3D week 5

Right! It’s been a stressful week after Dreamhack, but I think I’m finally beginning to catch up again! So here we have the final version of the crate with the Diffuse map, Specular map and Normal map included.

And here is the crate without normal map or specular map for comparison!

So what are the main differences with and without normal and specular maps? Well, firstly the normal map adds more depth to certain areas, the darker areas. this helps this particular crate in separating the individual planks a lot more. Originally it was more obvious that those were just black lines, but now it looks a bit more like a space in between the planks. It also adds a tad bit of that same depth to the spotted areas in the wood. This is only a optical illusion though, and if the crate were to be in a game and the player looked at it from certain angles, it would be obvious that the surface is flat. That’s why it’s important to consider the ”role” of the item before leaving everything up to the normal map. For an example, if this wasn’t a crate but rather a weapon in an FPS game, leaving too much up to Normal maps would be hurtful because the player will always be closely inspecting that weapon.

What the specular map has added is ”Shine” to the object. This is clearly vissible in the inner parts of the crate on picture one and 2 and is very visible on the entire left side of picture 3. The specular map is important since it is the main thing that will help the object reflect light. If you look at the ”original” picture (picture number 4) that has no specular map, it isn’t reflecting any light what so ever, this would probably make the crate look out of place and strange if it was put into a game.

The diffuse map is simply the texture itself, it’s where the object gets all of it’s color and details. Without the diffuse map, the crate would be far harder to recognize as a crate.
Without the diffuse, leaving only the normal map and specular map the crate would look something like this:

 (on this picture the work that the normal map does is more obvious as well)

And without any of the maps the crate would look something like this:

 

So to summarize: The diffuse map adds in color and details. The normal map adds in ”bumps” and gives certain areas more depth and enhances the details on the object. The specular map adds in the shine of the object and can help define the edges and help identify materials in objects (a screw for an example is shinier than wood).

 

And here’s how the three different maps look:

Diffuse map

Normal map

This picture clearly shows the details added to the black lines when rendered with the normal map included. The darker areas from the specular map become ”bumps”.

Specular map

The lighter areas in the specular map reflect more light than the darker areas.

 

Well, that’s it for this weeks blog post!

3D week 4

For week 4 we have textured the crate that we optimized! This blog post is coming out a little bit later than I had originally hoped since I went to Dreamhack, and in doing so I lost a couple of days worth of 3D practice.

When I was looking at my crate, I felt like I wanted to really distinguish the separate planks really well. The way I decided to go about this was by marking the edges of the planks with thick black lines, this ended up giving the crate a cartoon-ish look so I decided to roll with it and try to let the textures reflect that. The things that contributed to that look was both the colors I chose to work with (two shades of brown with black outlines) and also the way that I drew the markings on the wood. The dominant color is a light brown color, I decided to go with this color after looking at several different wood textures, most of these were either too dark, too grey or looked too ”waxed” (in lack of better words) but eventually I found one with a color I liked and decided to go with it. Also, since the crate is made up entirely out of wood, I wasn’t really left with an infinite amount of possibilities color wise so it all just came down to testing until I found one that matched what I had in mind.

The colors are both highly saturated, warm colors (the darker one is slightly more saturated than the lighter one) and mainly differ in the darkness of the color.

When I started trying to figure out how to make the textures, I looked through something like fourteen different pictures of planks and wood textures, I kept them on the side to have constant references to see how make the darker areas of the planks. I did this because I realized that creating those areas was really difficult, especially since I had to try and keep them somewhat random. Creating anything randomly when drawing is REALLY hard, if my mind slips for just a second it ends up looking too uniform and I didn’t want this (I know that regular planks can have constant markings, but this was not my wish for these planks). An example is that some of the dark spots in the wood ended up looking like hearts, but I didn’t mind that too much since I have seen that in real planks as well.

Since I had around 37 or so, planks to texture, I realized that it would almost be too time consuming to make each plank individually so what is shown in these pictures are something like 10 different planks that have been flipped, turned around and had slight changes to them. I was hoping that this would be something that is practically unnoticeable, it’s hard for me to judge how visible it is since I made them and when I look at them it’s almost instantly visible. But so far I haven’t had anyone of the people I’ve shown it to point it out to me so hopefully it works!

And here’s what the diffuse texture looks like, the UV with wireframe next to it and how the texture looks rendered. I was supposed to show the rendered version in UDK or Unity, but unfortunately UDK didn’t load my textures and I was unable to download unity, so instead, the rendered version is from 3Dsmax. The post for week 5 will probably have the same issues.

The post for week 5 will be posted in a little while as well!

 

3D Week 3

This weeks post will be coming in 2 parts, first a part where I write about the optimization of the crates and later a part about the upcoming task where I will show pictures of an object that I’m going to model in the near future.

