Material Editor' yet again
From here, I proceeded to select a blank-slot, click on the 'Standard' icon next to the 'Material' drop-down menu, access the 'Multi-Sub Object' parameter and then be greeted with the details shown in the window of the image above
The 'Multi-Sub Object' option allows the user to import and equip several materials into the 'Editor' (selecting their choice from each 'ID-Assigned' channel under the 'Sub_Material' icon), which correspond to the relevant areas of the object on which the 'Multi-Sub Object' will be placed




In this context, all I required for the 'Multi-Sub' was for it to choose a concurrent ID tag for each of the three materials I had imported (the three seperate images I had saved in PS according to my facial features and named accordingly, if you remember...) and progress onto using the 'UVW Unwrap' modifier to adjust the positioning of the 'Unwrapped' material to line up with the three materials making up the 'Multi-Sub Object' -
the image here showing how the 'Unwrap' drop-down menu allowing the user to register each map, selecting each one and adjusting the mapped co-ordinates before moving onto the next



The process was seemingly working as I had hoped, with the top section material clearly lining up pretty much exactly with the model itself, with a little manipulation needed to stretch the 'Unwrap' to make sure it followed the material closely underneath
There was indeed the small matter of the top half not applying anywhere near my forehead/hairline, as well as the other two material seemingly M.I.A, without any clue as to where there were as part of the 'Multi_Sub Object' modifier
Thoughts returned to the PNG file from which I had imported as a BitMap for the 'Multi-Sub': perhaps the white background that had to fill up the gaps between the three seperate sections of the image (a trait of the PNG file format) was actually interfering with Max's application of the 'ID_Dependant' ordering...?!




Once more into Photoshop, I then had the notion of reducing the gaps between all sections (named Top, Middle and Bottom) physically with the cursor-control, but also by reducing the 'Canvas Size' by 20 or so cms, thankfully avoiding any collisions by making only minor incremental adjustments at a time...
 



If anything, the adaptations used up alot of time and almost worsened the situation as, shown here, the materials were now distorting and not applying to the model anywhere near as well without the editing (there was no possible way to manipulate the 'Multi-Sub Object' modifier any more so I dropped it altogether)



This is where the backtracking came into play - I regrettably brought the complete, single-piece PNG reference file back into the 'Material Library' and postured myself into making greater use of the 'Unwrap UVWs' Editor
By minimising each change and assessing the visual changes they brought about with primary adjustments to the scaling and co-ordinant positions,I was able to get the model to show promise as the example here defines



This is a prime example of the intricacy of this method of applying and effectively 'Unwrapping' a UVW even as reduced as my frontal reference shot - here, I am moving the 'Vertices' around the eye so that the white background of the PNG file can be hidden behind the eyeballs themselves!
The MESH itself is able to be manipulated and forced to follow the material clearly displayed underneath - there is only so far that the MESH can be moved around before the material begins to present 'Uneveness' and 'Patching' where it has to overlap or contacts with the other areas of the 'Map' itself



 Gradually, the step-by-tentative-step manipulation was beginning to take shape, as I now had the whole MESH overlapping the material underneath with (almost) all extraneous points-of-reference containing some material data
Here I am reducing the height of the forehead section using the 'Scale and Reposition' tool within the 'Unwrap UVWs' window to allow the fringe to cover the white background of the reference map

 

 This final step was made easier by introducing two parameter options under the 'Unwrap UVWs' modifier tab, both are labelled under the 'Display' sub-menu: one called 'Show Pelt Seam' and the other 'Show Map Seam' - they meant I could ascertane the positioning and border edges of the model, avoiding any problems with the MESH running over the sides and causing more distortion
With these options selected, when I was straightening up the centre dip below my lower lip and the central-nostrils section, I was able to maintain a greater degree of consistency throughout the whole frontal section


Now all thats needed this week is an application of a basic material around the neck and over the skull - at least I'm sure the 'Multi-Sub Object' tool will be useful during that final stage

*The 2nd part of this week will conlcude the whole 5 weeks and give some idea of the highlights and lowlights of this complex process, as well as detailing thje possiblities given extra time and also the final rendered image to base final judgement on*

Monday, 19 October 2009

Week 4 Part 2 - Skull and (Really!) Crossbones

You will notice that I have an image of my reference photograph (minus the topography) on the 'Mesh_Model' of my face during this post - this was only due to me experimenting last this week to see how the image was to look before my backtrack into creating the skull and neck area
Of course, this will also mean that when I present the post for Week 5, the images that show how the first application of the reference images and the progress from then on will be divided between screen-shots of me with no skull behind and some that will indeed show me with skull and neck present - hope that makes sense!




