Thursday 22 October 2009

Group Meeting 1 & 2

The first 2 meetings didn't go particularly well. We didn't manage to get the entire group in one place. This was due to travel problems for a couple of the group members.



In the first meeting me and Kiel decided to take initiative and decide that the group should work on recreating a 2 stroke engine.

We have discussed what styles of shots could be used and what kind of information could be displayed.


In the 2nd meeting we got to know Shuo and discuss the previous weeks ideas with her.

Evaluation

Even though this is my first time at creating a human head. I am slightly disapointed with what I have created. There are a number of errors in the modelling as well as the texturing.

I struggled at understanding how the nose should be created and in the end it was a mess, it was far too angular and was too big in scale compared to the reference plates.



From this picture it is also possible to see that the head isn't completely in line with the reference plates either.

I believe that given more time I could produce a more satisfying piece of work. But as time was always going to be a factor with this project it makes me think that in the future I should possibly use my time more wisely to allow for more alterations on each stage.

Another problem is how I connected the ears, I stated earlier in my blog that I had problems with the welding tool and only certain vertices would attach. In the future I will try to keep such a difficult object simple. Even though I managed to get the initial detail of the ear quite high, it lead to problems. Due to these problems it lead to lost time and an end product that was of poor quality. The back of the ear is overly stretched and lets down the final piece.



In the future I will also look into becoming more familiar with the UVW mapping. With a mixture of the poor unwrapping as well as poor formations for the quads this lead to textures being stretched.



This made me realise jsut how important each stage is, and the initial "simple" stage of the topology is probably the most important if you want an incident free and easy route to completing such a piece of work.

Completed head

I was just about to write up my evaluation when I was thinking, "I'm not happy with what I've produced, somethings not right". So I decided to edit the texture slightly after I realised that the texture around the chin was too far up the face. I have since edited the eyes that I put into place so I don't look like an alien.







The eye was produced using a piece of Photoshop work that I had produced for a previous piece of uni work but couldn't use.

Bump Map & Specular Map

It will appear that I have not added a bump map to my character. This is not through lack of trying. I produced a bump map within Photoshop to add to the materials but when I went to add it as a map nothing happend. I am not sure whether its incorrect settings on my 3DS or whether I have inputted the map into the wrong area. But this was the map that I setup within Photoshop.



The purpose of this is for 3DS to use the darker areas as indents (paws of the skin) on the face and the lighter areas are used as petrusions (spots and moles).




The specular Map was meant to be setup up to bring out the greasy areas of the skin. Again I couldn't seem to get this working, but tbh I think the initial texture produced in Photoshop has a little bit of difference on.

I am disapointed that I could not get either of these effects completed as it would mean that I managed to complete all of the tutorials.

Texturing

When it came to texturing within Photoshop, I stumbled across alot of problems.

After getting the initial texture down, I found that the UVW map wasn't quite right and left me with unpainted areas around the mouth.



At first I thought I could go back into 3DS and simply move the mouth around...and this lead to disaster.



As you can see it turnt my character into a bit of a scarecrow where the texture had tiled itself. I am unsure how to get round this, but luckily id made a previous save so could go back.

It is probably good that I did have to go back as I managed to solve the initial problem. I found that the reason for part of the mouth not being textured was that I had missed a small spot within Photoshop.

I then had a similar problem occur around the front of the neck, again where I had missed a spot on the material.



If you notice ont he head is the dark area under the chin, this is not shading, it is infact poor alignment of my material within photoshop. Here I found how important it is to be accurate with the painting of the face, something that may seem a large area of the face, may only be a small area upon the UVW Map.



My next failure with the texturing and modelling was the ear. During the tutorial it states how important it is to make the ear as flat as possible when using the Pelt tool. I started to realise why after texturing the ear.



The texture of the ear appears quite square, I am unsure whether this is due to the poor modelling of the ear, or whether the texture is struggling to wrap itself around the detailed object. Sadly due to running out of time I cannot look into this further but at somepoint it would be beneficial to me to understand why this has happend.

So at present this is the state of my face...



Welding the Symmetry line & exporting the UVW Map

With a symmetry modifier already on the face, the initial UVW map is set below the symmetry modifier. The next process is to add an Unwrap UVW modifier above the Symmetry and weld the two sides of the face.

The reason for adding the second Unwrap UVW is to allow me to select the right side of the face.
after selecting the right ear and right side of the face, I had to flip the sections so that it produced a map of the full face. By wedlding the central points together I was left with the following UVW Map.



This then gave me a basis in which to apply a texture within photoshop.

UVW Unwrap of the ear

This task required me to follow a similar basis to what I did with the UVW Unwrap on the head.

