Jorge Cardoso's Blog

The  Public Sound Objects project installed at the Casa da Música will appear on the next issue of Exame Informática. 

In the installation at Casa da Música the following people participated:

Álvaro Barbosa - The project's author.

Jorge Cardoso - The guy (me) that implemented it all in Java.

Jorge Abade - The guy that designed those nice looking orange stands for the Mac computers.

Álvaro Barbosa explains how it works in this video:

This is the chart comparing the java.util.logging mechanism with plain simple System.out.println to log messages in a Java program.

In each test, N (1, 5 and 10) threads were launched. Each thread produces 100000 log messages in a loop and records the total time. The output of the program was redirected to a file (both standard output and standard error).

The chart shows the average time each thread took to log the 100000 messages.

The thing to note is that java.util.logging scales a lot worse than plain System.out.println calls. And things get worse if you're using a FileHandler on the Logger...

This might not be a problem in normal programs, but in my simulation case, it was since I was launching several instances on the same machine.  At some point, logging was a bottleneck...

Inserido na iniciativa Cursos de Verão na Católica, irei leccionar um "workshop" sobre interacção através de vídeo em Processing:

Este "workshop" tem como objectivo apresentar algumas técnicas que permitem criar interacção entre utilizador e computador usando  visão por computador. Usando a linguagem/ferramenta Processing, desenvolvida por professores do MIT, serão introduzidas as principais técnicas de isolamento de objectos numa sequência de frames de vídeo. Iremos ver como permitir que um utilizador interaja com elementos virtuais usando apenas oo movimento do seu corpo.

Podem ver no final desta página vídeos de alguns exemplos que iremos desenvolver.

Inserido na iniciativa Cursos de Verão na Católica, irei leccionar um "workshop" sobre Processing:

Este "workshop" tem como objectivo fornecer conhecimentos introdutórios à programação de computadores. Usando a linguagem/ferramenta Processing, desenvolvida por professores do MIT, serão introduzidos os conceitos básicos de programação ao mesmo tempo que se constroem programas com uma forte componente visual: "screensavers". Os conceitos abordados permitirão a subsequente aprendizagem autónoma de conceitos mais complexos e a adaptação a outras linguagens de programação como PHP, Actionscript de Flash, Java, etc.

Podem ver no final desta página vídeos de alguns exemplos de screensavers que iremos desenvolver.

I couldn't resist to play a bit more with the Face Detection library for Processing... see my previous post.

 This time the eyes follow the mouse

After seeing Bryan Chung's face detection Processing library for windows I decided to have a go with it.

First, I had trouble running the original example because of this error (see these comments on Bryan Chung's blog):

Exception in thread “Thread-2″ java.lang.UnsatisfiedLinkError: D:\processing-0135-expert\libraries\pFaceDetect\library\PFaceDetect.dll: This application has failed to start because the application configuration is incorrect. Reinstalling the application may fix this problem
at java.lang.ClassLoader$NativeLibrary.load(Native Method)
at java.lang.ClassLoader.loadLibrary0(Unknown Source)
at java.lang.ClassLoader.loadLibrary(Unknown Source)
at java.lang.Runtime.loadLibrary0(Unknown Source)
at java.lang.System.loadLibrary(Unknown Source)
at pFaceDetect.PFaceDetect.(PFaceDetect.java:10)

I ended up having to recompile the C++ code and replacing the original dll with my own, but Bryan's tested on several computers and he had no problems, so this was probably something wrong with my setup. Here's my pFaceDetect.dll, if yours doesn't work either.

I adapted slightly the original example so that it would track a single face (but detects all faces in the image) and replace it with a smiley.

In the video you can note some of the library's limitations (actually, OpenCV's haarcascade tracking limitations): sometimes it finds faces where there are none. Sometimes it doesn't find a face where there is one.

You can't see this in the video, but the maximum framerate I got in my laptop was 10 fps. This is low, but it doesn't stop you from doing something cool with this library, and there are probably ways to make it faster.

PS: I was chewing gum in the video...

I have just discovered that my old Freecom Media Center 35 can be used as a Network Attached Storage (NAS)!

I bought this a couple of years ago and I used it a standard external usb hard disk and also connected to a network to play media on a computer with a Windows Share.

But now (I don't know how long it has this functionality, the manual doesn't mention any of this) I found (while searching for a firmware upgrade) out it can also be used as NAS. All you need is the NDAS Software.

To get it working I had to upgrade the firmware to version 1.4.9 first. Then configure the media center with DHCP and NDAS on. Register the NDAS device ID and Write Key on the NDAS Device Management that I had just installed and that's it! I now have a new drive mapped.

You can even copy files while it is reproducing a video, but I don't recomend: it doesn't have enough resources to play the video smoothly while transfering files...

This code is still very messy, but I've decided to make public anyway..

This is the code for the last example. 

edit: You need Java 1.5 installed to run the examples. Install the JDK 1.5, or higher, and install Processing (without Java).