Java-Gaming.org Hi !
 Featured games (88) games approved by the League of Dukes Games in Showcase (681) Games in Android Showcase (196) games submitted by our members Games in WIP (744) games currently in development
 News: Read the Java Gaming Resources, or peek at the official Java tutorials
 Show Posts Pages: [1] 2 3
1  Game Development / Newbie & Debugging Questions / Re: Algorithm to loop through the color spectrum? on: 2011-12-06 06:38:04
 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 `import java.awt.Color;import java.awt.Dimension;import java.awt.Graphics;import java.util.Random;import javax.swing.JFrame;import javax.swing.JPanel;/* * To change this template, choose Tools | Templates * and open the template in the editor. *//** * * @author David */public class DrawPaletteShiz {    /**     * @param args the command line arguments     */    public static void main(String[] args) throws Throwable {      JFrame jf = new JFrame("Frame");      final Dimension size = new Dimension(100,100);      JPanel jp = new JPanel() {         @Override         public void paint(Graphics g) {            float f = new Random().nextFloat();            for (int i = 0; i < size.width; i++) {               for (int j = 0; j < size.height; j++) {                  g.setColor(Color.getHSBColor(f, 1.0f-(float)i/size.width, 1.0f-(float)j/size.height));                  g.drawLine(i,j,i,j);               }            }            for (int i = 0; i < size.height; i++) {               g.setColor(Color.getHSBColor((float)i/size.height,1.0f,1.0f));               g.drawLine(100,i,130,i);            }         }      };      jf.setResizable(false);      jp.setPreferredSize(new Dimension(size.width+30,size.height));      jf.getContentPane().add(jp);      jf.pack();      jf.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);      jf.show();      while (true) {         jp.repaint();         Thread.sleep(10);      }    }}`

have fun brah
2  Game Development / Performance Tuning / Re: WeakHashMap, canonical cache and value recycling? on: 2011-08-16 15:16:31
well then you could just override hashCode and use this

 1  2  3  4  5  6  7 `public int hashCode() { K key = get(); if (key == null) {  return 0; } return key.hashCode();}`

put:

 1  2  3  4 `public void put(K key, V value) {KeyValuePair kvp = new KeyValuePair(key, value);       M.put(kvp, kvp);   }`

full:

 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 `class WeakHashMap {    ReferenceQueue Q = new ReferenceQueue();    HashMap> M = new HashMap();    class KeyValuePair extends WeakReference {        V value;                 KeyValuePair(K key, V value) {            super(key, Q);            this.value = value;        }            @Override      public int hashCode() {         K key = get();         if (key == null) {            return 0;         }         return key.hashCode();      }    }         public void put(K key, V value) {      KeyValuePair kvp = new KeyValuePair(K, V);        M.put(kvp, kvp);    }         public V get(K key) {        KeyValuePair KVP = M.get(key);        if (KVP != null) {         return KVP.value;        }        return null;    }         public void recycle() {        while (true) {            KeyValuePair KVP = Q.poll();            if (KVP != null) {                V value = KVP.value;                // do something with value            }        }    }         // etc..     }`
3  Game Development / Performance Tuning / Re: WeakHashMap, canonical cache and value recycling? on: 2011-08-16 13:31:14
 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 `class WeakHashMap {   ReferenceQueue Q = new ReferenceQueue();   HashMap> M = new HashMap();   class KeyValuePair extends WeakReference {      V value;            KeyValuePair(K key, V value) {         super(K, value);         this.value = value;      }   }      public void put(K key, V value) {      M.put(key, new KeyValuePair(K, V));   }      public V get(K key) {      KeyValuePair KVP = M.get(key);      if (KVP != null) {       return KVP.value;      }      return null;   }      public void recycle() {      while (true) {         KeyValuePair KVP = Q.poll();         if (KVP != null) {            V value = KVP.value;            // do something with value         }      }   }      // etc..   }`

HTH

- David

Thanks, David, but isn't that storing the value as a weak reference and not the key?  I need the same semantics as WeakHashMap, with the key being wrapped in a weak reference.

The key is the weak reference, look at the KVP constructor

Btw typo:

 1 `super(K, value);`

should be

 1 `super(key, Q);`
4  Game Development / Performance Tuning / Re: WeakHashMap, canonical cache and value recycling? on: 2011-08-16 12:59:45
Use a ReferenceQueue when adding keys/values to the map
Do you mean basically reimplement WeakHashMap, which I'm happy to do but am hoping not to have to, or just wrapping the keys in a second weak reference?  I'm pretty sure the second wouldn't work because by the time I try and get the value out of the map it'll already be nulled.

I don't understand your example - aren't you calling map.put() with a single argument?

WeakHashMap is very unlikely to be a good choice. Its non-strongly-referenced references will be cleared on every GC (in Suns/Oracles JVM), as opposed to only during a full GC, which you probably want. It can easily run a couple of times per second, which kinda invalidates it as a cache.

What you probably need is a SoftHashMap, which (annoyingly) is not in the standard API. Pick a 3rd party one.

I could have put bets on someone saying this!

No, I want a WeakHashMap!  I understand it's not much use as a cache in itself, though neither usually would a SoftHashMap be - you want a map to soft referenced values.  In fact, I have a slight feeling of deja vu - I'm pretty sure we've had that conversation in another thread.

There's a good article here on this http://www.codeinstructions.com/2008/09/weakhashmap-is-not-cache-understanding.html

So why do I want a WeakHashMap?  Quote from same article on "What is the WeakHashMap good for?"

Quote
If the WeakHashMap is no good for caching, then what is it good for? It is good to implement canonical maps. Lets say you want to associate some extra information to an object that you have a strong reference to. You put an entry in a WeakHashMap with the object as the key, and the extra information as the map value. Then, as long as you keep a strong reference to the object, you will be able to check the map to retrieve the extra information. And once you release the object, the map entry will be cleared and the memory used by the extra information will be released.

This is basically what I'm trying to do - associate a cached object (most likely soft referenced) with a particular object - should have been clearer earlier!  I want the keys to be GC'd as soon as possible, because once the key is unreferenced, the value is unreachable.  The problem is, I want to get the values added to a queue on clearing the key so that I can do other things with them, rather than just having them set to null.

 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 `class WeakHashMap {   ReferenceQueue Q = new ReferenceQueue();   HashMap> M = new HashMap();   class KeyValuePair extends WeakReference {      V value;            KeyValuePair(K key, V value) {         super(K, Q);         this.value = value;      }   }      public void put(K key, V value) {      M.put(key, new KeyValuePair(K, V));   }      public V get(K key) {      KeyValuePair KVP = M.get(key);      if (KVP != null) {       return KVP.value;      }      return null;   }      public void recycle() {      while (true) {         KeyValuePair KVP = Q.poll();         if (KVP != null) {            V value = KVP.value;            // do something with value         }      }   }      // etc..   }`

HTH

- David
5  Game Development / Performance Tuning / Re: WeakHashMap, canonical cache and value recycling? on: 2011-08-16 11:26:40
Use a ReferenceQueue when adding keys/values to the map

example:

 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24 `ReferenceQueue mQueue = new ReferenceQueue();Map map = new HashMap();class MyData extends SoftReference/WeakReference {  YourKey key;  MyData(Referent r, YourKey key) {   super(r, mQueue);   this.key = key;  }}to add to cachevoid put(YourKey key, Referent r) {  map.put(new MyData(r, key));}then to read the recycled objectswhile (true) { MyData data = mQueue.poll(); if (data == null) break; ...process here}`

HTH

- David
6  Game Development / Newbie & Debugging Questions / Re: List<List<List<Object>>> or List[][] on: 2011-08-08 12:53:18
Thanks for your replies! It made me rethink how I should do this. I have some questions to the code Dx4 and counterp provided.

