One of the BEST, (Beginner to a VERY Advanced), SHORT computer game course

Hi again Igor!
Next question:
I made a wheel and to connect it inside I've made bridges between polygons. But they didn't have joint vertices at the intersection and I cant edit that polygon in the middle (inset, extrude ect.)
I've tried to add vertices on the edges intersections but it work only for one of them.

Please help)

 

Almost forgot...
Same wheel... Trying to make it looks like wheel on the pirate ship)))
I have 18 polygons outside, and need first to inset all of them one by one and than extrude. Tried shift+left click to select all of them and edit but it not really useful.

Maybe it some short cut to do that faster?

 

Sure)

First create a Tube object (18 segments), convert it to poly. Create a Cylinder with the same 18 segments. Align the cylinder with the tube. Attach tube to cylinder or other way arround and bridge the polygons.

When you inset, extrude or bevel, there is like 3 options it provides, one of which is a dropdown list. pick your number of polygons both on tube and cylinder, inset via individual mode and bridge them for a single-object wheel.

 

Thanks for quick response.

As I understand we need to create most of the objects in the 3ds and then send them to the Unity. But how to do it right? Should we make all of them in separate scenes, but those who interact with each other at one? And can we animate them in Unity or make them interact?

 

That's exactly what part #7 will be about.
Basically make use of export and output different objects into Unity. They are nothing but a graffic "capsule", empty inside, just a bunch of vercicies.

After producing such assets, we make use of them in unity, describe what each "capsule" should do, how to behave under physics (if applicable), how to shine under the light etc. We also assign different behaviors in game engine, that is - in unity.

Yes, we do animate them in 3ds max, and all of the world animations are pre-made in 3d software, such as 3ds max. As you export objects, you will be able to output the companion animations belonging to this particular object as well. Animations are nothing but another parameter native to this object we are exporting. It's a description of where each vertex should be in space as the time progresses.

Later, we can trigger such pre-made animations in unity under certain conditions, to create an illusion that they are living their own life. However, a more advanced approach is required sometimes, to for example lean against a wall in real time. Of course you can pre-make an animation for that, but it's more clever to drive the bones in unity in real-time, according to an assigned script.

If any questions - don't hesitate to ask. It's interesting to answer them and keeps the thread in top.

Igor

 

Igor,

What you can recommend instead of ZBrush? I've tried to use Autodesk MudBox, but it doesn't has that sphere tool that you show us at the tutorial, so its take more time to modify basic sculpt.

Can't wait for part#7))) When you planning to post it?

 

I would say those two are the 2 major ones. I still would go with Zbrush, so by all means choose it.

Part 7 is going to be on character animation, right now I've got a vacation after a big number of uni projects.
I might start to do it here, on a laptop or wait until I get back to compose a video.

 

Replying to check out later.

Otherwise, from the sneak peak I had, this should be pretty good!

 

Hey Igor!
How's your vocations?
Hope you are inspiring to continue recording such a good tutorials)

One question. Can you recommend where to start in Unity?

 

Hey

Currently gathering material for animation tutorial, and working on AI.

If you are going for unity, make sure to go only with c# scripting language.
This particular website has a very nice introductory lessons which will provide you a solid ground for tutorials that will be released by me

 

Hey There)

Question about terminology) How do you call game where you have 3D char, 3D runway, 2D background and camera view is like in 2d but with some angle?

 

3d platform genre)

 

How to turn on second autokey line that located under the line with the slider in 3ds?

 

Autokey is turned on with a pecial button that turns red to indicate a recording state. or simply press N

click here (its a video with a timestamp) http://youtu.be/5JoiacPDUjw?t=3h15m35s
The button to the right of this one is actually planting a combination of keys into a current frame. It's used rarely.

 

Yep, that seems like part 5 content)

 

yep, you're right) I probably forgot it, cause it was some time ago. I'll delete my previous post)

Can you write down some curriculum (or even make video) that people can follow during game making process. Like what start with (game concept, samples, character...). It will make life more easier)
I'll appreciate

 

Part 7 is out

 

Hey Igor! Have an issue with exporting object from 3ds to unity. How to improve object quality in unity? In unity it's looks much worse then in 3ds. Please help.

