technical architecture.

Scripting

My mashup of Unity experiments (Hydra/Oculus Rift/Spatial Graffiti)

by on Apr.17, 2013, under Interactive, Oculus, Unity, Unity3D

AKA: Things I play with instead of sleeping, after my son goes to bed.

EDIT: hey there people coming from Bruces video! check out the newer version for a nicer demo: cheers! 😀

SpatialGraffiti

Architecture – an interactive basketball arena 3d scene to wander around. Loosely based off of a project that we didn’t get paid for, so, mileage ++!
Hydra – for movement, and glory!
Line work – Spatial Graffiti, draw lines in 3d space with the hydra, It’s more fun than you’d think.
Basketball – kinda. It wasn’t too exciting, so it’s more or less removed by now.

How it works:

  • Rift, head tracking turns the player’s viewpoint, it just works? Thanks oculus!
  • Hydra, unity integration is nice, thanks!
    • Base unit:
      • Leave yourself enough room between the unit and the monitors! NO MONITOR PUNCHING ON MY WATCH!
      • I’ve found it works best if you have the controllers ‘docked’ at game start. It just saves some configuration hassle.
      • Pick up the controllers from the base station, move them to a comfortable position closer to your body, about as far away from each other as they appear in the rift view, and click the ‘Start’ buttons (The vertical button between the 1 and the 2 buttons) to take control of the hands.

[TL,DR]

  • Right Hand:
    • Thumb stick turns the player body.
    • Button 1: Points the finger, begins drawing a line in 3d space.
    • While button 1 is down, how much the trigger is depressed controls the width of the line, up to a point.
    • button 4 will launch a basketball, holding the button adds power per frame, letting go fires it off.. I know I know, throwing motions.. soon.
  • Left Hand:
    • Thumb stick moves the players body
    • Button 2: Points the finger, Inserts a ‘move’ helper near the hand
    • While button 2 is down, hold the trigger down to ‘MOVE’ the player body in 3d space. The movement isn’t 1:1, so go gentle to start.. but everyone I’ve tried it on has been moving this thing almost as fast as you can move your arm within 30 seconds, with minimal icky stomach feels.
    • The icky stomach feels is one of the main things I would like some feedback on, so I can pursue or steer differently.

[/TL,DR]

The main OVR menu scripts are in there, so if you manage to say, jump off of the world ‘by accident’ and look up at the things receding, you can hit the Right Shift key, and then Return to reset the scene.

Things I’m working on and experimenting with still

  • UI stereoficiation, NGUI draws well, but need something better than shooting a ray out for interaction. So far I don’t like the feel of the sixsense UI bits. working on some things here.
  • Networking. Multiple people in a small space drawing stuff, So far it works on lan, but who really has 2 rifts/hydras they can put in a room? I’ll figure it out!
  • Fly mode for the hydra, throttle style movement, basics started, but needs some tweaking before it’ll not make people feel strange.
  • 1:1 rotations, started, but I wanted some feedback on other stuff first!
  • Holding both bumpers will allow you to scale the player height/viewpoint up based on how close/far the ‘hands’ are from each other. nearly works, just a little too fall-thru-the-floory for my tastes to leave it enabled in this version.

I do architectural interactive/animation/work for a paycheck (haha, I know right!), so I spend a lot of time working with the fundamental problem of how to get non-game people empowered to get around in 3d space. I love toying with 3d spatial navigation systems, so being able to immerse in space with the Rift and prototype with the responsiveness of the hydra has been a lot of fun for me this week. There’s a lot more to play with here!

Most of my position/rotation stuff is made for placing items in the world, but the hydra tuscany demo covers a decent portion of the basics of that idea already. Besides, it was fun to mess with the player positions/orients, so that’s what I’m playing with so far.

So, if by chance you made it this far into my ramblings, here is the link to download the test app and give it a go!: 2013-04-18-BuckHydraLines.zip

4 Comments more...

Unity per project compiler options, -unsafe

by on Nov.28, 2011, under csharp, Unity3D

I had to integrate a small set of third party code that calls down to some win32 dllimport bits, to do this I needed to be able to work with some pointers. which in c# need to be declared in an unsafe function, so that it knows to be unmanaged.

Slightly awkward, but it turns out that there is an -unsafe compiler option that needs to be turned on in unity/mono, and that you can do this very easily on a per project basis by adding a file ‘smcs.rsp’ to the root of your assets dir, and in there you can add the text ‘-unsafe’ to add this to the compiler commandline.