I found that I had very little to edit in terms of the inverted faces or the overlapping faces, most of the overlapping faces were caused by the outer edge of the ear.

The only difference of the Unwrap on the ear was that I had to use the "Pelt" tool. The purpose of the "Pelt" tool is too take an over complicated surface and flatten it out as much as possible which makes it easier when applying the texture.
(Before and after shots of the ear below.)

From looking at the two screenshots the main difference is how complicated the ear is in the initial one.



It displays far more quads, even though this may improve quality, it will cause alot more problems with rendering times due to the amount of pixels it will take up. The second picture shows the simple form that the ear has after using the "Pelt" tool with Unwrap UVW.


UVW Map

Now then, I'd already had one shot at using the Unwrap UVW modifier before but realised i'd forgotten to produce lips and attach the ears meant I had to stop. To be honest im glad, if you take at view at one of my earlier entries the texturing was a shocker, I had patches of the face where they shouldn't have been and it was too easily distinguishable between the areas.



So the first stage of producing the texture is to layout the coordinates for each area of the face within 3DS.

By adding a chequered material you are able to see what areas of the head are stretched. This is not a good thing as it means the quality of the texture on that area of the head will lack quality.

I moved the vertices within the Unwrap modifier as much as possible until the areas were less stretched (see below for before and after shots)



I decided to go with attempt two as the stretched areas were more around the back of the head and I hoped that with the hair being there it wouldn't cause much problem in loss of detail.



The squares around the face were relatively even so I decided to move onto the next stage to remove any inverted or overlapped faces.



The inverted faces were fairly easy to fix. The majority of the time it was where multiple quads were flowing into each other so it meant just defining each quad more clearly by moving the vertices.
Other times it could be where a quad wasn't really clear and appeared more like a triangle or a really odd shape.

The only inverted face I did not managed to correct was the nose. I believe this is due to the poor standard of modelling when producing the nose. I am unsure how much this will effect the texture map result.



After deciding to leave the inverted nose as it was I moved on to edit the overlapped faces of the UVW Map. I found this a simple task to complete, with only two overlapped faces to deal with.

Wednesday 21 October 2009

Lips and attaching the ear

Since getting used to the UVW mapping, I have gone back over the modelling and produced lips as well as attached the ears to the head.

To produce the lips I decided to extrude edges from around the space where the mouth should be. Using the same technique used for the head, I based the vertices for depth and width using the reference plates.



A problem that I had when setting up the lips was knowing how many quads to have. When smoothing the lips I found that the top lip appeared considerably thinner than the bottom lip. I found this was due to not extruding the edges far enough.



Attaching the ear:

The next stage was to attach the ear, from the initial modelling of the ear I was extremely pleased with the amount of detail that I managed to get.

But I soon stumbled across problems when trying to attach it to the head.

The initial problem was finding what vertices on the ear should be attached to the head. The main reason for this was I was getting confused what were vertices of the ear, and what were the head as it just looked like a giant mess.



The next problem I came to was the vertices that seemed the most obvious to weld to the head, were not welding. When using target weld you usually select what one you'd like to move, and what vertex to move it to. Some of the vertices that I was selecting welded fine, but every so often I'd find a vertex that just wouldn't weld.




This lead to me having to adjust the modelling of the ear. To do this I had to delete vertices and then redraw quads. During this process I managed to make a mistake which lead to a loss in modelling quality of the ear.



I believe this happend due to deleting incorrect vertices. If time wasn't an issue then I would go back over the ear and reattach it with the level of detail that I had before. I believe that this version (above) is still acceptable for the purpose of the project. With this now complete I will now move on to producing the texture of the character using UVW mapping and Photoshop.

Wednesday 14 October 2009

Giving the character some face!

Before I go on to produce some lips for my character, I have decided to give the UVW mapping a quick try.

At the start of the project i was thinking..."How hard can texturing the face be, its just a bit of Photoshop". Truthfully, I thought that all I'd have to do is cut out the face from the original topology and throw it in the materials editor within 3DS...How I was wrong!

I first had to produce a flat version of the head on which to paint onto. This was done using the "Unwrap UVW" editer.

Here it is possible to see what parts of the face aren't connected properly. I found that the vertices around the centre of the forehead were slightly out of place and had left a triangle.

Once I had managed to edit this out I had to export the map into Photoshop.

This is where the fun began. I would have to cut out parts of the topology and resize them to fit the UVW map. This process took up alot more time than I was expecting.

But for a first attempt I thought I did a good job...



All the vital parts of the face are in the right area. I haven't gone into much detail for this area as I will have to redo it for when I finish adding the lips and ears. But now I know what to expect when it comes to the real thing.

Ear

I have now finished connecting up the remainder of the characters neck (after I had somehow managed not to notice!!!)