The first step that was needed to begin the formation of the skull was to create a 'Sphere Object' and convert it into an 'Editable Poly' This allow me to select the vertices that make up the Sphere itself and compress in the sides so it becomes almost eliptical in shape



( I know, it looks as though I'm a Jet Fighter pilot with a photochromatic visor on - good choice of object-material colour eh)
The image shows that after modifying the sphere into the eliptical form, I then moved the soon-to-be 'Skull_Object' along the 'Z' axis so that the outer edges made contact with my temples and the main central-bulge of the object appeared to collide into my eye-area
I used the HotKeys of CTRL and 4 to see the object in a wireframe-mesh form for ease of manipulation and guaging the formation of its Polygons






 Yes, the reasoning behind the positioning of the 'Skull_Object' shown here is so that when the object's Polygons are clicked and dragged (much easier when the shape of the highlight cursor is changed from a regular square shape into a circular-shaped cursor), it it much easier to select the ones needed to be deleted, rotating the camera around as much as necessary in order to get the ones required (all those in red here!)
Here, I needed all Polygons that covered my face, as well as the first two sections around the objects central emitent-point
 Another tip I use when doing a step like this is to check the 'Check Backfacing' option so that when selecting the Polys that are needed, you are not constantly de-selecting un-needed Polys that just happen to be in the path of your wandering cursor!



 Once I was happy with the Polys that I had selected, all I then needed to do was hit the 'Delete' button
Usually, doing this to a "solid-face" object causes the outer edges to lose their formation at this point, but this deletion was crucial to following the most efficient techniqur of creating the skull and neck area
There is of course a gap of the lower half shown here, but material will be extruded and this will be joined later by means of another previously used technique

 

 I was needing to use the 'Snap_To_Grid' tool as I had during the formation and 'Welding' and my lips
Initially, the tool wasn't recognising that when I highlighted the vertex on the outer edge of the skull (the hairline area) I wanted to 'Snap' to the requeset vertex that was part of the facial hairline - the skull vertex attempted to attach to where two vertices where attached somewhere else on the model!
I thought of a logical solution, which brought me to going into the 'Grid/Snap Settings' where I noticed under the 'Grids' tab, 'Grid Points' was checked, whereas I needed the 'Vertex' option checked
Once I had done that, all that needed to be done was to go into the 'Skull Objects' 'Editable_Poly' sub-menu and make adjustments to the amount of segments there were, just to make sure that the topology/grid of my face model would match up to the grids along the skull itself - makes the next step easier too...!




Once all requeset vertices were attached to one another (I did need to alter the positions of the outer vertices of my facial model as I had them curved too far around and the outermost vertex of the skull couldn't attach without pinching inwards - inherent part of the process when I wasn't planning to re-create the skull!), I could then move onto joining up the lower half of the skull to my chin area
This was, of course using the 'Scale' tool once again - the tool allows the user to select the edges they want extra Polys to be generated from and once pulling them out when holding the CTRL and left-click keys, extra material will be pulled away along the axis of 'Z'  This was I could re-create the lower skull and gradual curvature in the nape of the neck too



After using the scale tool to pull down the Polygons further towards the chin area, I reached the point where, as the image above shows, I couldn't match up the amount of internal edges of the gridlines of the extruded material to the gridlines of the facial-model!
An oversight that was certainly attributed to the discordant setup of the original topology and my own modifications when creating the mesh, I needed to use the 'Snap_To_Grid' tool on vertices that wouldn't have been suitable from one part to another, as displayed here
 


 The next step was to take the edges of the side-portion of the skull and 'Scale' them towards the adjoining (hopefully) edges of the face-model, taking into account that there would need to be a fair amount of adjustment after the extrusion took place
As shown, I decided to pull out a single 'Scale' row of Polygons as I could add in the opposing edges with the 'Cut' tool afterwards



Unfortunately, upon reaching the top part of the skull during the 'Scaling' process, I came across this complication - how was I to manipulate the edges to follow the designated-curvature of the outer edges of my face when all the Polys wanted to do was angle outwards and not snap-to the pre-determined vertices I was trying to get them to attach to...
I overcome this issue by rotating the edges once attached first, to each other, and then created a Poly which then had to incorporate all 4 vertices of the two edges into the facial model chosen by the 'Snap_To_Grid' tool