Reference threads: Setting Compiler Options, Unsafe code in Unity, How To: Set Project-Wide #pragma Directives with JavaScript

Other apparent uses stated in the threads are setting project wide defines and pragmas

3 Comments more...

Fun with Generics and Lists in C#

by on May.19, 2011, under .net, csharp, Scripting, Unity

So I’ve been doing a lot of List juggling lately and have run into a few tidbits that I wish I had learned a few months ago. I figure that is always a good cue to write them down for future reference and for any others who are digging into this C# world headfirst also. this is all sort of an exansion to try and get a concrete example of mike_mac’s lamda/delegates/callback articles

Say I have a custom class that I’m using to store a structured set of data, this time we’ll play with “ArmDefs”. ArmDef’s just hold a little string and int based data and a list of ArmDef’s as children, and a second list of GameObjects that contains all of the models for this section of arm for use to serialize to disk, but thats more a story for another time.

Overall this is used to build up a hierarchy that holds the “contents” of a runtime modular rig, and handles the Saving/Loading/Reloading.

To work with this I needed to set up a pile of recursive functions to edit the tree at runtime, since I don’t know the depth of the hierarchy, and rebuild the system as needed.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
[System.Serializable]
//Inheriting from ISerializable gives us
//an easy access to Serialization to disk
public class ArmDef : ISerializable {
    public string Name;
    public List<ArmDef> Children = new List<ArmDef>();
 
    public ArmDef(string iFile)
    {
        this.Name= iFile;
        this.Children = new List<ArmDef>();
    }
 
    public void AddChild(ArmDef child)
    {
        Children.Add(child);
//just wrapping the Add so that we can do "extra"
//stuff here if we want, like add links to parents etc
    }
}

not really building these thru scripts out in the wild but just to build something to work with here.

1
2
3
4
5
6
7
8
9
10
public ArmDef config = new ArmDef("Base Level");
 
//add a couple children
config.AddChild(new ArmDef("Level2a"));
config.AddChild(new ArmDef("Level2b"));
config.AddChild(new ArmDef("Level2c"));
 
//add a couple grandchildren
config.Children[1].AddChild(new ArmDef("Level3a"));
config.Children[1].AddChild(new ArmDef("Level3b"));

ok thats all the boring stuff for now! here is a Generic Recursive function that walks down a tree and does ‘Action<>‘ to all children of ‘parent’. The key here is that the Action<> pretty much rocks. !

1
2
3
4
5
6
7
8
9
10
11
public void HandleArmDef(Action<ArmDef> item, ArmDef parent)
{
    if (parent.Children.Count > 0)
    {
        foreach (ArmDeft in parent.Children)
        {
            HandleArmDef(item, ti);
        }
    }
    item(parent);
}

Now the fun begins when you start to chain this stuff together to “act” on all children.
we can define that “Action item” in the above function to be something else using some Lambda or Delegate type stuff. (I’m learning here, those are the reading points for the moment!)

This will go through the hierarchy and remove all children with the Name of “Level3a”.

1
2
string toDelete = "Level3a";
HandleArmDef(item => { item.Children = item.Children.Where(child => child.Name != toDelete).ToList(); }, config);

In another example, i wanted to search the hierarchy and see if there was a certain value anywhere in the hierarchy. easiest way i found was to just quickly build a flattened list and use the Linq “Any” function to see if the string was found anywhere in the hierarchy.

1
2
3
List<string> flatList = new List<string>();
HandleArmDef(item => { flatList.Add(item.Name ); }, parent);
bool isCurrentlyFlat = flatList.Any(x => x.Contains("Flat"));

in another case, i wanted to insert a new child into the hierarchy as a child of a specific object.

1
2
3
4
5
6
7
8
9
10
11
12
13
void AddChildAddon(ArmDef parent, string addAsChildOf, string newChild)
{
    //for each object 'item' down the hierarchy
    //if the item.Name contains our addAsChildOf string
    //we'll add the newChild as a child to the item.
    HandleArmDef(item =>
    {
        if (item.Name.Contains(addAsChildOf))
        {
            item.Children.Add(new ArmDef(newChild));
        }
    }, parent);
}

there’s plenty more possibilities of things to do here if I’ve explained well enough!
More info on Linq by example

1 Comment more...

Unity: Toying with Scriptable objects and Custom Editors

by on May.18, 2011, under .net, csharp, Scripting, Unity

Expanding on the previous post about Scriptable Objects as an easy way to save/edit some data at runtime, here’s a small extension looking at an example for a simple custom editor that appears when you select the Scriptable Object we built previously in the Project view.