 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  135  136  137  138  139  140  141  142  143  144  145  146  147  148  149  150  151  152  153  154  155  156  157  158  159  160  161  162  163  164  165  166  167  168  169  170  171  172  173  174  175 `public class TileHashtable {   public class Entry {      Entry next;      int x;      int y;      int hash;      UserLayer.Tile value;      public Entry(Entry next, int x, int y, int hash, Tile value) {         this.next = next;         this.x = x;         this.y = y;         this.hash = hash;         this.value = value;      }      int hash() {         return (x << 5) + y;      }      public int getHash() {         return hash;      }      public Entry getNext() {         return next;      }      public Tile getValue() {         return value;      }      public int getX() {         return x;      }      public int getY() {         return y;      }   }      int size;   Entry [] table;   public TileHashtable() {      table = new Entry[64];   }    static int hash(int h) {        // This function ensures that hashCodes that differ only by        // constant multiples at each bit position have a bounded        // number of collisions (approximately 8 at default load factor).        h ^= (h >>> 20) ^ (h >>> 12);        return h ^ (h >>> 7) ^ (h >>> 4);    }   static int getPointHash(int x, int y) {      return hash((x & 0xFFFF) << 16) | (y & 0xFFFF);   }    static int indexFor(int h, int length) {        return h & (length-1);    }   public void clear() {      for (int i = 0; i < table.length; i++) {         table[i] = null;      }      size = 0;   }   public UserLayer.Tile lookup(int x, int y, boolean createIfNotExist) {      int hash = getPointHash(x, y);      Entry e;      int index = indexFor(hash, table.length);      for (e = table[index]; e != null; e = e.next) {         if (e.hash == hash && e.x == x && e.y == y) {            return e.value;         }      }      if (createIfNotExist) {         UserLayer.Tile val = new UserLayer.Tile();         table[index] = new Entry(table[index], x, y, hash, val);         size++;         return val;      }      return null;   }   public int size() {      return size;   }   public boolean isEmpty() {      return size == 0;   }   public Iterator entryIterator() {      return new EntryIterator();   }   public Iterator valueIterator() {      return new ValueIterator();   }   public Iterator keyIterator() {      return new KeyIterator();   }    private abstract class HashIterator implements Iterator {        Entry next;   // next entry to return        int index;      // current slot        Entry current;   // current entry        HashIterator() {            if (size > 0) { // advance to first entry                Entry[] t = table;                while (index < t.length && (next = t[index++]) == null)                    ;            }        }        public final boolean hasNext() {            return next != null;        }        final Entry nextEntry() {            Entry e = next;            if (e == null)                throw new NoSuchElementException();            if ((next = e.next) == null) {                Entry[] t = table;                while (index < t.length && (next = t[index++]) == null)                    ;            }       current = e;            return e;        }        public void remove() {         throw new UnsupportedOperationException("My Dick");        }    }    private final class ValueIterator extends HashIterator {        public UserLayer.Tile next() {            return nextEntry().value;        }    }    private final class KeyIterator extends HashIterator {        public void next(Point p) {        }      public Void next() {         throw new UnsupportedOperationException("My Dick");      }    }    private final class EntryIterator extends HashIterator {        public Entry next() {            return nextEntry();        }    }}`

 1  2  3 `static int getPointHash(int x, int y) {      return hash((x & 0xFFFF) << 16) | (y & 0xFFFF);   }`

Why is it that we have to do that AND operation on 0xFFF? To me, it looks like it will return the same, if we exclude that. Also, I don't see a change if I exchange the OR operation with a "+". Can you please explain this?

This is only required if you need to make sure x and y dont exceed a short max.

 1  2  3  4  5  6  7 `static int hash(int h) {        // This function ensures that hashCodes that differ only by        // constant multiples at each bit position have a bounded        // number of collisions (approximately 8 at default load factor).        h ^= (h >>> 20) ^ (h >>> 12);        return h ^ (h >>> 7) ^ (h >>> 4);    }`

This also I can't seem to get much meaning out of. I'm unsure what this ensures, and I'd very much like a more in-depth explanation that the comment provided

basically this just mixes up the bits of a hashcode, as some hashcodes only differ in bits in lower bits, this function will ensure the bits become more spread from upper->lower

 1  2  3 `int hash() {         return (x << 5) + y;      }`

This again makes me feel like I don't really know what the hashes do. I though we'd use the last 16 bits to define y, and the first 16 ones to define x. Also, what happended to the AND-operation with the 0xFFFF byte?

this hash function doesnt really matter, you could make it anything that involves x and y, it would work the same.
7  Game Development / Newbie & Debugging Questions / Re: List<List<List<Object>>> or List[][] on: 2011-08-07 13:29:52
Here's a solution that will be faster than the one provided by counterp but uses the same semantics (there is no unboxing/boxing each read)

 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  135  136  137  138  139  140  141  142  143  144  145  146  147  148  149  150  151  152  153  154  155  156  157  158  159  160  161  162  163  164  165  166  167  168  169  170  171  172  173  174  175 `public class TileHashtable {   public class Entry {      Entry next;      int x;      int y;      int hash;      UserLayer.Tile value;      public Entry(Entry next, int x, int y, int hash, Tile value) {         this.next = next;         this.x = x;         this.y = y;         this.hash = hash;         this.value = value;      }      int hash() {         return (x << 5) + y;      }      public int getHash() {         return hash;      }      public Entry getNext() {         return next;      }      public Tile getValue() {         return value;      }      public int getX() {         return x;      }      public int getY() {         return y;      }   }      int size;   Entry [] table;   public TileHashtable() {      table = new Entry[64];   }    static int hash(int h) {        // This function ensures that hashCodes that differ only by        // constant multiples at each bit position have a bounded        // number of collisions (approximately 8 at default load factor).        h ^= (h >>> 20) ^ (h >>> 12);        return h ^ (h >>> 7) ^ (h >>> 4);    }   static int getPointHash(int x, int y) {      return hash((x & 0xFFFF) << 16) | (y & 0xFFFF);   }    static int indexFor(int h, int length) {        return h & (length-1);    }   public void clear() {      for (int i = 0; i < table.length; i++) {         table[i] = null;      }      size = 0;   }   public UserLayer.Tile lookup(int x, int y, boolean createIfNotExist) {      int hash = getPointHash(x, y);      Entry e;      int index = indexFor(hash, table.length);      for (e = table[index]; e != null; e = e.next) {         if (e.hash == hash && e.x == x && e.y == y) {            return e.value;         }      }      if (createIfNotExist) {         UserLayer.Tile val = new UserLayer.Tile();         table[index] = new Entry(table[index], x, y, hash, val);         size++;         return val;      }      return null;   }   public int size() {      return size;   }   public boolean isEmpty() {      return size == 0;   }   public Iterator entryIterator() {      return new EntryIterator();   }   public Iterator valueIterator() {      return new ValueIterator();   }   public Iterator keyIterator() {      return new KeyIterator();   }    private abstract class HashIterator implements Iterator {        Entry next;   // next entry to return        int index;      // current slot        Entry current;   // current entry        HashIterator() {            if (size > 0) { // advance to first entry                Entry[] t = table;                while (index < t.length && (next = t[index++]) == null)                    ;            }        }        public final boolean hasNext() {            return next != null;        }        final Entry nextEntry() {            Entry e = next;            if (e == null)                throw new NoSuchElementException();            if ((next = e.next) == null) {                Entry[] t = table;                while (index < t.length && (next = t[index++]) == null)                    ;            }       current = e;            return e;        }        public void remove() {         throw new UnsupportedOperationException("My Dick");        }    }    private final class ValueIterator extends HashIterator {        public UserLayer.Tile next() {            return nextEntry().value;        }    }    private final class KeyIterator extends HashIterator {        public void next(Point p) {        }      public Void next() {         throw new UnsupportedOperationException("My Dick");      }    }    private final class EntryIterator extends HashIterator {        public Entry next() {            return nextEntry();        }    }}`

replace UserLayer.Tile with whatever obj type you need

HTH

- David
8  Game Development / Game Mechanics / Re: Slight jerkiness on: 2011-07-28 11:17:43
I've had this problem, and apparently it's a bug in Windows ever since 98.