This part really was taking alot of time up, with each newly assigned vertex and Poly needing to attach somewhere, with many causing a 'Collision' with the modified 'Face Object'
I decided that I needed to cut up the large expanse of Polygons by using the 'Cut' tool, thereby allowing me to adjust the flow of the Splines of the skull in a more competant manner, as well as inherently improving the models natural form



It was at this point that a rather important realization dawned on me...
As not every vertex of the outer edges of the skull was attached to the requeset edges of the face itself due to the 'Snap_To_Grid' tool refusing to pick up on many of said vertices (a result of my misconducted topological study, I regret to add), the Symmetry modifier that is placed on the 'Face_Object' will not recognise the 'Skull_Object' because not every vertex has been 'Welded'!
And so, I had to manually add the 'Symmetry' modifier onto the skull and adjust the settings which Max would have sorted out for me - the tedious task is displayed here...



The metaphorical penny dropped once more when, during the process of selecting the new 'Mirrored' half of the skull shown here to 'Re-Scale' and repositioning it to actually line up momentarilly with the opposite side of my face, the thought occured to me that if the 'Symmetry' mod. didn't work, why on earth wouldn't the 'Mesh_Smooth' mod. follow suit

Sure enough, I had to apply the 'Mesh_Smooth' modifier once more, meaning that it was certain that, as the face and skull had two differing iterations of the modifier, around 70% of the edges were likely not to remotely tesselate with each other - once again, making the process much more complicated for me as I would need to manually monitor the movements of each of the vertices to avoid the scenario of "...moving one so it attachs accurately, only to find another one has moved out of position..."!



 The evidence - as I predicted in a semi-sarcastic manner, Max has indeed caused a major upset with the next step of me assigned to manually tweak positionings of all the outer-edged vertices that are struggling to join with their opposing vertices on the skull...
To avoid even more confusion, I had to add in another iteration of the 'Editable Poly' modifier under the 'Mesh_Smooth' modifier just so that I could work with the vertices without them smoothed and distorted (I wouldn't plan to make this technique any more stressful than it was currently feeling)

 

 After what seemed like hours spent fiddling around with indecipherable vertices around the neck and cheek areas, I had to follow up towards the hairline where I had the sinking feeling that the positionings and placement of vertices and Polygons there were just as, if not more, jumbled
I knew that the final model would need the 'Mesh_Smooth' modifier on to compare with the 'UVW Mapping' that needed to be placed (more on that stage next week) so I enabled it once more - it presented me with the image shown here and I spend another good while adapting yet more vertices in this section to comply with my intended design decisions

 

The neck was comparitively simple to sort out when brought against other items mentioned this week!
As shown, I highlighted edges at the base of the skull, keeping the selection and symmetrical as I could (given the situation) and I was able to pull material downwards to create a neck out of material distinguished from the rest of the 'Skull_Object'



 When the material was pulled downwards, there was a gap underneath the face itself, allowing me to see through the mesh to the rear-ward Polygons - not exactly a feature to keep!
I used a bit of ingenuity and fashioned up a frontal throat and nape section by multi-selection of the front edges running along the bottom of the 'Extruded_MESH' and used the 'Scale' tool to pull another row of Polys upwards, happily creating a natural-looking cleft running down the centre of the section (a benefit of a manually-applied 'Symmetry' modifier at last!)

 

 This ties in with the image/caption below: once the material was 'Scaled' upwards and positioned carefully, I double-checked that there were not leading edges that would protrude outwards from the lower left/right of the facial model - the vertices that were out-of-line were assigned the 'Target Weld' function and snapped to the area I determined, hiding them from view...

 

To improve the sense of realism to the model and to remove the admission that my neck was displayed in true accuaracy, I left-click-highlighted the two outermost Polys of the side of the neck and the whole of the front section and used the rotate tool to pull it backwards and giving the impression of muscle definition and natural undulation in the skin itself


*Okay, this week is over and I can honestly say that I have never had to backtrack so much when modelling an item like this before!
I know that many discrepancies were due to my lack of experience when constructing something like this before, but I'll leave the full excused-conclusion till my last post as a Part 2 to Week 5...*


Week 4 - I Can See, Taste and Smell!

This post will describe how using common 'Polygon' manipulation, Max allows me to create realistic looking lips, as well as nostrils and eye detailing (even a set of teeth thrown in too!)