Here’s what you get clicking on the file previously:
Before Custom Editor

And the result with the custom editor:
with editor

Now, this isn’t the sexiest of examples, but hopefully it helps someone along the way!

First a small addition to the data class, lets add a constructor to it, since our input values are going to be basically the same every time (Mimic ‘this’ camera), we’ll add a secondary constructor for the class that just takes a unity Camera as input and records the values for us.

We’ll use the same ‘CameraLocationHolder.cs’ from below:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
//File: CameraLocation.cs
using UnityEngine;
using System.Collections;
//the Serializable attribute is key to what makes the custom class visible in the editor !
[System.Serializable]
public class CameraLocation
{
    public string name;
    public bool isOrtho;
    public float orthoSize;
    public float fov;
    public Vector3 position;
    public Quaternion rotation;
 
    //base constructor that builds an empty CamLoc
    public CameraLocation()
    {
        name = string.Empty;
        isOrtho = false;
        orthoSize = -1;
        fov = -1;
        position = Vector3.zero;
        rotation = Quaternion.identity;
    }
    //build a CamLoc from a supplied camera
    public CameraLocation(Camera c)
    {
        if (c)
        {
            name = "New Camera";
            isOrtho = c.isOrthoGraphic;
            orthoSize = c.orthographicSize;
            fov = c.fov;
            position = c.transform.position;
            rotation = c.transform.rotation;
        }
    }
}

And we replace the ‘CameraListboxWizard’ that we previously used to edit the Scriptable Object with this (very similar) ‘Custom Editor’

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
//File: CameraLocationEditor.cs -- Place in an Editor directory under the unity project
 
using UnityEngine;
using UnityEditor;
 
//this limits the editor to running on object that have type CameraLocationHolder
[CustomEditor(typeof(CameraLocationHolder))]
public class CameraLocationEditor : Editor {
 
    //This will just be a shortcut to the target, ex: the object you clicked on.
    private CameraLocationHolder db;
 
	void Awake()
	{
		db=(CameraLocationHolder)target;
	}
 
	public override void OnInspectorGUI()
	{
		GUILayout.BeginVertical();
        if (GUILayout.Button("Add current Camera.main position", GUILayout.Height(50)))
        {
			Camera m = Camera.main;
			if(m)
			{
                //Add the new camera to the list
                db.content.Add(new CameraLocation(m));
                //tell the unity editor that the list has changed and needs to be serialized
                EditorUtility.SetDirty(target);
			}			
        }
 
        GUILayout.Space(5);
        DisplayCurrentCamList();
        GUILayout.Space(5);
        GUILayout.EndVertical();
    }
 
    void DisplayCurrentCamList()
    {
		if (db.content.Count == 0)
			return;
 
        //For each of the cameralocations in the saved scriptable object db
        for (int i = 0; i < db.content.Count; i++ )
        {
            EditorGUILayout.BeginHorizontal();
 
            //A text field to rename the camera
            db.content[i].name = GUILayout.TextField(db.content[i].name, GUILayout.Width(200));
 
            //a few simple buttons to juggle the array order
            if (GUILayout.Button("UP"))
            {
                if (i > 0)
                {
                    CameraLocation item = db.content[i];
                    db.content.RemoveAt(i);
                    db.content.Insert(i - 1, item);
                }
            }
            if (GUILayout.Button("DN"))
            {
                if (i < db.content.Count)
                {
                    CameraLocation item = db.content[i];
                    db.content.RemoveAt(i);
                    db.content.Insert(i + 1, item);
                }
            }
            if (GUILayout.Button("Remove"))
            {
                db.content.Remove(db.content[i]);
            }
 
            //sets the maincamera to be at the selected cameralocation
            if (GUILayout.Button("MoveTo"))
            {
                MoveToView(db.content[i]);
            }
 
            //updates the cameralocation to match up with the Camera.main (Usable at runtime or at edit time)
            if (GUILayout.Button("UPDATE"))
            {
                UpdateView(db.content[i]);
			}	
			EditorUtility.SetDirty(target);
            EditorGUILayout.EndHorizontal();
        }
    }
 
	void MoveToView(CameraLocation cam)
	{
		Camera MainCamera = Camera.mainCamera;
	    Quaternion currentRotation = cam.rotation;
 
