Saturday, January 30, 2010

Animated Render and Turnaround of Tower

I had a few hours free last night to make an animated render of the latest model, the tower at the end of the pedestrian zone where the player will be inserted into the game level. Add a snowstorm this morning that makes the streets somewhat impassable, and you get a weekend with Blender, GIMP, YouTube, and the project blog. Anyway, here is the result of my latest endeavors:

Although clearly not at the artistic level of many 3D modelers out there, and probably violating numerous design pet peeves as described by blenderguru, I'm still impressed with what I've accomplished and realize that I still have a lot more to learn. I'd particularly be interested in learning how to make the corners of my models more "rounded" and real-life while keeping the model within the low-poly mesh range. The roof also looks much too reflective for a typical slate roof like you find in Germany, which I noticed only after doing a turnaround of the model and seeing the it catches the light of lamp at certain angles. Need to figure our how to correct this (probably with a specularity map, right?).

Instead of baking the textures and importing the 3D model into Unity 3D, at the moment I'm thinking that I will continue to develop models for the game level and continue with my research on 3D digital game-based learning for second language acquisition. It has been fun and also interesting to be involved in the design process while simultaneously reading primary and secondary literature on the topic. I'll be interested in seeing what come of this whole thing in the coming year.

Tuesday, January 26, 2010

Tower Finished!

After nearly two weeks of working on the pedestrian zone tower in my spare time, I am happy to say that I finally got it finished. And here is the final product:

Not bad, considering that I have no 3D art experience at all and am pretty much self-taught on Blender and GIMP, the two open source platforms I used to create the tower model. The source of inspiration for the tower is Das Obertor in Bad Camberg. Not an exact copy, but fairly close:

Most of the textures I grabbed off the Web, although for the sides of the tower and the passage I found some high-quality textures at Arroway Textures. I can't begin to tell you how happy I am with the textures I found there. Although I grabbed a few to test them out on this project, I'm most certainly going to purchase the some of the CDs they offer on their site. In the coming days, I will post a video of a turnaround render, a FRAPS of how the model was put together, and a few other things.

Monday, January 25, 2010

Fun With Blender Texture Channels

Had an interesting problem a few days ago with Blender's texture channels. I had developed two models separately, a sign for the pedestrian zone and the tower, and wanted to bring them together (a) to see how they looked and (b) to get my workflow ironed out a bit. I am planning to develop the models separately, stage them within Blender and join them, and then import them into Unity 3D. Should be easy enough, I thought, and I was quickly able to package the image resources for the sign, import it into the tower model, and scale it to size. According to my Blender interface, things were going well:

The texured viewport indicated that, indeed, the model for the sign was where it was supposed to be and looked the way I wanted. Imagine my frustration, then, when I got this render:

The problem, which I discovered after a long time of tinkering and swearing, was with my texture channels. To simplify things a bit, I had one channel called "colormap" and the other called "bumpmap" for all my meshes. In all my models. Although Blender can keep these channels separated in individual files, once these models are imported into other models, Blender gets the channels confused. Thus, my pipe colormap being applied to my sign colormaps. I thought that each file would maybe have a unique namespace, or something like this, to avoid such a problem, but apparently not. To avoid this problem in the future, I will create unique texture channels for all models.

Wednesday, January 20, 2010

FRAPS Test of Blender Interface

As I mentioned in my previous post, I just found the FRAPS shareware online and downloaded it for doing screen captures to support my book project. I also noted that the software supports games using OpenGL technology, which meant that - perhaps - it would also support Blender video captures as Blender is OpenGL API-based software. Previous attempts with Camtasia proved less than satisfactory at doing Blender video captures, so I thought I would give it a try. And here is what I got:

Although the shareware version only supports 30-second video captures, I was extremely impressed with the quality of the capture and the audio track. I did this particular capture at half-size 30 fps, although FRAPS allows a user to go up to 60 fps with a full-size capture. I felt a little pressed to explain everything regarding the tower and how I see it fitting into the game withing the alloted 30 seconds, so I just may shell out the $37 for the full unlocked version in the near future.

As for the tower itself, I envision the passage through the tower being the insertion point into the game level, with the pedestrian zone connected to one end of the passage and the other exit being blocked by a car. This would naturally entice the player to move forward into the pedestrian zone without giving him or her the sense of being forced into this direction. I am nearly finished with the texture map for the tower and will then move on to making bump maps for it. When these are finally done, I will do a turnaround render and move on to the next model.

Sunday, January 17, 2010

FRAPS and Feedback from The Modern Language Journal

On account of the recent holidays and an unfortunate problem with my computer's motherboard, I haven't been blogging as much as I would have liked. Now that the computer problem is resolved and the powers that be have turned the heat back on in my office, I am once again in full production mode. A new year ahead of me and new ideas to explore!

I just finished reading Nitsche's Video Game Spaces and Gee's What Video Games Have to Teach Us About Learning and Literacy. I have started the latter on numerous occasions, but never really got around to finishing it. So, as the computer was in the shop, I went back to the library, checked it out, and finally read it cover to cover. I'm now on to reading Juul's Half-Real and am about half-way (no pun intended) through it.

Based on what I have read so far, I'm thinking about radically redoing the level in which I insert the player into the game. The main problem, I think, with the prior level design was the layout of the level itself, which might lead to a lack of player guidance and direction. It occurred to me that a real-life pedestrian zone is functional in that it creates an environment in which people can walk around and explore, without having any clear objective or destination. We want people to stay and shop, not quickly move through while looking for something else. A virtual gamespace, on the other hand, will have specific objectives and instructional goals, and the player/learner will need to move through this space in order to maintain a sense of momentum in the game. We want players to move forward in the game and continue learning, but while simultaneously maintaining a sense of the real pedestrian zone that allows limitless lateral movement and exploration. And it has to be fun. A tricky design task, no doubt. I have some ideas of what the new level should look like and will post these later.

Looking for a way to create screen captures of gameplay for my current book project on digital game-based learning for second language acquisition, I came across FRAPS. It's shareware, so some of its functionality is limited unless you decide to purchase it, but it makes documenting gameplay so much easier than taking screenshots of a game in windowed mode. I'll certainly be using this a lot as I work on the book. It works with programs running on OpenGL or DirectX, which means that it also makes nice screen captures of Blender. Here, my recent work on a game model:

Finally, I just found out that the article I submitted to Modern Language Journal was rejected. This is what they had to say:
This is an insightful and well-argued paper on the design features and potential value of gaming environments for language learning. It makes a distinct contribution in this area. Although we have published in this area recently (most recently the 2009 Focus Issue), as the author notes, we do not feel that The MLJ is the appropriate venue for this paper. The MLJ readership would be interested in implementation and empirical investigation of the uses and effects of such games, rather than in the complexities of the design. While some designers and CALL specialists no doubt consult The MLJ from time to time, they are more likely to focus on specialized technology journals. I strongly recommend the author to approach one of the technology or CALL oriented journals, e.g. LLT, ReCALL, System.
I am very thankful for the positive feedback and high regard of my scholarship. On the other hand, what threw me was the emphasis on implementation and empirical investigation. After all, somebody has to make these types of learning environments and without initial and well-formulated inquiries into how to build them, there would never be anything to evaluate anyway. So, I added some extra references to research I recently read and sent it off the next day to The Foreign Language Annals. We'll have to see what comes of it there.