This is how I fixed it:

 1  2  3  4 `new Thread() {{ setDaemon(true); start(); }public void run() { while(true) {try {Thread.sleep(Integer.MAX_VALUE);}catch(Throwable t){} } }};`

9  Game Development / Newbie & Debugging Questions / Re: how to read keyboard and mouse input when window not in focus( if possible?) on: 2011-07-28 10:49:53
Well, first off I know a lot of you think I am writing  keylogger, and I s'pose I technically am. BUT not for any illegal intentions I assure you.

My plan is to write 2 programs, on the reads input, and then saves everything I did (or atleast the basics of it), and then another that reads the file and does the stuff(this one I know how to do).

But when I thought about itm ore, I realized that I wouldn't be able to listen to the input because my program would not be in focus.

So, if anybody has any suggestions, I would be very grateful

h3ckboy

here

 1  2  3  4  5 `package yourpckage;class yourclass { public static native int GetAsyncKeyState ( int v );}`

and

 1  2 `JNIEXPORT jint JNICALL Java_yourpackage_yourclass_GetAsyncKeyState  (JNIEnv *, jclass, jint v) { return GetAsyncKeyState(v); }`

remember to #include < windows.h > !!

example of usage:

 1  2  3  4  5  6 `for (int i = 0; i < 256; i++) { final int vk = VK_BASE + i; if (yourpackage.yourclass.GetAsyncKeyState(vk) > 0) {   fireKeyDown(vk); }}`

define VK_BASE to be the base of all virtual keys you want to read.

- David
 10 Discussions / Business and Project Management Discussions / Re: Advertising/Marketing for Mobile Games on: 2011-07-26 10:20:48 make some viral facebook pages (preferably a lot with 100k+ likes) and then post your links on the facebook pages = profit(you have to understand current social situations, etc for this to work though.)
 11 Game Development / Shared Code / Re: some new blendmodes (add, multiply, overlay, etc..) on: 2011-07-25 12:00:47 Quote from: nsigma on 2011-07-25 10:32:17Right, license headers changed to All-Permissive in RGBMath and RGBCompositehttp://code.google.com/p/praxis/source/browse/ripl/src/net/neilcsmith/ripl/rgbmath/RGBComposite.javahttp://code.google.com/p/praxis/source/browse/ripl/src/net/neilcsmith/ripl/rgbmath/RGBMath.javaI haven't look at adding in any of your changes or additional blend modes yet.  Did try various versions of the code you posted with continuously overlaying yellow but can definitely not replicate the drift to red - must be something we're doing differently outside of this code.Incidentally, as I was posting it in another topic, have you seen the various blend modes from http://filthyrichclients.org/ -> Examples -> Chapter 6.  They're for using with Java2D and aren't as optimal as the code we've got between us, but there might be some good pointers to some missing blend modes in there.Best wishes,  Neilthanks. I noticed the drift to red only happened when I used darkish yellow colors, eg 0xffffaa00, it didnt happen when I used brighter yellows.Took a look at their blendcomposite class and it seems like their blender would be orders of magnitudes slower than the one we have between us, though I agree it would help us write more blend modes. will write more blend modes when I get some time - David
 12 Game Development / Shared Code / Re: some new blendmodes (add, multiply, overlay, etc..) on: 2011-07-25 03:11:25 Quote from: nsigma on 2011-07-24 15:46:53I am interested, though, in why we're getting different results.  I wonder what we're doing differently  I'll be doing some work on Praxis in a couple of days, so might try plugging in your altered blend function and see if that causes it, but I can't see at the moment why any of your optimisation changes would alter the behaviour.  That altered mult() method, while more accurate, is also going to slow things down, and negate the optimisations you've made elsewhere - if something else is the root cause, it seems a shame to change it!yeah, I can only try to optimize further now, but in my application accuracy is more important than speed - in yours probably not so much as you need to render frames at a fast rate.If I optimize anything further I'll post it here - David
13  Game Development / Newbie & Debugging Questions / Re: Character movement with Mouse on: 2011-07-24 13:03:04
and now to move towards the point

 1  2  3  4  5 `x += movex;y += movey;if (x == mouseX && y == mouseY) { // you can use epsilon here stop();}`

hth
That is NOT going to work. In 99.999999999999% of all cases it will just overshoot and continue forever. You would need to change the if-statement to detect overshooting and then set the position to (mouseX, mouseY) to actually land on the mouse position.

thats why I also write use an epsilon here in the comment, its up to OP to implement an epsilon though.

- David
 14 Game Development / Shared Code / Re: some new blendmodes (add, multiply, overlay, etc..) on: 2011-07-24 12:08:32 Quote from: nsigma on 2011-07-24 11:23:32Quote from: Dx4 on 2011-07-23 17:30:39yea I was experimenting and happened to come up with that. I noticed an extremely different result - sometimes I used the color 0xffffaa00 with an opacity of 50% and after stroking a couple times, it became red (0xffff0000) which was obviously a major bug for a drawing application.After switching to this method for mult, it stayed dark yellow instead of changing to red after a certain number of strokes. - DavidI just had a quick test in Praxis as this would be a fairly serious bug for me too, but I can't reproduce this effect using my original code.  I assume you're using Normal / SrcOver blend mode for this?  I only see this effect (just red) when the opacity goes below 2%, when rounding errors really start to kick in - no way around that while using an integer pipeline.  Could it be errors in the premultiply() -> unpremultiply() causing it to round down?yeah I am using SrcOver. it also happens when I use premultiplied colors (eg: not calling the premultiply and unpremultiply functions)switching to the mult function I posted above fixed it.
15  Java Game APIs & Engines / Java 2D / Re: Loading custom font security issues on: 2011-07-24 07:55:12
Thanks for the reply David. That sounds like a good way to do it.
I found another way too:

 1  2  3  4  5  6  7  8 `URL url = null;      try       {         url = new URL(getCodeBase(), "font.ttf");         InputStream in=url.openStream();         m_font = Font.createFont(Font.TRUETYPE_FONT,in).deriveFont(20f);      }       catch (Exception e) {}`

no worries, glad i could help.
16  Game Development / Newbie & Debugging Questions / Re: Character movement with Mouse on: 2011-07-24 07:40:32
first you need to find the angle

 1 `double angle = Math.atan2(mouseY-playerY,mouseX-playerX);`

then you need to build a vector to calculate the movement threshold

 1  2 `double movex = Math.cos(angle);double movey = Math.sin(angle);`

and now to move towards the point

 1  2  3  4  5 `x += movex;y += movey;if (x == mouseX && y == mouseY) { // you can use epsilon here stop();}`

hth
17  Game Development / Shared Code / Re: some new blendmodes (add, multiply, overlay, etc..) on: 2011-07-23 17:30:39
this seems to be a more accurate alternative to your mult():

 1  2  3  4 `   public static int mult(int a, int b) {      int t = a * b + 128;      return (t >> 8) + t >> 8;   }`

Interesting.  Is that yours or from a particular source?

Tried a few values at random and does seem to be slightly more accurate, though only 1 away on all that I tried - have you noticed any greater difference?

Personally, it's unlikely I'll use that fix as it's adding 2 extra operations every time it's called - which is going to add up considering this method may be called 8 or more times per pixel.  I'm not even sure about adding in the +1 fix we had in the previous thread.  But then our use cases are slightly different - I'm working with video so am wanting to process many full frames per second, whereas I can understand accuracy being a greater concern in your application.

yea I was experimenting and happened to come up with that. I noticed an extremely different result - sometimes I used the color 0xffffaa00 with an opacity of 50% and after stroking a couple times, it became red (0xffff0000) which was obviously a major bug for a drawing application.

After switching to this method for mult, it stayed dark yellow instead of changing to red after a certain number of strokes.