Below some object history in screenshots

 

I found solution this morning) Maybe for somebody it'll be useful

- choose material in assets that looks bad
- in inspector window change parameters in "max size" and "format"

 

Yes VoLkizz, that's the solution to the mentioned problem.
Nice work on texturing!

However, 4k textures are never an option for games. 2k is the absolute maximum, mostly due to GPU limitations. You also wouldn't want to use huge number of texture files. I would say a max of 10 diffuse and 10 other maps for the whole enviroment of the level.

When working on such objects - you would require to write special, custom shader. It will blend the supplied, tiled textures as layers directly on the object. It will do so according to a lower-rez opacity map which also has to be supplied into this shader. You could even paint the object (alter the opacity map) just like you do in a rather limited unity terrain.

I might do shaders at some stage of those tutorial series, but I am not sure atm

 

Thanks Igor!
You read my thought. I was googled question that you give answers so many time. I'm starving about discussion like that.

By the way I left message for you on the linkedIn.

 

You see, I thought that creating bump map and playing with lights will save me a lot of polys, i don't even talk about paint layers (should it be for example one for couple different objects?). About limited textures - i want to have have a big variety of objects in the game with alot of animation and interactions. How to achieve that goal and save productivity and lightness (i mean for pc, ps or xbox)?

 

A little bit of dynamite incoming))

You will have to examine closely your targeted devices and GPUs they are using. Have a look at VRAM, since it's directly related to the numbers of textures that you can use.

Also, keep in mind that vram is going to be used for anything that is drawn on screen, greater means less throttling and swapping of data between RAM and VRAM (slow process). Bigger VRAM means higher resolution, more verticies and accompanied data, shaders, greater number of textures and their detail.

Search for a relation between the power of graphics processing unit and the video random access memory. If you find a particularly interesting link, please post one in this thread.

If you are using layers, you can have a 1k texture for depicting several materials: rock in the first quadrant (square quarter), grass in the second, let's say wood in the third and gravel in the fourth. You also have a texture for the normals to accompany the diffuse texture with four quadrants (normals for rock, grass, wood and gravel in respective quadrants).
If the shader allows to do manipulations on the UV coordinates you will be able to take a step further than normal map - do a Parallax Maping or even Parallax occlusion mapping (just offsetting UVs and textures depending on the view angle vs offsetting the texture so that it self-occludes).

Anyways, you currently have 4 textures in a single 1024 image. Notice how we've stored 4 pictures in a single image, - a single texure. That's how you save on number of textures. Shader will use the quadrants respectively and tile resulted textures across the object containing this custom shader. Since they are above each other, you will have to blend them according to the low-res opacity map, (another texture to go into shader). We cover tiled textures I think in tutorial 4 (or 5), somewhere at the end.

Each object has a set of UVs that are 2D coordinates (2d vectors) accompanying every vertex (3d vector). Using those UVs, you can group objects and unwrap them on one UV table, similar to tutorial #4. You will only use those UVs from objects to sample the Opacity map texture in the UV area specific to that particular object (which is grouped with others). The diffuse textures will be tiled across the object using the UVs, where those uvs get offset into first, second, third (and so on) quadrant consequently.

To create an opacity map (how the tiled textures will be shown from under each other) you will have to either edit it in Photoshop or paint directly in Unity, if your custom shader allows it. You can have many opacity maps, depending on the number of layers you want to blend between. Keep in mind that they are low-res opacity maps, so you can have them lined up in different parts of a big image, even a 2k one (like 4 quadrants or other stacking). You can reduce the number of opacity maps if you use not only black-white, but an RBG opacity map and force shader to pick it up. It's a more complicated approach, but now you can blend between 3 textures with one opacity map instead of 2 textures - resulting in less opacity maps overall. Remember that alpha channel is used for specular color, so it can't be used by opacity map (tutorial #6, beginning).

Keep in mind the material batching Unity provides and try to share textures and materials between objects.