In preparation for creating the lips, I returned to using the 'Left' viewport to gauge the distances between the upper and lower lips themselves
The technique I would need to employ to get them modelled would be to use the 'Scale' tool - this involves selecting the 'Move' icon, going into the 'Select Edge' option under 'Editable Poly' sub-level, clicking on the lower edge of the polygon you want the new poly to be created from whilst holding the 'SHIFT-key'
I knew I would need to create my lips so that the lowers were 'taller' in a sense that my upper lips, to coincide with other scaling throughout the model





Once I extruded a polygon from all existing polys around the lip area, I then had to choose to move my viewport over to the 'Left' view as shown here so that I could move the edges along the 'Z' axis (an important step, otherwise I would end up having a parrot's beak instead of fleshy lips!)
You can clearly see that the various 'Scaled' polys will not tesselate well with each other as this stage, but the image/paragraph below describes the next step




As part of this step, once all the edges and resulting vertices were close to one another and no gaps were entirely visible, I needed to use the 'Weld' tool once again - highlighting each vertex perpendicular to the other so that two seperate polygon's would be joined by their parent vertices (again, solving any issues with gaps along the modelled area itself)
The result is that roundness and fullness is brought back to the lips and I can begin to form them as per my reference images




The tool that I used to repair distorted and discordant polygons and vertices in the last post has become a lifesaver when recreating topology that I hadn't realized when Photoshopping the original reference-photograph would be so disastrous to do without
The 'Connect' tool has allowed me, here, to insert the vermillion border of the upper lip which the philtrum and nasolabial folds lead into before reaching the lips themselves and the vermillion of the lower lips that curve upwards from the chin





This the the result of the work that has been carried out - I decided to draw out the lower edges of the lips into a neutral-like expression if only to give me greater flexibility if I feel the need to change the form of the lips at all or conjure a reason to open the mouth back up
Note that the nose had also gained another 'Connected' joining edges which perhaps may allow me to transform the nostrils slightly before adding in the inner nostril detailing



I was now needing to create a pair of eyeballs that would (hopefully) fit inside the sockets that I had spent a good deal of time adjusting and repositioning to achieve as close to real-life dimensions as I could
By creating a 'Sphere' from the 'Primitive Shapes' tab, I then converted it to an 'Editable Poly' so that I could adapt it form easily - this image shows the use of 'Soft Selection' which allows the user to manipulate the standard form of an objects order of vertices, edges and co-ords
I am causing the Sphere to compress in on itself here so that I could achieve the desired form needed for this purpose


 

 Once again (most certainly not for the last time) the 'Connect' tool meant I could overcome another potential time-wasting spectacle of having to recreate any part of the eye-socket should my manipulations of the sphere not quite sit right with the modelled feature
This has meant I could add further precise detailing as to how I wanted to socket to appear - including the amount at which the lids would fold upwards when the sphere is inserted, as well as to make sure the 'ball itself will not appear to be falling out of the socket when rendered!





Upon closer inspection of my eye area, I thought that I could add in the small section at the inner corner of the eye where the tear ducts are situated and where the eyeball itself is actually away from the socket ever-so slightly
I used the 'Scale' key-combination of left-click and the SHIFT-key to drag out two extra polygons from the outer edges of the existing outermost polygons of the corner - I did find I was unsuccesful in using the 'Weld' tool and attach the outermost vertices to the existing socket-polys but, using the 'Perspective' view I was able to align them as closely against the other polygons around the area


 

There was a delicate line to take whether to follow the reference images, taken to maintain accuracy when constructing the model, or in fact, to alter the form of various parts to allow new segments and additions to fit into their places
This was the scenario I faced when altering and adjusting the eyelids elevation and depth when I was planning to position in the eyeball - here, you can see that from this angle, the sphere is coming into contact with the upper lids at the corners...

 

Once I had sorted the left-hand eyeball, I then thought all I needed to do was to use the 'Clone' tool and transfer the cloned 'Copy' along the 'X' axis so that it could slide into t'other eye socket, with only a small tweak to the 'Y' axis needed to compensate for any slight miscalculations with the 'Spline' modelling
Typically, I realized that due to my oversight of needing to use the 'Rotate' function to comply with the flowing facial structure of the left-hand socket, I needed to go into the 'Type-In' floating menu shown here to retain precise measurements to ensure a snug fit!