Its now time to create the ear and finally the lips.

In all honesty I got lost during the tutorial of producing the ear. The problem I found is that I started concentrating too much on what the ear in the tutorial is meant to look like, instead of using the technique and producing my own ear.

After watching the tutorial around 3 times, I decided to turn it off thinking I was familiarised with the entire process.

With some helpful tips from my group member Robert, I found that the ear wasn't so daunting after all and managed to produce this...



I found that the best way to produce the ear was to start off with the basic outline. (Using the line tool in corner mode)

I then used the extrude modifier to produce a block. This would allow me to use the cut tool (within edges mode of editable poly) to produce vertices that I could adjust.

By using the edges mode it allowed me to layout the shape of the ear and add vertices when the depth of the ear changed.


Now that I have created the ear I have decided to use target weld to attach the ear to the head. But now I have a problem with this process.

I haven't set out the area on the head where the ears would need to be connected. (Vertices are out of place and would lead to stretching if I connected the two parts in their current state).

I plan to create the lips and work on the structure of attaching the ear and head components after. Allowing me to be fully focussed on the task ahead.

Group Meeting 2

After having a 2nd group meeting, it has been agreed that an engine would be the most practical exhibit to create.

This is because we believe it to the the most interesting compared to the pumps, telephones, radios and other exhibits.

We believe that with the engine, it won't limit us on how creative we can be.

Not only can we present an engine in three dimensional form, but we believe it can also be made interesting to explain how the engine works.

With the use of sound, speech and animation we believe it possible to create an engine.

The group discussed the benefits of producing a 4 stroke engine as well as a 2 stroke engine.
In the end it was decided to go for the 2 stroke. The problem with the 4 stroke is there are far too many parts within the engine that are unseen, so to produce an unfamiliar product would seem senceless.

Thanks to the museum offering a demonstration 2 stroke engine, it allowed group members to take apart and rebuild a basic 2 stroke engine from scratch. This meant that certain group members were familiar with was made such an engine.



The group then discussed what elements would need to be created in order to reproduce the 2 stroke engine, they were:

- Engine casing
- Spark Plug
- Fan
- Pistons
- Animation of Pistons/Fan
- Simulation of firing order

Once created the group would then have to look into produce:

- Textures
- Lighting
- Camera work
- Animation

I believe that the group has a very strong idea for what they want to achieve from this project.

We aim to have a meeting on Friday 16th October where we aim to get dimensions so that each group member can produce their part of the engine so it fits with the others. Once the ratios have been setup it will be time to delegate the tasks.

Tuesday 13 October 2009

Back & Sides

Since my previous blog entry I have been concentrating on producing the back and sides to the head.
This consisted of copying existing quads to form the basic shape. Then by using the target weld tool, connecting the vertices together.

Here I learnt how important it is to have the features lined up. Where I was working using only one view I didn't notice that i was moving the points so far out of place that the head appeared lop sided. Even though I thought it was easier working from one view, I decided to use multiple views so I could keep track of where every vertex was being moved to.

I found the tutorial for this section to be poor. The explanation for completing this process seemed rushed, and lead to panic when I wasn't sure how to produce the side of the head.

I realised it best to think back over techniques that I had taught myself during the tutorials so far and was able to produce a full head.

After completing this process I realised that the forehead of my character was petruding too far at the front so decided to adjust this while it was fresh in my memory.



Another problem that I had spotted was the jaw line of the character. It appeared stronger than what the picture appears.



I have decided to edit this area one I have produced the lips. This is incase I am forced to edit the region anyway as it saves time by editing it twice.

Thursday 8 October 2009

Group Meeting 1

The first group meeting was setup for Monday 5th, and even though only two members managed to make it to the meeting some good ideas were discussed. We looked at what exhibits were on offer at the museum.

We ruled out producing an animation that represented how electricity is produced. This was due to the museum already having a physical representation.

The first idea we had was to produce an animated version of an engine that the museum owns. In the meeting notes (below) you can see what me and Kiel discussed as to what aspects of the animation would be important.

Another suggestion to what the group could produce was the pendulum clock. With its simple design, made of basic shaped parts. The only problem was making this into an interesting animation.



The next offical group meeting has been set for Monday 12th October. Here we will discuss in detail what the group believe should be animated within the engine, a shot sequence (storyboard) and possibly a delegation of tasks.


After the initial discussion I am pleased with what Kiel has managed to collect so far. He has a number of detailed pictures of the couple of engines, and has managed to find a design of inside an engine.

If the group can get together often enough then I believe we can succesfully answer the brief.

Mirroring & Nose

Even though it is possible to build the entire face without symmetry, a mirror style technique was used to produce an entire face. This allows both sides of the face to be adjusted with only selecting one point.