Speaking of linkedIn, currently I just see everything I am doing as a hobby of mine. I wouldn't want to give any lessons other than these ones (videos), plus I am constantly learning and think the current approach is the most efficient one. I have time for editing and can produce a juicy content at the end for everybody to enjoy))

Don't want to participate in majority of projects at the moment, but will definitely tank-drive head on into the gaming industry a little later.

But I am always happy to answer questions here

Igor

 

Wow! Little bit overwhelming)) I think you get the idea which mechanics of this game I want to make. Avatar gonna running back and forth, jumping up and down, interacting with object in the right order to complete the level.
I'll appreciate if you'll help me with such things as:
- Is it right that i'm making landscape in 3ds & mudbox (if yes how to optimize in steps, link or video), or for this game its better to do terrain in unity?
- You said that per level i can have around 10 maps and diffuses, what if i have complicated object (not just one in the level) that contain 22 meshes some of them animated ( should i put all VW's in one file or how does it work?) Just dont want to spend time on useless things.

Thanks for helping

 

For a platform game you want to create a set of assets kind of like the one you had problem a wile ago. You can then combine them in Unity and create a level from there.

I mean 10 textures max per environment of the level, - if you need to texture characters or items it's up to you how many additional images you would want to use. If some parts of the objects will never be visible - cut the vertices and remove uvs. Look for How to make a horror game #4, it's exactly about grouping objects for uvs and tillable textures + how to texture walls of house with just 1 texture - so that they don't look identical.
When you animate an object (such as spinning wall), - make sure to link it to a helper or a dummy (object with no geometry). Feel free to animate the first object, but don't touch the dummy, - it should have no annimation. Don't export dummy.

If you want to have objects share one texture, group them in the following way:

1) attach all of them together into an editable poly.
2) Stick unwrap UVW and unwrap vertices next to each other.
2.5) render the UV table as a .jpeg
3) Take a screenshot of UV table and annotate where each object is.
4) detach objects, check their pivot points.
5) animate objects if you need to, don't forget about linking to dummy etc.

For horror game I would be making landscape mesh in 3ds max in squares. You can then load/unload them at runtime.

 

How to flip faces of the object? I have one object where when im trying to select faces from outside in unwrap UVW modifier it doesn't allow me to do that, so i can't get proper uvw map.
Thanks

 

Flipping normals of specific polygons is only possible to be done from Editable poly, sub object level. Your normals are probably flipped, so
before you stick on Unwrap UVW, first apply a normalize modifier --> unify normals.

 

Hello Igor!
Have question about C# scripting in Unity.
I've wrote the script for character movement that allows me to move in all direction on the plane, can flip graphics body from right to left, but i want character look at the direction where it moves. But this is secondary. I have camera not tight to the character and when char flip to the left controller doing the same - right and left buttons works fine, up and down - invert. I thought, if I'll set invert option on "Vertical" in Input Manager to turn on automatically when character facing left it can solve this problem. Can u please help me to do that, what should i write in charactermovement script to achieve that?

Here is the script
using UnityEngine;
using System.Collections;

public class CharacterMovement : MonoBehaviour
{

public float maxSpeed = 6.0f;
public bool facingRight = true;
public bool facingUp = true;
public float moveDirectionX;
public float moveDirectionY;

void FixedUpdate ()
{
rigidbody.velocity = new Vector3 (moveDirectionX * maxSpeed, moveDirectionY * maxSpeed);
transform.Translate (Vector3.forward * Input.GetAxis ("Vertical"));

if (moveDirectionX > 0.0f && !facingRight)
{
Flip ();
}
else if (moveDirectionX < 0.0f && facingRight)
{
Flip ();
}
}
void Update()
{
moveDirectionX = Input.GetAxis ("Horizontal");
moveDirectionY = Input.GetAxis ("Vertical");
}

void Flip ()
{
facingRight = !facingRight;

transform.Rotate (Vector3.up, 180.0f, Space.World);
}
}

 

1) Grab an empty game object and center it at your main character made of cubes.
2) Throw your script on this empty game object. Get rid of the script from your cubes.
3) Parent the main char made out of cubes to this empty game object with the script. (rest of the cubes have to be linked to the first cube that's about to be linked to the empty game object.
4) Make a variable public GameObject firstCube
5) stick your cube that is linked to the empty game object into the new variable.
6) in the script apply flipping to the firstCube object. This will flip visually the character, but the Empty game object (where the script is actually located) will not have it's axis turned.