- David
18  Game Development / Shared Code / Re: some new blendmodes (add, multiply, overlay, etc..) on: 2011-07-23 13:43:44
this seems to be a more accurate alternative to your mult():

 1  2  3  4 `   public static int mult(int a, int b) {      int t = a * b + 128;      return (t >> 8) + t >> 8;   }`
19  Java Game APIs & Engines / Java 2D / Re: Loading custom font security issues on: 2011-07-23 13:27:10
Hi, I have this code here that works fine offline, but fails online because of security IO issues.
 1 `font = Font.createFont(Font.TRUETYPE_FONT,new File("font.ttf")).deriveFont(20f);`

I think the problem is that it is trying to load the font from the user's file system instead of from the server. How can I fix this?
thanks

try putting the font in the jar then using getClass().getClassLoader().getResourceAsStream( ... ) and passing the stream into createFont instead. the accesscontrolcontext created by createFont with an input stream should have temp folder write permission.

HTH

- David
 20 Game Development / Shared Code / Re: some new blendmodes (add, multiply, overlay, etc..) on: 2011-07-23 12:03:35 Quote from: nsigma on 2011-07-23 11:49:02Quote from: Dx4 on 2011-07-23 11:29:24I pass in source pixels as single values because it seems easier that way, I don't have to write a massive loop around every single function which reads each source pixel, etc.and if you need to use a source array, it should be easy to add that anyhow. also there are no virtual calls in this class, which removes the overhead of a vtable (hotspot should inline the calls).OK. Makes sense. I'm just used to pushing the loop into the composite method because I needed RGBComposite to be an abstract class. I'll port some of your optimisations and other blend modes back to mine, so if people want to work with arrays then that's already there - I'm used to passing in scanlines or whole regions at a time.go ahead. I just wanted it to modify one pixel at a time because that seemed the most flexible to me. For example, in my project I have to construct a source pixel when blending (for example when mixing brush alpha with brush color: int source = alpha<<24|(color&0x00ffffff) ). I don't want to have to make an array each time I just want to modify a single pixel in a specific location.
22  Game Development / Shared Code / some new blendmodes (add, multiply, overlay, etc..) on: 2011-07-23 09:32:35
this class allows you to add new blendmodes to java. included are:

 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17 `   public static enum Mode {      Normal,      Add,      Multiply,      Erase,      Average,      Overlay,      Screen,      Glow,      Reflect,      Red,      Green,      Blue,      ColorDodge,      ColorBurn,a      HardLight,   }`

supports both premultiplied and unpremultiplied colors.