		//if we had an ortho cam, lets just set it there before we start moving.
		MainCamera.orthographicSize = cam.orthoSize;
		MainCamera.isOrthoGraphic = cam.isOrtho;
 
	    //a parent check incase we have a 'fpswalker' type script setup also
	    if (MainCamera.transform.parent)
	    {
	        //move the parent object
	        MainCamera.transform.parent.position = cam.position;
 
	        Vector3 ang = currentRotation.eulerAngles;
 
	        //rotate the parent object on the Y axis
	        MainCamera.transform.parent.rotation = Quaternion.Euler(0, ang.y , 0);
 
	        //rotate the child camera on the X and Z axis
	        MainCamera.transform.localRotation = Quaternion.Euler(ang.x , 0, ang.z );
 
	    }
	    else // there is no FPSWalker parent, so just move the camera itself
	    {
	        MainCamera.transform.position = cam.position;
	        MainCamera.transform.rotation = currentRotation;
	    }
 
	    MainCamera.fov = cam.fov;	
	}
 
	void UpdateView(CameraLocation cam)
	{
		Camera MainCamera = Camera.mainCamera;
	    cam.isOrtho = Camera.main.isOrthoGraphic;
		cam.orthoSize = Camera.main.orthographicSize;
		cam.position = Camera.main.transform.position;
    	cam.rotation = Camera.main.transform.rotation;
		cam.fov = MainCamera.fov;
	}	
}

Now having built that up, we can simply click on the asset file that we had created previously and see the editor in the Inspector panel, making it easier to handle multiple lists.

6 Comments more...

Some Useful Unity3D resources

by on Feb.16, 2011, under Unity, Unity3D

Since I’m doing a lot of unity here lately, and starting to lay down some teaching to coworkers, here are a couple useful references for unity3d stuff that aren’t too commonly thrown out there.

http://feedity.com/rss.aspx/unifycommunity-com/UVtQUlVb – RSS feed that updates when new scripts are added to the community WIKI (And feedity in general is pretty nifty.)

http://www.google.com/cse/home?cx=002470491425767499270:iugs1ezlsfq a google custom search engine maintained by some anonymous saint, searches the unity references, answers, forums, and a few other popular blogs.

http://www.google.com/cse/home?cx=001712401338047450041:csfhqk-trfa a second google CSE that includes MSDN and the wiki among others, for when you need that extra bit of microsoft reference.

Also, don’t neglect the boys in irc.freenode.net #unity3d, definitely the single most useful resource @ http://webchat.freenode.net/ or your friendly neighborhood irc client.

*Oops! bolger fail. edit to fix the links.

Leave a Comment : more...

Unity: toying with Scriptable Objects

by on Oct.19, 2010, under Architecture, csharp, Scripting, Unity3D

In architecture there is often a need to keep track of and have handy many different viewpoints of a model at a time. with my early unity experiments, I made the mistake of creating a Camera for every Viewpoint that I wanted to have available in my scene. this was faulty for several reasons, not the least of which was that every ‘Camera’ object in a scene is actually rendering everything in the scene at once. Each camera has a few fun/useful options in the Clear Flags that makes this useful for things like rendering UI overlays with 3d objects and other assorted multipass ideas, but thats a whole ‘nother story.

Next attempt was to just import a large list of dummy objects in from 3dsmax that would represent camera orientations and allow the end user to select from these dummys and align the camera with the dummy objects using some UI bits in unity. This worked fine, but there is the base problem that the user cannot modify/edit these objects easily and have the changes save across play sessions.

This is where Scriptable Objects started to come in to play. (Taken partly from the unity Character Demo where they were explaining some of the joys of AssetBundles if you wish to look up more in depth examples.) this may not be the most useful demo of Scriptable objects, but this is a case study for when you are trying to store Runtime data for use at editortime.

There are several parts: First and probably the only important part is the Holder, this is really just a list of ‘Somethings’ that we save to disk. in our case, CameraLocations.

//CameraLocationHolder.cs
using System.Collections.Generic;
using UnityEngine;
 
public class CameraLocationHolder : ScriptableObject {
    public List<CameraLocation> content;
    public CameraLocationHolder(List content)
    {
        this.content = content;
    }
}

the above empty list consists of ‘CameraLocation’s which is just a simple class I used to store a few variables that are needed to re-create a Viewpoint. Camera locations is a Serializable Class (Serializable in this case means that its viewable in the inspector in Unity the same way you would view for instance a Vector3..) and it has some default values in there of pulling the maincameras location/rotation and field of view.