You can then get rid of the checks for "is the object facing one or the other direction?".
Just move left right with empty game object and do the visual flip on its children, the cubes

 

Hey Igor! Thanks a lot!!! I was suffering with that issue for couple days))) It works just how I need. Buy the way, as you see I start learning C# thru video tutorials and also bought the book that i'm very happy with for now called "Learning C# by developing Games with Unity 3D" beginners guide. I wonder to ask you, as person who stady that staff alot, is there any tricks and tips to make process of learning and writing scripts less painful?

 

From my experience, learning the scripting language is nothing like maths, where you start with one concept and others come out from the first one. It's more like learning a language.

Trust me, once you learn a chunk of sentences, words and expressions you will see how flexibly you can express your thoughts. You then will start to see why it all works and has a meaning. Remember - it's nothing but a cloud of electrons traveling in your piece of metal called PC, in a harmonious way.

Book is a good way to start

 

"Remember - it's nothing but a cloud of electrons traveling in your piece of metal called PC, in a harmonious way."
I like it. Philosophical stuff...

also want to ask if C# good to start with, for beginners?

 

For sure, I started with unityscript and it was only good in the sense that I could look at how coding works from two slightly different views. If you get used to Csharp, u can then get into C++

 

I've add jump function to my script, but some thing wrong - I can manage jump speed but i cant do anything with gravity (i tried to adjust it in Input M, Physics, Inspector nothing help) its falling to long and look unnatural. What i done in the script is add variable

public float jumpSpeed = 600.0f;

and in da void Update

if (Input.GetButtonDown ("Jump"))
{
rigidbody.AddForce (new Vector3 (0, jumpSpeed));
}

It used to work in my other unity project, but not really with current one.
also have problems with colliders - if pressing the collider with character holding key down at some point it just go thru the collider. how to fix that?
will appreciate any suggestions

 

Hey Igor! I've finished my book! it's get some time to start understand code) Also unity reference helps A LOT.

I fix my CharacterMovement script, so now it's look acceptable, but still need some further adjustments. Just want you to look at it and tell what you think.

using UnityEngine;
using System.Collections;

public class CharacterMovement : MonoBehaviour
{

public float maxSpeed = 5.0f;
public bool facingRight = true;
public float moveDirectionX;
public float moveDirectionY;
public GameObject GraphicsCharacter;
public float jumpSpeed = 300.0f;

void FixedUpdate ()
{
rigidbody.AddForce (new Vector3 (moveDirectionX, moveDirectionY));
transform.Translate (Vector3.forward * Input.GetAxis ("Vertical") * Time.deltaTime * 5);
transform.Translate (Vector3.right * Input.GetAxis ("Horizontal") * Time.deltaTime * 5);

if (moveDirectionX > 0.0f && !facingRight)
{
Flip ();
}
else if (moveDirectionX < 0.0f && facingRight)
{
Flip ();
}
}

void Update()
{
moveDirectionX = Input.GetAxis ("Horizontal");
moveDirectionY = Input.GetAxis ("Vertical");
if (Input.GetButtonDown ("Jump"))
{
rigidbody.AddForce (new Vector2 (0, jumpSpeed));
}
}

void Flip ()
{
facingRight = !facingRight;
GraphicsCharacter.transform.Rotate (Vector3.up, 180.0f, Space.World);
}
}

P.S. thanks that you've been not answering this question)

 

Well done for using Fixed Update function for your physics. One thing you should change is Time.deltaTime - it's not applicable in Fixed Update. In that case, use Time.fixedDeltaTime, since it the sample time of the physics loop.

Find more explanation on this variable here.

if (Input.GetButtonDown ("Jump"))
{
rigidbody.AddForce (new Vector2 (0, jumpSpeed));
}

this piece of code above needs to be in FixedUpdate, just as any other physics / rigidbody things would be

 

Thanks for advice)

I'm kind of stuck with other thing now.
Trying to make my character be able to pick up objects, represent them in inventory and also in the hands. Everything work, but not how it should.