 

It did strike me as peculiar how, no matter how a user can keep concentration and a level-mind when going through each stage of this process, you can take a step back and see parts where you think "that isn't exactly looking as I imagined it to be" This shows the stage after the eyeballs had been positioned
 I thought it best to keep an equal level of detail shown so far in the modelling when adding in the material for the eye-detailing with the next section below describing that process






To get a detailed Iris/Pupil material that was close to my own without me needing to photograph myself (painful!), I found the image shown here within Google's Image search and decided that I could modify it to seem more realistic - the image was ideal to wrap around the spherical eyeball as it had the square border round it which would accept a 'UVW Mapping' modifier much better than the eye in circular form...
Taking it into Photoshop CS4, all I wanted to do was to add a 'Levels' Adjustment layer , resulting in an enchancement to the 'Red' Channel which contributed to the visible magenta border around the pupil itself





I still felt there was a touch more I could do to further enhance realism and true-to-life detail


I used the 'Brush' tool with the same magenta hue to add in some faintly-visible nerve endings across the surface of the cornea (perhaps a close representation of the original photograph!) and by introducing a 'Gradient-Fill' Adjustment Layer I could throw a dark-to-pale-grey gradient shown here so that when applied to the eyeball object will show darker around the lower eye-lid, again to add further depth to the object




 This is the 'BitMap' taken from the upload to the Material Library, applied through a 'UVW Map' as a 'Box-Mapped' texture and aligned to the 'Z' axis so that I could get the Pupil/Iris to align in such a way as that half will be hidden as shown on the photo and to avoid the appearance of me having a wide-eyed frog-like stare
At this stage I also employed the technique of incorporating a 'Specular-Highlight' so that its looks as natural light would reflect of the glass-like surface

 

 The positioning of the eyes themselves was troublesome when trying a maintain a semblence of symmetry, rotating and altering the depth at which they sit inside the sockets so that eventually they look relatively natural
I had to use the 'Freeze' option when working on the rear of the head so that wasn't accidentally selecting the innapropriate object and causing various vertices, edges or elements from moving around without me noticing until it was too late!

 

 Now it was time to move onto creating the inner nostrils - at this point, it was pleased with  how well formed the outer areas of the nostrils were looking, considering I spent a good deal of time adjusting and altering the vertices around to create natural and flowing topography around the outsides
Going into the Sub-Object level of the Modifier List down to the 'Editable Poly' level, selecting 'Polygon', I used the 'Create' tool to generate another poly across the open-space, joining each vertex up and creating a solid 2-D cap over the end - I made sure they all joined properly by ensuring I had the 'Snap-To-Vertex' icon engaged

 

 

With the newly-created Poly still selected, I used the 'Extrude' tool under the 'Edit Polygons' menu
By using the 'Group_Extrusion Type', and specifying the 'Extrusion Height' of -0.57, this is
how the Polygon looked afterwards - a efficient way of creating nostril inners without having to use the time-engaging 'Scale' tool and worrying about 'Welding' all the vertices up inside the nostril canal


 

 A brief update as to how the the modelling looked at this current stage where all facial features were completed: eyes inserted w/ material applied, nostrils 'Extruded' to a neccessary amount and lips that are as well formed as time would allow ( As part of this stage. I have also enhanced the 'Subdivision' method to a 'Quad Output' with 2 'Iterations' accounted for)


*I have, so far, certainly felt the consequences of ill-preparation of the reference and concurrent modelling process - without supremely accurate initial photographs and resulting topology, problems that surely can hold a user back and cause many incidents of mis-informed decision making!*






Although this decision wouldn't have been based entirely on fact in accordance with the photographs, I wanted to show that it would be possible to display another technique and an example of individual persona by incorporating a material onto a 'Primitive Object' and using it in the overall rendered image - this example is of course a set of teeth!

A cylinder was created, modified with a 'UVW Map' through 'Box Mapping' and a internet-downloaded BitMap of a shining pair of pearly-whites were applied, as displayed here







Now, I was to position them inside the head, behind the now slightly open lips (modified by selecting vertices and adjusting them along the 'Y' axis to give a natural presentation of the upper incisors)
Of course, I needed to avoid the cylinder coming into contact with the edges of the lips near the nasolabial folds - I was able to gain greater incisive adjustment by right-clicking on the 'Move_Transform' icon and get the detailed 'Type-In' window shown here

 

 

 I decided to, once I was relatively happy with the intial positioning of all details, render the image for the first time to check its appearance before advancing any further - I was well aware at this stage in the process that the untidy looking edges near my cheeks and neck needed to be sorted out - perhaps I was to use this as a chance to incorporate a full skull and neck after all (it's be a shame to avoid attempting them altogether I admit...!)



* There will be a part 2 this week as well, which will detail how I doubled-backed, retrofitting the skull and created a fully 3-D object (a process which strongly highlighted the weakness my initial plans were created in)*