When adding the symmetry modifier, I found that the face looked severely pinched around the centre.



I found that the reason for this pinching was I had Vertices too close to the line of symmetry, so moving them across slightly allowed for a more natural curve to the face.


The next stage was to produce the nose. I had a lot of trouble with this process and at the moment the nose still isn't perfect. The main reason for this is due to the way the quads were formed around the nose and mouth.



This is the current state of my characters face, as you can see from the 2 angles the nose doesn't appear natural. The nostrils are currently too small and are set too far forward .

So that I do not waste too much time, I have decided to move onto the next stage and work on the nose at a later in the project if there is time. This way I can continue to gain knowledge with using the techniques that I have already learnt, as well as gain knowledge with using new techniques.





Updated...



I have since managed to edit the area of the nose and have pin pointed the reason for my problems.


I had managed delete a vertex within the nose so not all quads were actually connected to the face.

After correcting this problem I have managed to move the nostrils so they sit deeper. A new problem since moving the nostrils is that the nose appears extremely angular. With the use of the turbosmooth modifier I have found a way of quickly correcting this problem.

Sunday 4 October 2009

Welding and modelling

Today I went to weld all the vertices together, this is done to make sure that when you move one part of the face, the surrounding areas are adjusted.

Next I had to start forming the shape of the face, this gives the three dimensional element to the head. Well this process started well, but I didn't really understand where I was supposed to be moving the vertices to. Here I decided to stop viewing the remainder of the tutorial and mess around with the vertices until I understood what I was actually aiming for.



This was the right thing to do for me,it meant that I understood why the lines curved in certain directions and also allowed me to learn more of the settings within 3ds through mistakes i was making.

One technique that i have managed to learn that will become useful for the future, is producing extra vertices and also deleting unused lines. The reason that this will become useful for this particular project is that i can now recreate areas where i have made mistakes.
For example, areas of my topology did not weld togeter within 3DS. This means that holes were left within the face when setting the topology as a editable poly. (Below I explain how i got round this problem and was able to produce a cleaner looking face design).



With this technique it allowed me to delete lines easily if i had accidentally produced a triangle on the face. All I had to do was select a vertice (vertex mode) or a line (edge mode) and press backspace on the keyboard. This would open up a larger quadand reduce pinching on the face.

A problem I found with this though, is that by deleting an edge it would produce a straight line across the face, this was frustrating if it was the cheek area which required a smooth rounded surface.

But there is another technique that can easily rectify this. By selecting the "Vertex" mode, it gives us the option to added points on the face which will allow you to produce more quads.



With this step you have to select the Edges option and check the "Insert Vertex" option. This allows you to select anywhere on the face where you would need a vertex.

Once the vertex has been setup, go into "Vertex" mode and highlight the vertices that you would like to connect together and press the "Connect" button.

I found this an invaluable technique to use when finding problems with producing the basic shape of the face, and no doubt I will have to use it again at some stage further on with this project.

Saturday 3 October 2009

Character face continued

Over the next week I am to complete the next three tutorials based on the building of a head.

I have just completed adding my topology to 3ds using the line tool. I won't express my thoughts fully, but my word was it just a little bit tricky!

I finished my first attempt, went to connect the entire set of lines and found giant holes in it. So I thought, ok this isn't so bad i'll just go back an edit the lines. More and more triangles started to appear and I was panicing.

The good thing about finding the triangles at such an early stage was that it didn't take long to put things right. I did this by using the line tool again and reattching vertices to produce more squares.

the next stage is to start producing the shape of the face in three dimensional form. I found that when I went to view the vertices, that far to many were appearing, the tutorial shows around 240 vertices for the face. Mine was appearing with a figure closer to 5000.



I'm not sure what setting I had selected for this but I had to produce the head references again.

Museum first impression

So far i believe my tutorial work has been progressing quite well. Even though I have hit a couple of problems, I have managed to get round them thanks to feedback or exploring what options are given with the software.
Now that the first three tutorials have been completed I decided to take a trip to the Museum of Power. I wanted to get a basic idea as to what the museum is about and what exhibitions are on offer.

My initial thought was how small the museum is. It didn't seem the most welcoming of attractions (with it being impossible to track someone down that worked for the museum).

When finally making my way round the exhibits, I was shocked at just how technical they were. The museum was home to numerous styles of pumps, information on how electricity is made and what it is used for.


It was intriguing to learn how the pumps work and what each was used for, but the exhibit that impressed me the most was the car engines on display.



After collecting an varied amount of photos from the museum I have had the chance to think of ideas as to what the group could do to promote an exhibit. A group meeting has been setup for Monday 5th, to compare ideas and detail what the group would like to achieve from the project.