 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  135  136  137  138  139  140  141  142  143  144  145  146  147  148  149  150  151  152  153  154  155  156  157  158  159  160  161  162  163  164  165  166  167  168  169  170  171  172  173  174  175  176  177  178  179  180  181  182  183  184  185  186  187  188  189  190  191  192  193  194  195  196  197  198  199  200  201  202  203  204  205  206  207  208  209  210  211  212  213  214  215  216  217  218  219  220  221  222  223  224  225  226  227  228  229  230  231  232  233  234  235  236  237  238  239  240  241  242  243  244  245  246  247  248  249  250  251  252  253  254  255  256  257  258  259  260  261  262  263  264  265  266  267  268  269  270  271  272  273  274  275  276  277  278  279  280  281  282  283  284  285  286  287  288  289  290  291  292  293  294  295  296  297  298  299  300  301  302  303  304  305  306  307  308  309  310  311  312  313  314  315  316  317  318  319  320  321  322  323  324  325  326  327  328  329  330  331  332  333  334  335  336  337  338  339  340  341  342  343  344  345  346  347  348  349  350  351  352  353  354  355  356  357  358  359  360  361  362  363  364  365  366  367  368  369  370  371  372  373  374  375  376  377  378  379  380  381  382  383  384  385  386  387  388  389  390  391  392  393  394  395  396  397  398  399  400  401  402  403  404  405  406  407  408  409  410  411  412  413  414  415  416  417  418  419  420  421  422  423  424  425  426  427  428  429  430  431  432  433  434  435  436  437  438  439  440  441  442  443  444  445  446  447  448  449  450  451  452  453  454  455  456  457  458  459  460  461  462  463  464  465  466  467  468  469  470  471  472  473  474  475  476  477  478  479  480  481  482  483  484  485  486  487  488  489  490  491  492  493  494  495  496  497  498  499  500  501  502  503  504  505  506  507  508  509  510  511  512  513  514  515  516  517  518  519  520  521  522  523  524  525  526  527  528  529  530  531  532  533  534  535  536  537  538  539  540  541  542  543  544  545  546  547  548  549  550  551  552  553  554  555  556  557  558  559  560  561  562  563  564  565  566  567  568  569  570  571  572  573  574  575  576  577  578  579  580  581  582  583  584  585  586  587  588  589  590  591  592  593  594  595  596  597  598  599  600  601  602  603  604  605  606  607  608  609  610  611  612  613  614  615  616  617  618  619  620  621  622  623  624  625  626  627  628  629  630  631  632  633  634  635  636  637  638 `/* * To change this template, choose Tools | Templates * and open the template in the editor. */package as.blend;/** * * @author David */public class Compositor {   public static final int ALPHA_MASK = 0xff000000;   public static final int RED_MASK = 0x00ff0000;   public static final int GREEN_MASK = 0x0000ff00;   public static final int BLUE_MASK = 0x000000ff;   public static enum Mode {      Normal,      Add,      Multiply,      Erase,      Average,      Overlay,      Screen,      Glow,      Reflect,      Red,      Green,      Blue,      ColorDodge,      ColorBurn,      HardLight,   }   public static int premultiply(int argb) {      int a = argb >>> 24;      if (a == 0) {         return 0;      } else if (a == 255) {         return argb;      } else {         return (a << 24) | multRGB(argb, a);      }   }   public static int multRGB(int src, int multiplier) {      multiplier++;      return ((src & RED_MASK) * multiplier) >> 8 & RED_MASK            | ((src & GREEN_MASK) * multiplier) >> 8 & GREEN_MASK            | ((src & BLUE_MASK) * multiplier) >> 8;   }   public static void blendPixelPremultiplied(int [] pixels, int offset, int srcPx, int alpha, Mode mode) {      switch (mode) {         case Normal:            blendPixel_SrcOver(pixels, offset, srcPx, alpha); return;         case Add:            blendPixel_Add(pixels, offset, srcPx, alpha); return;         case Multiply:            blendPixel_Multiply(pixels, offset, srcPx, alpha); return;         case Erase:            blendPixel_DstOut(pixels, offset, srcPx, alpha); return;         case Average:            blendPixel_Average(pixels, offset, srcPx, alpha); return;         case Overlay:            blendPixel_Overlay(pixels, offset, srcPx, alpha); return;         case Screen:            blendPixel_Screen(pixels, offset, srcPx, alpha); return;         case Glow:            blendPixel_Glow(pixels, offset, srcPx, alpha); return;         case Reflect:            blendPixel_Reflect(pixels, offset, srcPx, alpha); return;         case Red:            blendPixel_Red(pixels, offset, srcPx, alpha); return;         case Green:            blendPixel_Green(pixels, offset, srcPx, alpha); return;         case Blue:            blendPixel_Blue(pixels, offset, srcPx, alpha); return;         case ColorBurn:            blendPixel_ColorBurn(pixels, offset, srcPx, alpha); return;         case ColorDodge:            blendPixel_ColorDodge(pixels, offset, srcPx, alpha); return;         case HardLight:            blendPixel_HardLight(pixels, offset, srcPx, alpha); return;      }      throw new BlendModeNotAvailableException(mode);   }   public static class BlendModeNotAvailableException extends RuntimeException {      public BlendModeNotAvailableException(Mode mode) {         super(mode + " is not available");      }   }   /**    * blends two non premultiplied colors together    * @param pixels    * @param offset    * @param srcPx    * @param alpha    * @param mode    */   public static void blendPixelUnpremultiplied(int[] pixels, int offset, int srcPx, int alpha, Mode mode) {      pixels[offset] = premultiplyEx(pixels[offset]);      blendPixelPremultiplied(pixels, offset, premultiplyEx(srcPx), alpha, mode);      pixels[offset] = unpremultiply(pixels[offset]);   }   public static void blendPixel_SrcOver(int pixels[], int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      final int oneMinusSrcA = 0xff - srcA;      pixels[offset] =            (blendOneMinus(srcA, destA, oneMinusSrcA) << 24            | blendOneMinus(srcR, destR, oneMinusSrcA) << 16            | blendOneMinus(srcG, destG, oneMinusSrcA) << 8            | blendOneMinus(srcB, destB, oneMinusSrcA));   }   /**    * Blends pixel using DstOut algorithm    * @param pixels    * @param offset    * @param srcPx    * @param alpha    */   public static void blendPixel_DstOut(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      if (alpha != 255) {         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);        final int oneMinusAsAe = 0xff - mult(alpha, srcA);      pixels[offset] =            ( mult(destA, oneMinusAsAe) << 24            | mult(destR, oneMinusAsAe) << 16            | mult(destG, oneMinusAsAe) << 8            | mult(destB, oneMinusAsAe));   }   public static void blendPixel_Average(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            ( (srcA + destA) << 23            | (srcR + destR) << 15            | (srcG + destG) << 7            | (srcB + destB) >> 1);   }   public static void blendPixel_Multiply(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      final int oneMinusSrcA = 0xff - srcA,              oneMinusDstA = 0xff - destA;      pixels[offset] =            blend(srcA, destA, srcA) << 24 |            (mult(srcR, destR) + mult(destR, oneMinusSrcA) + mult(srcR, oneMinusDstA)) << 16 |            (mult(srcG, destG) + mult(destG, oneMinusSrcA) + mult(srcG, oneMinusDstA)) << 8 |            mult(srcB, destB) + mult(destB, oneMinusSrcA) + mult(srcB, oneMinusDstA);   }   public static void blendPixel_Add(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =             min(srcA + destA, 0xff) << 24 |            min(srcR + destR, 0xff) << 16 |            min(srcG + destG, 0xff) << 8 |            min(srcB + destB, 0xff);   }   public static void blendPixel_Overlay(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      /*      if a < 128 then        result := (a*b) SHR 7      else        result := 255 - ((255-a) * (255-b) SHR 7);      */      pixels[offset] =            min(255, srcA + destA) << 24 |            overlay(srcR, destR) << 16 |            overlay(srcG, destG) << 8 |            overlay(srcB, destB);   }   public static void blendPixel_Screen(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      /*        result := 255 - ((255-a) * (255-b) SHR 7);      */      pixels[offset] =            min(255, srcA + destA) << 24 |            screen(srcR, destR) << 16 |            screen(srcG, destG) << 8 |            screen(srcB, destB);   }   public static void blendPixel_Reflect(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            min(255, srcA + destA) << 24 |            reflect(srcR, destR) << 16 |            reflect(srcG, destG) << 8 |            reflect(srcB, destB);   }   public static void blendPixel_Glow(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] = // inverse reflect            min(255, srcA + destA) << 24 |            reflect(destR, srcR) << 16 |            reflect(destG, srcG) << 8 |            reflect(destB, srcB);   }   public static void blendPixel_ColorDodge(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            min(255, srcA + destA) << 24 |            colordodge(srcR, destR) << 16 |            colordodge(srcG, destG) << 8 |            colordodge(srcB, destB);   }   public static void blendPixel_ColorBurn(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            min(255, srcA + destA) << 24 |            colorburn(srcR, destR) << 16 |            colorburn(srcG, destG) << 8 |            colorburn(srcB, destB);   }   public static void blendPixel_Darken(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            min(255, srcA + destA) << 24 |            min(srcR, destR) << 16 |            min(srcG, destG) << 8 |            min(srcB, destB);   }   public static void blendPixel_Lighten(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            min(255, srcA + destA) << 24 |            max(srcR, destR) << 16 |            max(srcG, destG) << 8 |            max(srcB, destB);   }   public static void blendPixel_Red(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            min(255, srcA + destA) << 24 |            srcR << 16 |            destG << 8 |            destB;   }   public static void blendPixel_Green(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcG = (srcPx & GREEN_MASK) >>> 8;      if (alpha != 255) {         srcG = mult(srcG, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            min(255, srcA + destA) << 24 |            destR << 16 |            srcG << 8 |            destB;   }   public static void blendPixel_Blue(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      pixels[offset] =            min(255, srcA + destA) << 24 |            destR << 16 |            destG << 8 |            srcB;   }   public static void blendPixel_HardLight(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            min(0xff, srcA + destA) << 24 |            hardlight(srcR, destR) << 16 |            hardlight(srcG, destG) << 8 |            hardlight(srcB, destB);   }   public static void blendPixel_Difference(int [] pixels, int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] =            blend(srcA, destA, srcA) << 24 |               (srcR + destR - (2 * min(mult(srcR, destA), mult(destR, srcA)))) << 16 |               (srcG + destG - (2 * min(mult(srcG, destA), mult(destG, srcA)))) << 8 |               (srcB + destB - (2 * min(mult(srcB, destA), mult(destB, srcA))));   }   public static int hardlight(int a, int b) {      return a < 128 ? (a * b) >> 7 : 0xff - ((0xff-a) * (0xff-b) >> 7);   }   public static int reflect(int a, int b) {      return b == 255 ? 255 : min(255, a * a / (255 - b));   }      public static int colorburn(int a, int b) {      return b == 0 ? 0 : max(0, (0xff - ((0xff-a) << 8)/ b));   }   public static int colordodge(int a, int b) {      return b == 0xff ? 0xff : min(0xff, ((a<<8) / (0xff-b)));   }   public static int overlay(int a, int b) {      if (a < 128) {         return (a * b) >> 7;      } else {         return 0xff - (((0xff-a) * (0xff-b)) >> 7);      }   }   public static int screen(int a, int b) {      return 0xff - mult(a, b);   }   public static int min(int a, int b) {      return a < b ? a:b;   }   public static int max(int a, int b) {      return a > b ? a:b;   }   public static int blendOneMinus(int src, int dest, int oneMinus) {      return src + mult(dest, oneMinus);   }   public static int blend(int src, int dest, int alpha) {      return src + mult(dest, 0xFF - alpha);   }   public static int mult2(int val, int multiplier) {      return (val * (multiplier + 1)) >> 8;   }   public static int mult(int a, int b) {      final int t = a * b + 128;      return (t >> 8) + t >> 8;   }   public static int premultiplyEx(int rgb) {      return premultiply(rgb, rgb>>>24);   }   public static int premultiply(int rgbColor, int alpha) {      if (alpha <= 0) {         return 0;      } else if (alpha >= 255) {         return 0xff000000 | rgbColor;      } else {         int r = (rgbColor >> 16) & 0xff;         int g = (rgbColor >> 8) & 0xff;         int b = rgbColor & 0xff;         r = (alpha * r + 127) / 255;         g = (alpha * g + 127) / 255;         b = (alpha * b + 127) / 255;         return (alpha << 24) | (r << 16) | (g << 8) | b;      }   }   public static int unpremultiply(int preARGBColor) {      int a = preARGBColor >>> 24;      if (a == 0) {         return 0;      } else if (a == 255) {         return preARGBColor;      } else {         int r = (preARGBColor >> 16) & 0xff;         int g = (preARGBColor >> 8) & 0xff;         int b = preARGBColor & 0xff;         r = 255 * r / a;         g = 255 * g / a;         b = 255 * b / a;         return (a << 24) | (r << 16) | (g << 8) | b;      }   }}`

mult and blend functions taken from praxis (http://code.google.com/p/praxis/source/browse/ripl/src/net/neilcsmith/ripl/rgbmath/RGBMath.java)

reference: http://www.pegtop.net/delphi/articles/blendmodes/intro.htm

example:

 1  2  3  4  5 `int [] destination = .. // your destination arrayint source = 0x80ff0000; // red with 50% opacity, can be any colorint offset = 0x20; // offset in destination array to blendint extraAlpha = 100; // extra alpha to blend withblendPixelPremultiplied(destination, offset, source, extraAlpha, Mode.Multiply);`

the above code blends the pixel in offset 0x20 with the color red, with 50% opacity, adding 100 extra alpha while blending, using the multiply blendmode.
 23 Game Development / Newbie & Debugging Questions / Re: Storing player data on: 2011-07-21 12:52:40 Quote from: Mickelukas on 2011-07-21 12:02:06Nice post Dx4 but it'll still be possible to backwards engineer and change the save file.obviously, people who want to hack your game and are dedicated enough WILL be able to, it's impossible to protect it from everyone, but with what I wrote above when mixed with some fancy obfuscation will prove to be a very burdensome task to reverse engineer and change.- David
24  Game Development / Newbie & Debugging Questions / Re: Storing player data on: 2011-07-21 09:55:54
Im working on a rpg game and i have a problem, storing player data.
I was thinking about storing save games in plain text files within the jar or in directories placed near the jar file, but that would be really easy for the player to modify, which will make the game less challenging and less fun. I have been thinking about serelization as well, but i don't know anything about serelization and it seems awfully complicated. Is there any other alternatives?

