26 November 2010

Got my hands on some utilities that we can use to convert .dae files to .pod files. It can be found here:
http://www.imgtec.com/powervr/insider/powervr-collada2pod.asp

Managed to get some sample .dae files to test out the conversion, and this website provides some good models:

http://www.collada.org/owl/

Once I got the conversion done, I got it tested out with PVRShaman. This utility allows me to import the .pod files that I have and view it. I can even do things like rotation, panning, etc. Changing display to wireframe can be done too. So with that, I'm actually able to see that the .dae file is converted to .pod nicely. Here's the utility:
http://www.imgtec.com/powervr/insider/powervr-pvrshaman.asp

Many 3D modeling tools out there are able to export the models into .dae files. Now the question is, am I able to use Blender to export the models into .pod files directly? If that can be done, we can easily skip the step for conversion. Did quite a lot of research, and realized that most people out there do the conversion using this method.

I then looked into the requirements for having animations in .pod files. Did some analysis on the .dae files that I have downloaded, and found out that the animations are actually inside the .pod file itself. After going through some forums, many users are saying that they create animations by capturing the frames. Once they got the animations up, they could just easily export the 3D model into a .dae file, which can then be converted to a .pod file. There, we would then have our animations.

25 November 2010

Carried on with correcting the textures implementation today. Still could not get things working properly.

Kevin then came in and brought up his concern on the different file formats that we can use to load our 3D models. Hence, we went on to do some research to find out which file format is best supported by the hardware that's being used by the iPad, which is the POWERVR chipset.

After lunch, we discussed on what we have researched on and finally decided to work with .pod files (aka POWERVR object data). Zac was tasked to look into loading the .pod files, while I looked into the utilities that are available to obtain .pod files, be it converting or exporting directly from 3D modeling tools.

Had a short feedback session with Kevin before leaving the office, and I received some really valuable feedback from him. Gonna keep his feedback in mind and work actively to improve myself. It's very seldom that we meet people who are willing to go out of their way and give you constructive feedback and let you know about yourself, be it the positive or the negative side. I'm gonna have to cherish this opportunity to improve.

"Never be afraid to question"

24 November 2010

Realized that my textures wasn't working very well. Had to do something to the texture indices and model indices as they're different, and these affect how the model is being drawn.

Spent the day trying to understand the sample codes that we have. The codes are kinda complex, hopefully I can get it working soon.

23 November 2010

Zac and I started work on lighting and textures today. We decided to split to work and Zac was to work on lighting, and I was to work on textures. But before we started on our work, Zac did some explanation on how he solved the translation issue. It really is the problem that we suspected during our small discussion on the train home yesterday.

Read up this link and I got a fairly clear picture of what to do. I find it really useful as the author explains every step clearly, thus, giving me a clearer picture of what is going on. Check it out here:

http://iphonedevelopment.blogspot.com/2009/05/opengl-es-from-ground-up-part-6_25.html

Got the textures up at the end of the day. Zac got lighting up too. However, there's still much improvement to be made on our work. Gonna work on that tomorrow.

22 November 2010

Spent the entire day working on the pan gesture for translation. Zac and I had a hard time trying to cater for 3D translation. Thus, we went back to understand how the frustum is being created again.

Although we have understood what was going on, we still could not come up with a proper solution. The center of the object does not follow our touch at all.

Later in the day, Kevin came in and brought up his solution. We tried it out, but there were still some minor problems. Gonna work out those solutions with Zac tomorrow.

19 November 2010

Touched up on the scaling of the 3D models in the morning. Now the object is scaling nicely. Kevin then came in and we discussed on what to do next. Here's a picture of what we have discussed:

We would be following the schedule stated on the top right corner of the board. Left the layout and gesture for the last as we have to wait for the iOS 4.2 SDK to be released, so that we can work on the iPad.

Lunched at Clementi with Kevin, Calvin and Zac. Went there as Kevin wants to buy the iPhone 3D programming book. We ate and then went to get Koi bubble tea. Hehe. It was a really funny experience. Shall not mention the details here! :P

After lunch, we came back and I added in the part on 3D model translations. I already had it done up previously, just had to port it over to this project. With this, we have our version 2.0 up and running. Spent some time solving memory leaks here and there.

*edit* translation was not working well. Had to tweak it here and there. Spent some time trying to understand what happens when the model moves in the negative z-axis. Somehow got the formula up, but still have to try it out on Monday to see if it works.

18 November 2010

Got quite a lot of things done today. Completed the port over for swipe gesture rotation to OpenGL ES 1.1 today. Not much changes had to be made, so it was fairly quick. Still, individual object interactions with multiple objects loaded was not catered for yet.

Moved on to make use of pinch gesture for scaling, and managed to get it up. Still got some tweaking to do tomorrow though. Got to watch the podcast lecture first.

With that done, Zac and I then moved on to research on how to interact with individual objects. We found 2 different approaches. One of which is to make use of radius, the other is to make use of colors. After weighing the pros and cons, we decided to go with colors first. Got things working out nicely at the end of the day :) we're able to do individual object rotations with multiple objects loaded onto the view.