//CameraLocation.cs
using UnityEngine;
using System.Collections;
 
[System.Serializable]
public class CameraLocation
{
    public string name = "New Camera";
    public float fov = Camera.main.fov;
    public Vector3 position = Camera.main.transform.position;
    public Quaternion rotation = Camera.main.transform.rotation;
}

Now, we create an asset in the editor where we will store our list of CameraLocation’s, this can be placed anywhere, its a small file so i tend to just store it in Resources for easy finding later.

//CreateCameraLocationAsset.cs
using UnityEngine;
using UnityEditor;
using System;
using System.Collections.Generic;
 
public class CreateCameraLocationAsset
{
    [MenuItem("Custom/Cameras/Create camera location holder")]
    public static void CreateMyAsset()
    {
        ListList<CameraLocation> content = new List<CameraLocation>();
        CameraLocationHolder asset = new CameraLocationHolder(content);  //scriptable object 
        AssetDatabase.CreateAsset(asset, "Assets/CameraLocationDatabase.asset");
        AssetDatabase.SaveAssets();
        EditorUtility.FocusProjectWindow();
        Selection.activeObject = asset;
    }
}

Now you have an ‘Asset’ where you can store/load/erase/whatever in a list of CameraLocations at runtime, and the results will still be available at Editor time, and at future Runs of the application.

For instance, you can browse around your scene as you would normally with your FlyingCameraScriptOfDoom, and occasionally call a script to add your current camera.main viewpoint to the list.

or accessing it as a runtime resource for use by a normal GUI pulldown:

//Excerpts from CameraListBox.cs
 
//declaring a list of stuff
            public List<CameraLocation> camList = new List<CameraLocation>(); //Declare our list of stuff
 
//....
 
//somewhere in Start() we load the list from the DB
            Object o = Resources.Load("CameraLocationDatabase", typeof(CameraLocationHolder));
            CameraLocationHolder CameraLocationDB = (CameraLocationHolder)o;
//fill our list of stuff, with the content from the Holder in the asset
            camList = CameraLocationDB.content;
 
//....
 
//and somewhere in our OnGUI()
            GUILayout.BeginVertical(GUILayout.Width(120));
            for (int i = 0; i &lt; camList.Count; i++)
            {
                if (GUILayout.Button(camList[i].name))
                {
                    SelectedListItem = i;//Set the index for our currrently selected item
                    CameraLocation cam = camList[SelectedListItem];
                    StartCoroutine(SmoothMoveCamLoc(cam.position, cam.rotation, cam.fov, duration));
                   }
            }
            GUILayout.EndVertical();
 
//the SmoothMoveCamLoc just lerps between the current positions/rotations/fov's 
//and the desired ones that were stored in the CameraLocation.
//although mine has a few other random logic items in there to decide
//which style of camera i have currently and deal with any parents/targets/initializing/etc
//needed if we move the camera.

For instance, editing the list by an Editor script, that builds a list for you to Add/Remove/Rename/and Re-Order the list, because thats oddly hard to do with the built in object interface? Maybe this is where a custom inspector should be for the .asset type? hmm! things to play with in the future.

//CameraListboxWizard.cs
using UnityEngine;
using UnityEditor;
using System.Collections.Generic;
 
public class CameraListboxWizard : ScriptableWizard
{
    public string DBLocation = "Assets/Resources/CameraLocationDatabase.asset";
    private CameraLocation camLocation;
    static private Object dbaseAsset;
    static private CameraLocationHolder db;
 
    [MenuItem("Custom/Cameras/CameraListBox - Database Helper")]
    static void DoSet()
    {
        ScriptableWizard.DisplayWizard("Add view to CameraListBox", typeof(CameraListboxWizard), "DONE");
    }
 
    void OnWizardUpdate()
    {
        helpString = "This will control the Camera Location asset database\n, and allow you to 'save' the Camera.main's position\n while in the player mode\n";
 