Sign the jar, and save the game state as a class using ObjectWeb ASM (or some other class modifying library). Replace the game state class file within the JAR, and generate a new checksum for the file. Save the checksum into MANIFEST.MF. When they try to modify the saved game state (which I doubt many could), save it and then try to run it again, the JVM will crash as the checksum for the class doesn't match. For extra security, check manifest.mf when the game starts and do a file integrity check on each file in the JAR file against the entries in the manifest. This will also ensure that people can't just remove the signing on the jar.

Example class:

 1  2  3  4 `public class State { private static int numLives = 100; private static int health = 100; .. etc}`

then use a class modifier to rewrite the values in the state class.

Alternatively, just use XML or OOS to save the state, and use JCE (DES, AES256, etc) to encrypt the data using a key generated with some special data, eg last modified time of the executing jar, OS type, etc

for example:

 1 `byte[] key = md5(System.getProperty("user.name") + System.getProperty("os.arch"));`
25  Game Development / Newbie & Debugging Questions / Re: Alpha blending on: 2011-07-20 19:11:49
That project looks pretty sweet!  Though MySQL seems to be down on your server at the moment.

Your experience on switching from Java2D is similar to mine - huge speed up - lots of things become faster (or even possible!)

it is down. subscription on my dedicated server expired i have yet to renew it

but you can still view the gallery - my artist drew those images in UxPaint: http://uxpaint.com/gallery

This is the most i've optimized the algorithm so far:

 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22 `   private void blendPixelEXT(int pixels[], int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      int srcA = srcPx >>> 24;      int srcR = (srcPx & RED_MASK) >>> 16;      int srcG = (srcPx & GREEN_MASK) >>> 8;      int srcB = (srcPx & BLUE_MASK);      if (alpha != 255) {         srcR = mult(srcR, alpha);         srcG = mult(srcG, alpha);         srcB = mult(srcB, alpha);         srcA = mult(srcA, alpha);      }      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] = (blend(srcA, destA, srcA) << 24            | blend(srcR, destR, srcA) << 16            | blend(srcG, destG, srcA) << 8            | blend(srcB, destB, srcA));   }`
 26 Discussions / General Discussions / Re: UxPaint - new collaborative paint program in development on: 2011-07-20 19:04:26 Update:Added square brush, and switched the entire rendering system to software instead of Java2D. - David
27  Game Development / Newbie & Debugging Questions / Re: Alpha blending on: 2011-07-20 18:55:23

Incidentally, what are you writing a software renderer for?  A project of its own or something bigger?

And I know I mentioned it before, but I'd really recommend moving to premultiplied data throughout - it's easier and better performing - trust me, I found out the hard way!

I'm writing a software renderer to handle the brushes for my project:

http://www.java-gaming.org/topics/uxpaint-new-collaborative-paint-program-in-development/24478/view.html

I wasn't happy with the performance of Java2D, especially in regards to colorizing images on the fly; therefore I had to devise something much faster.

Now that I have switched to a software renderer, I get the following benefits:

• Able to use any blendmode I want for brushes (Add, Multiply, Over, Overlay, Screen, Darken, etc)
• Able to save the state of any pixels I modify without doing a scan beforehand (VERY useful for undo states)
• Able to colorize brush pixels on the fly using a simple OR operation: (source << 24) | color, previously in Java2D I had to create a new image, get a graphics context, set alphacomposite, set color and call drawLine(x,y,x,y) for each pixel I wanted to recolor. INSANE speed increase in brush rendering once I made the switch.
• Uses much less memory, each brush image (with colors, etc) took about 4kb memory before, and about 1000 would be created for a single stroke, equal to 4MB. With my new software renderer, it is able to use a constant 400kb per brush (for 100 sizes)

overall, the software renderer is about 15x faster than Java2D and also supports blendmodes, state saving, etc, its a win-win situation.

I might switch to TYPE_ARGB_PRE later, if required, but I've actually tried it before and it results in weird artifacts when I zoom in, i'll save it for investigation on another day

BTW: Brush class is now:

 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  135  136  137  138  139  140  141  142  143  144  145  146  147  148  149  150  151  152  153  154  155  156  157  158  159  160  161  162  163  164  165  166  167  168  169  170  171  172  173  174  175  176  177  178  179  180  181  182  183  184  185  186  187  188  189  190  191  192  193  194  195  196  197  198  199  200  201  202  203  204  205  206  207  208  209  210  211  212  213  214  215  216  217  218  219  220  221  222  223  224  225  226  227  228  229  230  231  232  233  234  235  236  237  238  239  240  241  242  243  244  245  246  247  248  249  250  251  252  253  254  255  256  257  258  259  260  261  262  263  264  265  266  267  268  269  270  271  272  273  274  275  276  277  278  279  280  281 `package as.internal.unnatural;import as.internal.UndoManager.UndoState;import java.awt.AlphaComposite;import java.awt.Color;import java.awt.Graphics2D;import java.awt.RenderingHints;import java.awt.image.BufferedImage;import java.awt.image.DataBufferInt;import java.util.Arrays;/** * * @author David */public class Brush {   private int[][] brushData;   private int maxPixelValue;   public static int TypeRound = 1,         TypeSquare = 2;   public static Brush getSquareBrush(int count) {      return new Brush(count, TypeSquare, false);   }   public static Brush getRoundBrush(int count) {      return new Brush(count, TypeRound, true);   }   private Brush(int numSizes, int type, boolean antialias) {      brushData = new int[numSizes][];      for (int i = 1; i <= numSizes; i++) {         BufferedImage brush = new BufferedImage(i, i, BufferedImage.TYPE_INT_ARGB);         Graphics2D g = brush.createGraphics();         if (antialias) {            g.setRenderingHint(RenderingHints.KEY_INTERPOLATION, RenderingHints.VALUE_INTERPOLATION_BICUBIC);            g.setRenderingHint(RenderingHints.KEY_ANTIALIASING, RenderingHints.VALUE_ANTIALIAS_ON);         }         g.setColor(Color.BLACK);         g.setComposite(AlphaComposite.Src);         if (type == TypeRound) {            g.fillOval(0, 0, i, i);         } else if (type == TypeSquare) {            g.fillRect(0, 0, i, i);         }         g.dispose();         int array[] = ((DataBufferInt) brush.getRaster().getDataBuffer()).getData();         brushData[i - 1] = Arrays.copyOf(array, array.length);         int[] data = brushData[i - 1];         for (int j = 0; j < data.length; j++) {            data[j] >>>= 24;            if (data[j] > maxPixelValue) {               maxPixelValue = data[j];            }         }         brush.flush();      }   }   public Brush(BufferedImage image, int numSizes) {      brushData = new int[numSizes][];      for (int i = 1; i <= numSizes; i++) {         BufferedImage brush = new BufferedImage(i, i, BufferedImage.TYPE_INT_ARGB);         Graphics2D g = brush.createGraphics();         g.setRenderingHint(RenderingHints.KEY_INTERPOLATION, RenderingHints.VALUE_INTERPOLATION_BICUBIC);         g.drawImage(image, 0, 0, i, i, null);         g.dispose();         int array[] = ((DataBufferInt) brush.getRaster().getDataBuffer()).getData();         brushData[i - 1] = Arrays.copyOf(array, array.length);         int[] data = brushData[i - 1];         for (int j = 0; j < data.length; j++) {            data[j] >>>= 24;            if (data[j] > maxPixelValue) {               maxPixelValue = data[j];            }         }         brush.flush();      }   }   public int getPixelMaxAlpha() {      return maxPixelValue;   }   public int getMaxBrushSize() {      return brushData.length + 1;   }   int clamp(int value) {      return value > 255 ? 255 : value;   }   public static int premultiply(int rgbColor, int alpha) {      if (alpha <= 0) {         return 0;      } else if (alpha >= 255) {         return 0xff000000 | rgbColor;      } else {         int r = (rgbColor >> 16) & 0xff;         int g = (rgbColor >> 8) & 0xff;         int b = rgbColor & 0xff;         r = (alpha * r + 127) / 255;         g = (alpha * g + 127) / 255;         b = (alpha * b + 127) / 255;         return (alpha << 24) | (r << 16) | (g << 8) | b;      }   }   public static int unpremultiply(int preARGBColor) {      int a = preARGBColor >>> 24;      if (a == 0) {         return 0;      } else if (a == 255) {         return preARGBColor;      } else {         int r = (preARGBColor >> 16) & 0xff;         int g = (preARGBColor >> 8) & 0xff;         int b = preARGBColor & 0xff;         r = 255 * r / a;         g = 255 * g / a;         b = 255 * b / a;         return (a << 24) | (r << 16) | (g << 8) | b;      }   }   void alphaBlend(int[] pixels, int offset, int source, int alpha) {      int destRGB = pixels[offset];      int destA = destRGB >>> 24;      //destRGB = premultiply(destRGB, destA);      int destR = (destRGB >> 16) & 0xff;      int destG = (destRGB >> 8) & 0xff;      int destB = destRGB & 0xff;      int srcA = source >>> 24;      //source = premultiply(source, srcA);      int srcR = (source >> 16) & 0xff;      int srcG = (source >> 8) & 0xff;      int srcB = source & 0xff;      srcA *= alpha;      srcR *= alpha;      srcG *= alpha;      srcB *= alpha;      int oneMinusSrcA = 0xff - (srcA >> 8);      destR = (srcR + destR * oneMinusSrcA) >> 8;      destG = (srcG + destG * oneMinusSrcA) >> 8;      destB = (srcB + destB * oneMinusSrcA) >> 8;      destA = (srcA + destA * oneMinusSrcA) >> 8;      //pixels[offset] = (destA << 24) | (destR << 16) | (destG << 8) | destB;      pixels[offset] = //unpremultiply            ((destA << 24) | (destR << 16) | (destG << 8) | destB);   }           /*thanks nsigma*/   public static final int ALPHA_MASK = 0xff000000;   public static final int RED_MASK = 0x00ff0000;   public static final int GREEN_MASK = 0x0000ff00;   public static final int BLUE_MASK = 0x000000ff;   public static int premultiply(int argb) {      int a = argb >>> 24;      if (a == 0) {         return 0;      } else if (a == 255) {         return argb;      } else {         return (a << 24) | multRGB(argb, a);      }   }   public static int premultiplyEXT(int argb, int a) {      if (a == 0) {         return 0;      } else if (a == 255) {         return argb;      } else {         return (a << 24) | multRGB(argb, a);      }   }   public static int multRGB(int src, int multiplier) {      multiplier++;      return ((src & RED_MASK) * multiplier) >> 8 & RED_MASK            | ((src & GREEN_MASK) * multiplier) >> 8 & GREEN_MASK            | ((src & BLUE_MASK) * multiplier) >> 8;   }   private void blendPixelEXT(int pixels[], int offset, int srcPx, int alpha) {      boolean alpha255 = alpha == 255;      int destPx = pixels[offset];      destPx = premultiplyEXT(destPx, (destPx) >>> 24);      int srcA = srcPx >>> 24;      srcPx = premultiplyEXT(srcPx, srcA);      if (alpha255 == false) {         srcA = mult(srcA, alpha);      }      int srcR = (alpha255) ? (srcPx & RED_MASK) >>> 16            : mult((srcPx & RED_MASK) >>> 16, alpha);      int srcG = (alpha255) ? (srcPx & GREEN_MASK) >>> 8            : mult((srcPx & GREEN_MASK) >>> 8, alpha);      int srcB = (alpha255) ? (srcPx & BLUE_MASK)            : mult(srcPx & BLUE_MASK, alpha);      int destA = (destPx) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] = unpremultiply(blend(srcA, destA, srcA) << 24            | blend(srcR, destR, srcA) << 16            | blend(srcG, destG, srcA) << 8            | blend(srcB, destB, srcA));   }   public static int blend(int src, int dest, int alpha) {      return src + mult(dest, 0xFF - alpha);   }   public static int mult(int val, int multiplier) {      return (val * (multiplier + 1)) >> 8;   }   public void drawBrush(int x, int y, int size, int alpha, int color, int[] pixels, int width, int height, UndoState state) {      if (alpha == 0) {         return;      }      int brushIndex = 0;      int[] brushPixels = brushData[size - 1];      for (int i = 0; i < size; i++) {         final int ycoord = y + i;         if (ycoord < 0 || ycoord >= height) {            continue;         }         for (int j = 0; j < size; j++) {            int source = (brushPixels[brushIndex++]);            if (source > 0) {               final int xcoord = x + j;               if (xcoord < 0 || xcoord >= width) {                  continue;               }               final int pos = ycoord * width + xcoord;               state.putPixel(pixels[pos], xcoord, ycoord);               blendPixelEXT(pixels, pos, (source << 24) | color, alpha);            }         }      }   }   public void drawBrush(int x, int y, int size, int alpha, int color, UserLayer layer) {      if (size <= 0) {         return;      }      int brushIndex = 0;      int[] brushPixels = brushData[size - 1];      int limit = (alpha * maxPixelValue) / 255;      for (int i = 0; i < size; i++) {         for (int j = 0; j < size; j++) {            int source = (brushPixels[brushIndex++]);            if (source > 0) {               final int xcoord = x + j, ycoord = y + i;               if (xcoord < 0 || ycoord < 0 || xcoord >= layer.width || ycoord >= layer.height) {                  continue;               }               final int xcoordScaled = xcoord >> 5, ycoordScaled = ycoord >> 5;               UserLayer.Tile tile = layer.getTileForNoScaling(xcoordScaled, ycoordScaled);               int offset = ((ycoord - (ycoordScaled << 5)) << 5) + xcoord - (xcoordScaled << 5);               int destA = tile.pixels[offset] >>> 24;               destA = source + ((destA * (0xff - source)) >> 8);               destA = Math.min(limit, destA);               tile.pixels[offset] = color | (destA << 24);            }         }      }   }}`

just in case anyone else finds it useful..

David
28  Game Development / Newbie & Debugging Questions / Re: Alpha blending on: 2011-07-20 14:53:24
 1  2  3  4  5  6  7 `    public static int blend(int src, int dest, int alpha) {        return src + (((0xFF - alpha) * dest) >> 8);    }    public static int mult(int val, int multiplier) {        return (val * (multiplier + 1)) >> 8;    }`

I'm thinking of just capping the alpha so the resultant alpha can NEVER be less than the destination alpha.. seems like the easiest way out.

This is probably still rounding errors caused by bitshifting.  Could try -

 1  2  3  4  5  6  7 `    public static int blend(int src, int dest, int alpha) {        return src + mult(dest, 0xFF - alpha);    }    public static int mult(int val, int multiplier) {        return (val * (multiplier + 1)) >> 8;    }`

ie. force it to add 1 to the multiplier.

Haven't been able to test it, but I think it might work.  Should probably evaluate that fix for my code too.