(Below is couple scripts that I'm using for it)

//from script that attached to each item
public GameObject parentBone;

public void hendlingItem()
{
transform.parent= parentBone.transform; // - isn't it here something missing?
}


// from inventory script
private void OnTriggerEnter(Collider hitCollider)
{
if( "pickup" == hitCollider.tag )
{
PickUp item = hitCollider.GetComponent<PickUp>();
inventory.Add( item );
item.hendlingItem ();

}
}
Problem is that when I collide with an object it's just glue to me like I hit it. It doesn't transform position of item to the parentBone object.

Thank in advance

 

I fixed that problem

By the way, when you planning to release unity tutorials?

 

Yep, part 8 is getting out in a week. Recording and heavy editing is in process atm. So far really sweet and intense)

 

Can't wait to see it)

 

Hi Igor!
Did you ever had problems with disabling the script component in unity?
Cuz i'm using "script.enabled = false", but it just clear checkbox in the inspector and script continue to work like before.

It is no error. And somehow function from other script like "timer count down and restart the level" stop working. I tried couple times to restart unity, but it doesn't helped. When I'm switching back to method "trigger.SetActive (true);" every things work well, but it is disabling whole trigger.

I planned to put couple scripts on the same trigger to have ability manage events when character colliding with it and to reduce weight of the scene itself.
I don't wanna every script to have personal collider/trigger.

Please help resolve this issue.

 

Great to see that this has not been abandoned.
I think I may start checking out your videos this week sometime.

Thank You for sharing your time and work with everyone there isn't much of that these days in the world.

 

Nice tutorial
Thanks you

 

Part 8 is out!

click here

 

Thank you so much for this fantastic tutorial series.
Hope you found some time to make a part 9.

 

Good job!

 

Long post alert)))
TLDR: part 9 eta start of september, Advanced AI + Advanced Editor scripting incoming.

I think that there will be part 9, because it's very necessary.
It will lock anyone's grasp on C#. Huh, in the beginning of tutorial 8 the viewer could not know anything about c#, by the end of 9 he will be a pro))

In it, Its main theme will be Advanced, Custom AI that I am working on as my final year project 2014-2015. Tutorial itself will be made of the following things:
1) writing a highly complex tool for unity editor. It will allow for mesh distortion, manipulation, creation and modification right inside of Unity.
2) Using the mesh data to define custom boundaries of the zones (any shape - irregular quads, stretched triangles, etc). Connecting them, subdividing the sides and fixing any implications such actions cause.
3) Building relations between such zones and storing those relations as Assets. Linking up such assets to the zones, behind the curtains.
4) Looking over advanced data structures (Considering if I should show you guys my version of Red Black balanced search tree and if we should make it in part 9 as well.) - Balanced binary (tertiary, etc) search trees, min-max heaps, lists.
Also, going over the Order time concept and how it's used in algorithms.
5) Implementing them to further enhance our AI setup to solve the Planar Point Location (Google it for now).
6) Building the AI of creatures, perhaps as threaded tasks.

Static classes, asset database, mesh class, recursive functions, function delegates, LINQ, lambda expressions, heavy array work, generic lists, threaded tasks, advanced data structures, serialization - the few of many things we will focus on.

If I make it, - it will be the most epic tutorial ever, a rough estimate is - at least 14 hours long, probably more. I will try to pack information tightly, yet will ensure it's easy to comprehend.

I think I will start making it as soon as I finish my Year 3 ( end of May 2015). Because I will be doing MSc starting from september, part 9 should be done by the start of September.

But it will not be a tutorial for everyone. Even though I will ensure it's easy to understand, it is still very complex, and less than 1% of people will manage to finish it. You will have to code along with me and it's a very intense work. Whole research of my study year packed into part 9's 13 hours)))

However, if you will complete it, you will be able to tackle any problem with game development.
Potentially making yourself a very valuable member of your game design team.
Igor

 

That sounds amazing, thank you for doing this for the community. It's much appreciated.

 

Part 9 Demo

 

Igor Aherne interview, pre-final (how to make a horror game part #9)