        UpdateMyAssetLocation();
    }
 
    void AddExistingCameras()
    {
        int count = Camera.allCameras.Length;
        for (int i = 0; i &lt; count; ++i)
        {
            //except for the main camera, which we'll use to transition between the others
            if (Camera.allCameras[i] != Camera.main)
            {
                camLocation = new CameraLocation();
                camLocation.name = Camera.allCameras[i].name;
                camLocation.fov = Camera.allCameras[i].fov;
                camLocation.position = Camera.allCameras[i].transform.position;
                camLocation.rotation = Camera.allCameras[i].transform.rotation;
                AddCamera(camLocation);
            }
        }
    }
 
    void OnWizardCreate()
    {
        Debug.Log("done");
    }
 
    void AddCamera(CameraLocation cam)
    {
        db.content.Add(cam);
        EditorUtility.SetDirty(dbaseAsset);
    }
 
    void OnGUI()
    {
        EditorGUILayout.BeginVertical();
         EditorGUILayout.BeginHorizontal();
        GUILayout.Label(name, GUILayout.MaxWidth(0));
        DBLocation = EditorGUILayout.TextField(DBLocation);
        if (GUILayout.Button("Browse"))
        {
            DBLocation = EditorUtility.OpenFilePanel(name, DBLocation, "asset");
            UpdateMyAssetLocation();
        }
        EditorGUILayout.EndHorizontal();
 
        if (GUILayout.Button("Update Database Location"))
        {
            UpdateMyAssetLocation();
        }
 
        EditorGUILayout.Space();
        if (GUILayout.Button("Add current Camera.main position", GUILayout.Height(50)))
        {
            camLocation = new CameraLocation();
            AddCamera(camLocation);
        }
        EditorGUILayout.Space();
        DisplayCurrentCamList();
        EditorGUILayout.Space();
 
        if (GUILayout.Button("Add all existing 'Cameras' to the list\n this can 'import' max camera locations", GUILayout.Height(50)))
        {
            AddExistingCameras();
        }
        EditorGUILayout.EndVertical();
    }
 
    void DisplayCurrentCamList()
    {
        for (int i = 0; i &lt; db.content.Count; i++ )
        {
            EditorGUILayout.BeginHorizontal();
            db.content[i].name = GUILayout.TextField(db.content[i].name, GUILayout.Width(200));
            if (GUILayout.Button("UP"))
            {
                if (i &gt; 0)
                {
                    CameraLocation item = db.content[i];
                    db.content.RemoveAt(i);
                    db.content.Insert(i - 1, item);
                }
            }
            if (GUILayout.Button("DN"))
            {
                if (i &lt; db.content.Count)
                {
                    CameraLocation item = db.content[i];
                    db.content.RemoveAt(i);
                    db.content.Insert(i + 1, item);
                }
            }
            if (GUILayout.Button("Remove"))
            {
                db.content.Remove(db.content[i]);
            }
            //db.SetDirty();
            EditorUtility.SetDirty(dbaseAsset);
            EditorGUILayout.EndHorizontal();
        }
    }
 
    public void UpdateMyAssetLocation()
    {
        dbaseAsset = AssetDatabase.LoadAssetAtPath(DBLocation, typeof(CameraLocationHolder));
        CameraLocationHolder CameraLocationDB = (CameraLocationHolder)dbaseAsset;
        db = CameraLocationDB;
    }
    public void CreateMyAsset()
    {
        List content = new List();
        CameraLocationHolder asset = new CameraLocationHolder(content);  //scriptable object 
        AssetDatabase.CreateAsset(asset, "CameraLocationDatabase");
        AssetDatabase.SaveAssets();
        EditorUtility.FocusProjectWindow();
        Selection.activeObject = asset;
    }
}
12 Comments more...

Recursive Tiff to PDF

by on Apr.01, 2010, under Admin, Batching, utility

Another one of those semi random requests that turned out to be more of a pain in the arse than it really should be. the mission: Converting a large mass of scanned tiff CAD drawings into multi page PDF’s

There are two ways to do this fairly quickly, the easy way if you don’t have many individual sets to work with is to simply drag all the tif images into the window in Acrobat Pro and save the file from there.

If however, like us, you have to convert 42 gigs or so of tiff’s spread across 204 folders and group the PDF’s by the folders, you can programmatically do it via some batch scripting and the use of a few free command line utilities from the internet.

First thing to do is convert each individual TIF file to a PDF, for this, we use Imagemagick this is a tool that can do many things but we’re really just going to abuse the ‘convert’ application for now.

After installation, you can then in the dos prompt type ‘convert file.tif file.pdf’

For our purposes we want to convert ALL of the files, so we can set it up in a recursive for loop using

Saved into a file “Convert all Tiffs in Subdirectory to PDFs.bat” or something equally exciting:
FOR /R %%a IN (*.tif) DO convert "%%a" "%%a.pdf"

After that is done we will use a second utility to ‘join’ the pdf’s the PDF toolkit

With this util we can modify pdf files quickly; we’ll use the Append function

pdftk *.pdf cat output combined.pdf

This will concatenate all pdf’s in the current directory and write them into ‘combined.pdf’ in the same directory. Great!