Best wishes, Neil

Thanks so much Neil! That did the trick

BTW: Slight optimization:

 1 `int destA = (destPx & ALPHA_MASK) >>> 24;`

use

 1 `int destA = (destPx) >>> 24;`

no need to mask it when you know that the left 24 bits will always be zero, same thing applies to srcA.
29  Game Development / Newbie & Debugging Questions / Re: Alpha blending on: 2011-07-20 09:03:58

I've also written my own software renderer.  Just had my first morning coffee so it's a bit early for bitshifting - let's hope this is right ...

 1  2  3  4 `        srcA *= alpha;        srcR *= alpha;        srcG *= alpha;        srcB *= alpha;`

This bit looks wrong.  Assuming your alpha value is in the range 0..255, then that needs to be shifted down.

This should be right, as I'm doing a >> 8 shift after, to move it back into the range of a byte.

eg.
 1 `  srcA = (srcA * alpha) >> 8;`

In my code I'm also doing the equivalent of
 1 `  srcA = (srcA * (alpha + 1)) >> 8;`

which seems to be more accurate.

I will try this, thanks for spotting that.

The second issue may be that Porter-Duff algorithms expect pre-multiplied colour data.  You haven't mentioned whether the data is pre-multiplied or not (I'd recommend doing so, as it makes a lot of things easier for both you and the CPU).

If your data isn't pre-multiplied then I'd also change the code above to be

 1  2  3  4 `        srcA = alpha == 255 ? srcA : (srcA * (alpha + 1)) >> 8;        srcR = (srcR * (srcA + 1)) >> 8;        srcG = (srcG * (srcA + 1)) >> 8;        srcB = (srcB * (srcA + 1)) >> 8;`

ie. make sure to multiply colour values by srcA not just alpha.

Hopefully that's all correct - I haven't actually tried it!

This looks interesting. I will take a look into it.

If you're interested in looking at my blending mode code in Praxis, have a look here http://code.google.com/p/praxis/source/browse/ripl/src/net/neilcsmith/ripl/rgbmath/RGBComposite.java.  SrcOver is actually called Normal.  Praxis as a whole is GPL3, but feel free to use anything in this file, or the RGBMath file (which it also needs).

I will definitely take a look at your Praxis project. It looks quite interesting.

Hope that helps.

Best wishes,  Neil

EDIT:

it still results in pixels with less alpha than expected for some reason, when low alpha values are used, try 1

this is the code I used:

 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 `    public static final int ALPHA_MASK = 0xff000000;    public static final int RED_MASK = 0x00ff0000;    public static final int GREEN_MASK = 0x0000ff00;    public static final int BLUE_MASK = 0x000000ff;     public static int premultiply(int argb) {        int a = argb >>> 24;        if (a == 0) {            return 0;        }        else if (a == 255) {            return argb;        }        else {            return (a << 24) | multRGB(argb, a);        }    }   public static int multRGB(int src, int multiplier) {        multiplier++;        return ((src & RED_MASK) * multiplier) >> 8 & RED_MASK |                ((src & GREEN_MASK) * multiplier) >> 8 & GREEN_MASK |                ((src & BLUE_MASK) * multiplier) >> 8;    }   private void blendPixelEXT(int pixels[], int offset, int srcPx, int alpha) {      int destPx = pixels[offset];      destPx = premultiply(destPx);      srcPx = premultiply(srcPx);      int srcA = (alpha == 255) ? (srcPx & ALPHA_MASK) >>> 24            : mult((srcPx & ALPHA_MASK) >>> 24, alpha);      int srcR = (alpha == 255) ? (srcPx & RED_MASK) >>> 16            : mult((srcPx & RED_MASK) >>> 16, alpha);      int srcG = (alpha == 255) ? (srcPx & GREEN_MASK) >>> 8            : mult((srcPx & GREEN_MASK) >>> 8, alpha);      int srcB = (alpha == 255) ? (srcPx & BLUE_MASK)            : mult(srcPx & BLUE_MASK, alpha);      int destA = (destPx & ALPHA_MASK) >>> 24;      int destR = (destPx & RED_MASK) >>> 16;      int destG = (destPx & GREEN_MASK) >>> 8;      int destB = (destPx & BLUE_MASK);      pixels[offset] = unpremultiply(blend(srcA, destA, srcA) << 24 |            blend(srcR, destR, srcA) << 16 |            blend(srcG, destG, srcA) << 8 |            blend(srcB, destB, srcA));   }    public static int blend(int src, int dest, int alpha) {        return src + (((0xFF - alpha) * dest) >> 8);    }    public static int mult(int val, int multiplier) {        return (val * (multiplier + 1)) >> 8;    }`

I'm thinking of just capping the alpha so the resultant alpha can NEVER be less than the destination alpha.. seems like the easiest way out.
30  Game Development / Newbie & Debugging Questions / Alpha blending on: 2011-07-20 05:39:31
ive been writing a software renderer, but it seems like alpha blending isnt working so well..

algorithm for porter-duff SrcOver:

 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 `   void alphaBlend(int [] pixels, int offset, int source, int alpha) {        int destRGB = pixels[offset];      int destA = destRGB >>> 24;        int destR = (destRGB >> 16) & 0xff;        int destG = (destRGB >> 8) & 0xff;        int destB = destRGB & 0xff;        int srcA = source >>> 24;        int srcR = (source >> 16) & 0xff;        int srcG = (source >> 8) & 0xff;        int srcB = source & 0xff;        srcA *= alpha;        srcR *= alpha;        srcG *= alpha;        srcB *= alpha;        int oneMinusSrcA = 0xff - (srcA >> 8);        destR = (srcR + destR * oneMinusSrcA) >> 8;        destG = (srcG + destG * oneMinusSrcA) >> 8;        destB = (srcB + destB * oneMinusSrcA) >> 8;      destA = (srcA + destA * oneMinusSrcA) >> 8;        //pixels[offset] = (destA << 24) | (destR << 16) | (destG << 8) | destB;      pixels[offset] =            ((destA << 24) | (destR << 16) | (destG << 8) | destB);   }`

pixel = array of ARGB pixels
offset = position of pixel in array to blend
source = color in ARGB format to blend with destination (pixel[offset])
alpha = alpha to take into account while blending (between 0-255)

the problem is that sometimes it doesn't blend the alpha properly and instead just replaces the destination alpha with source alpha, can anyone see why this happens?

it happens when the source color has a very small alpha value eg: 1

thanks
 Pages: [1] 2 3
 CopyableCougar4 (55 views) 2016-06-25 16:56:52 Hydroque (89 views) 2016-06-22 02:17:53 SwampChicken (90 views) 2016-06-20 13:22:57 SwampChicken (92 views) 2016-06-20 13:22:49 SwampChicken (85 views) 2016-06-20 13:22:26 Hydroque (129 views) 2016-06-15 08:22:50 Hydroque (123 views) 2016-06-13 06:40:55 DarkCart (240 views) 2016-05-29 02:30:33 Hydroque (206 views) 2016-05-26 14:45:46 Mac70 (188 views) 2016-05-24 21:16:33
 Spasi 49x theagentd 44x NegativeZero 39x KaiHH 30x LiquidNitrogen 23x princec 19x CoDi^R 16x ra4king 13x ags1 12x Opiop 10x kingroka123 9x purenickery 9x TritonDreyja 8x Archive 8x SHC 8x DavidBVal 8x
 Making a Dynamic Plugin Systemby Hydroque2016-06-25 00:13:25Java Data structures2016-06-13 21:22:09Java Data structures2016-06-13 21:20:42FPS Camera Tutorialby Hydroque2016-05-22 05:40:58Website offering 3D Models specifically for games for freeby vusman2016-05-18 17:23:09Website offering 3D Models specifically for games for freeby vusman2016-05-09 08:50:56Website offering 3D Models specifically for games for freeby vusman2016-05-06 11:10:21Website offering 3D Models specifically for games for freeby vusman2016-04-29 12:56:17
 java-gaming.org is not responsible for the content posted by its members, including references to external websites, and other references that may or may not have a relation with our primarily gaming and game production oriented community. inquiries and complaints can be sent via email to the info‑account of the company managing the website of java‑gaming.org