Part 1: Optimizing the crates

Post apocalyptic crate

 

This is the crate that was optimized by Folland Rickard. To the left we have the original model and to the right we have the optimized one.

The area that has been highlighted (light green) is the place where most changes took place. For this one, that area was the lid. The biggest issue with the lid was that I had created it as a plane and than extruded specific areas of the plane to create planks. This resulted in overlapping faces and flickering in one area. Looking at the right part of the lid on the left picture, you can clearly see that it only consists of a plane, basically there is nothing there but 2 faces. The lid was given more of a thickness in order to fix this. The overlapping faces were also fixed by deleting them and creating new faces.

Urban Crate

This crate was fixed by Ludwig Lindsttål, same as above we have the original to the left, a zoomed in version of the original in the middle and the optimized one to the right.

For this crate, two areas were the main subject to change (red and light green). Starting by looking at the bottom of the original version (zoomed), you can see that there is a face at the foot of the crate where the planks are pushed in. To fix this, the entire bottom face was removed and replaced with a new one that fitted better.
The upper part of the crate (Light green) Was earlier covering a more or less identical version of the lower part. The entire upper bit was deleted (since it would not be visible anyways) and the cover was place back on top.

For both of these crates, the poly-count was severely reduced and had n-gons removed.

 

Medieval Crate

This is the crate that I ”optimized”. Same as above, (Red and lightgreen) highlights have been added to show the areas that where subject to the most changes, original on the left and fixed on the right.

The reason for me writing ”optimized” is because that is partially a lie. This crate had some 170 overlapping faces, plenty of overlapping vertices and too many n-gons to count on one hand.
So after careful consideration and inspection of the original crate, I decided to remake it from scratch and then optimize the new one.

So after recreating the crate I was left with about 1/4 of the problems I originally had. An example would be that the 170 overlapping faces were now down to just 40. I had no n-gons to fix, nor any overlapping vertices.

So, regarding the largest changes I made when reconstructing, lets start at the bottom! In the original model, the bottom had the same construction as the all the other sides. I felt that adding that extra side only increased the poly-count and the number of potential issues. Another argument for removing the bottom construction was that it’s an area that may not be visible most of the time.

The second zone that had large changes were the remaining sides (so essentially the entire crate). In the original version, the horizontal planks were far more uneven, this resulted in many more issues, especially since I extruded them more than once per plank. I also reduced the total number of horizontal planks, reducing the amount of issues I had with the crate as well as the poly-count. Instead of making them all diferent sizes, I made them with 2 different sizes. To make them look more ”random” I made sure that the planks on the same height didn’t have the same size (or amount of extrusion, this is visible in the image of the fixed crate).
The diagonal planks were also made to be all the same size, this made the creation of the box far easier on me.

Fixing my own crate turned out to be far easier than optimizing someone else’s work since I knew exactly what I had done with my own model. While saying that, having my own work looked at and fixed was a huge learning experience. Since this is among the first models I have ever created, being in a group with experienced people was really helpful.

 

Part 2: The future model

 

 

 

 

 

 

 

So the object I chose to try and model was this dagger. I chose this object because it was one of the best preserved items I found, I thought it looked cool and working with models like this seemed like a lot of fun! The dagger has supposedly belonged to a captain on a ship that sunk somewhere around the (if I remember this correctly) 1700’s. The reason why they think that it belonged to the captain was because of the patterns on the handle and the fact that no item of similar was found along with it.
I figured I was going to use the ”Skyrim” style when modelling it (this is also one of the reasons why I chose it) because it reminded me of a dagger from Morrowind (an earlier game in the elder scrolls franchise). According to me, this dagger would fit right into that universe.
I think that I will put most emphasis on the handle, especially the upper part of the handle, simply because that area is what makes the dagger itself special. Also, the pattern of the handle reminds me of the pattern on the upper part of the handle on this dagger from skyrim:

One of the things that I will try to avoid later on is making the blade too shiny, I might make it look a bit duller just to emphasize the handle area where most of the details are.
What I think will be the hardest part when modelling this dagger is creating the rounded ”valve” like area in the upper parts of the handle. All in all I feel like the entire process of modelling this dagger is going to be challenging since this will still be among my first models.
My biggest concern is that I won’t be able to do the dagger justice.
The edge might also become an issue, but hopefully I can avoid any major problems by not concerning myself with making it look ”sharp” (Naturally I will try to make it look sharp, but it won’t be my main concern).
One more thing that might become an issue (hopefully it’s just an optical illusion due to the angle that won’t become an issue) is that when I was drawing the edge flow, the blade didn’t always seem to align with the handle. It ended up looking crooked. This should however be easy to avoid since all I have to do is model the 2 parts separately and just attach them to each other afterwards (which was the plan all along).

 

Well, that’s the end for this weeks 3D post!