To expand this into a large subdirectory we do a little more batch scripting

The below is saved into a file “Join all PDF in directory tree.bat”

FOR /F "delims=" %%a in ('DIR /S /B /AD ^|SORT') DO (
CD "%%a"
pdftk *.pdf cat output "%%a_joined.pdf"
)

This will output a combined PDF for each of the individual directories.

A good time was had by all.

Imagemagick looks to have quite a few useful applications and will warrant some more exploration soon.

1 Comment more...

Interactive Lab Demo

by on Feb.10, 2010, under Architecture, csharp, Interactive, Unity3D

Here is another Architectural interactive test project. This time, a laboratory setup, a tighter interior space, set up without any real lighting (1 direct light + ambient colors + Ambient Occlusion currently) created with the goal of customizing the unity asset pipeline to be a little more friendly for architectural work.

First, the Project, roughly 6mb, clicking on the screenshot will open the project in a new window. its another quickly evolving work in progress, Warning: No consideration has been taken currently for speed on older hardware, ~100fps on an 8800GT was the target.
Screenshot

Some findings and random ramblings on arch and interactive 3d:

  • unlike with general gaming, with architecture you are in most cases going to already have a model provided in some form or another. odds are, the model is going to be pretty horrific, as arch models tend to be built for the purpose of either:
    • A: Viz… the Viz model will require a lot of cleanup in terms of material work, and stripping down a lot of the high poly fluff required for rendering.
    • B: Design Development… the DD model is normally plagued more by bad modeling conventions, no naming, poor materials, materials not fully assigned, or mangled geometry due to being stretched and squished for days on end to keep up with a design.
  • you will probably still want to break up the file by some logical layer type system, by default this will land you with a unique material for each material of each fbx file
  • a small AssetPostProcessor script that works on the OnMaterialAssign function can make the process of sharing materials across multiple FBX files much easier to handle, with the added bonus of being able to easily progressively build up a library of reusable realtime materials
 
//file: MaterialsPostProcessor.cs
//goal: Share a library of materials across different assets based on the name of material supplied in the FBX file
// Dave Buchhofer - 02.10.2010
 
using UnityEngine;
using UnityEditor;
using System.Collections;
 
public class MaterialsPostProcessor : AssetPostprocessor
{
    Material OnAssignMaterialModel (Material material, Renderer renderer)
    {
        //The path where you keep your "Standard" library of materials
        string StandardMatPath = "Assets/DMGStandardMaterials/" + material.name + ".mat";
 
        //the path where you want to keep your "temp" materials, that weren't found above
        //So you have a logical place to look for materials that need editing and tweaking for realtime work.
        string TemporaryMatPath = "Assets/Mesh/Materials/" + material.name + ".mat";
 
        //Check for it in the standard 
        if (AssetDatabase.LoadAssetAtPath(StandardMatPath, typeof(Material)))
        {
            Debug.Log("FOUND: " + StandardMatPath);
            return (Material)AssetDatabase.LoadAssetAtPath(StandardMatPath, typeof(Material));
        }
        //Else check to see if we've already built one in the temp path
        if (AssetDatabase.LoadAssetAtPath(TemporaryMatPath, typeof(Material)))
        {
            Debug.Log("FOUND temp: " + TemporaryMatPath);
            return (Material)AssetDatabase.LoadAssetAtPath(TemporaryMatPath, typeof(Material));
        }
        //Or create it?
        material.shader = Shader.Find("Diffuse");
        AssetDatabase.CreateAsset(material, TemporaryMatPath);
        Debug.Log("CREATED temp: " + TemporaryMatPath);
        return material;
    }
}

also check out the previous arch experiment a few posts down: Basketball Arena

3 Comments :, , , more...

Recursive file size listing output to excel

by on Dec.30, 2009, under Admin, Batching, Scripting, vbscript

Today I had to find all JPG files over a certain size in a large directory tree of a few thousand images, to find some poorly compressed jpg’s, fix the compression, and replace them on a website. Todays cavaet, the CMS is pretty weak and only lets you update 1 image at a time, so you can’t just blindly resave everything and reupload, todays Second cavaet, I dont have server level access to said website, so all the normal automated ways are out of luck.

joy.

So I found a little snip of code online, and threw a pretty simple hack job together to narrow the scope some. (PS: If you do end up doing any vbscript, I’d suggest VbsEdit for an IDE)

Anyway! the code:

It searches through a directory tree of your choosing
for any file matching the filetype: “jpg”
that is above a size threshold: fileSizeThreshold (In this case, 2kb.. so essentially everything)
and outputs it into an excell sheet with full path and size information
so that you can easily Sort, Filter, Delegate, or use as data for further automating!

' // **************************************
' //    ComputerHighGuys recursive search
' //     
' //    Date Created: 20 Aug 07
' //	http://www.tek-tips.com/faqs.cfm?fid=6716
' //
' //	Adjusted to a jpg file search and 
' //	added excel output for easy sorting/filtering
' //	
' // **************************************
 
' // If we'd like to save the output into an excel sheet
WriteExcel = "True"
 
Dim objexcel
excelRow=2
 
' // value the filesize needs to exceed to be visible in the output
fileSizeThreshold=2
 
If WriteExcel = "True" Then
	' // Create the Excel sheet to drop the information into
	Set objExcel = createobject("Excel.application")   
	objexcel.Workbooks.add
	objexcel.Cells(1, 1).Value = "Folder Name"
	objexcel.Cells(1, 2).Value = "Filename"
	objexcel.Cells(1, 3).Value = "Filesize"
	objexcel.Cells(1, 4).Value = "Filesize Unit"
	objexcel.Visible = True
	Wscript.Sleep 300
End If
 
' // Directory to search
searchDir = "C:\"
 
set objFSO=CreateObject("Scripting.FileSystemObject")
Set objFolder = objFSO.GetFolder(searchDir)
Set colFiles = objFolder.Files
 
' // Launch the function
ScanSubFolders(objFolder)
 
' // recursive function that will search through all subfolders for a specified
' // filetype and output some information about the files to an Excel Sheel
Sub scanSubFolders(objFolder)
	' // Grab sub folders    
    Set colFolders = objFolder.SubFolders
 
    For Each objSubFolder In colFolders
 
    	' // the files to search
        Set colFiles = objSubFolder.Files
        For Each objFile in colFiles
 
            ' // the extension of the filetype to search for  
            If lcase(Right(objFile.Name,3)) = "jpg" Then
 
                ' // Getting File size in KB
                If round(objFile.Size/1024,1) > fileSizeThreshold Then
 
                	' // echo the files to the console
                	WScript.Echo objSubFolder.Path & "  " & objFile.Name & "  " & round(objFile.Size/1024,1)  & "KB"
 
					' // write various shit to excel~                    
                    If WriteExcel = "True" then
                        wscript.Echo objSubFolder.Path & "  " & objFile.Name & "  " & round(objFile.Size/1024,1)  & "KB"
                    	objexcel.Cells(excelRow, 1).Value = objSubFolder.Path
    					objexcel.Cells(excelRow, 2).Value = objFile.Name
    					objexcel.Cells(excelRow, 3).Value =  round(objFile.Size/1024,1)
    					objexcel.Cells(excelRow, 4).Value =  "KB" 
    					excelRow=excelRow+1    
                    End If
                End if    
            End If
        Next
        ScanSubFolders(objSubFolder)
    Next
End Sub
2 Comments :, , , , , more...

ServerUtilities v.85

by on Jun.23, 2009, under Batching, maxscript, vray

Its been a while since I’ve posted anything, so its time for the habitual blog sorry/script update!

We recently picked up VrayRT, and i found myself again vnc’ing over to farm machines to restart services and well, that just wont do!

So I did a small update to the ServerUtilities script adding a few services, and doing a little work on the underbelly of it to make it a bit more stable. also packaged it into an installer because I was getting a few people that had just installed it strangely and were having issues.

screenshot and download link

screenshot and download link

    — .65 – added vray 2010 to services
    — .75 – added vrayRT
    — .75 – convert to simpler dos prompt
    — .75 – added an optional pause to the end of the commandline based tools
    — .75 – fixed the invert button in server selection
    — .75 – prints the commandline to the listener
    — .80 – convert to dos based loop for multiple systems (if pause: hit enter to advance to next machine in loop..)
    — .85 – update to support spaces in the max file path! oops!

Download the Installer

5 Comments more...

Looking for something?

Use the form below to search the site:

Still not finding what you're looking for? Drop a comment on a post or contact us so we can take care of it!