Started 7 years ago (2017-11-13T23:00:00Z).
Ended 6 years ago (2018-01-11T23:00:00Z).

I set this to daily but will mostly only update on weekdays unless I do work on weekends. U _ U This is an informal project log for my final project for Experimental Media Production, which will serve as material for my report later on.

Have a question for this streak?Ask a question...

Recent submissions (18 total)

I forgot to update my streak here, but while I've been gone I've mostly finished the project. The models have been brought into the Unity project and I've trained the program to recognise natural images instead of markers so that I can use images that look more like a picture book than markers do.

One issue I had with the Live2D models is that they're normally not intended for 3D environments, so when you make animations they don't clip the model to the background, because it's assumed that you'll place it in a normal 2D frame yourself as you please, but that's not the case when you work in AR and have a freely moving camera. I solved this by making a simple png file with a box which I set the model parts to clip to and then turned the opacity of the box to 0% so that the box clips the parts of the model I want to hide without being visible, and then keyed the parameters for the box in the animation files so that I can use the same model for all the animations and have custom heights of visibility for them all.

What's left to do is simply to make sure that everything works, put the book together neatly and do the required writing. I should have no problem getting it done by Wednesday.

Only have one more motion left to animate, then I can grab snapshots of it and start to arrange the pages for AR.

Ran into some software problems and couldn't get much work done, but read up on making markers.

Got it working with my own model and a static animation! (Is not apparent in the video, but it's working and that's the important part.) Next steps is to generate my own patterns and fixing the actual pages so that it will look right, and getting all the content in.

Finished setting up the model. Gave up on the idea of having her be able to stand up from an at least crouching position as I couldn't get the joints to behave nicely. Might revisit that later if I have the time.

Continued setting parameters for the model. Learned how to duplicate and flip a set of art meshes/deformers/parameters and editing them to make them correct. One thing that doesn't automatically copy over is any glued vertices. When I do the legs I'll have to check if duplicating the glue will make it carry over or it you have to manually re-glue every time.

Continued to work on the model. Consulted a bunch of tutorials and the Live2D documentation to get more systematic knowledge of how to set up a model properly as before I have only relied on intuition and the only partially translated Live2D official documentation. I now have a better understanding of what I'm doing, although I'm not sure this model I'm making now would work in eg. FaceRig out of the box. I'd probably have to modify it, but since that isn't the primary purpose of this project and all animations will be my custom ones anyway that doesn't matter much. I have now animated everything on the face, and I'll finish animating the entire head before stopping for today.

Finished the illustative part of the model and imported it into Live2D after updating the program. Might have to edit the model parts so they are easier to animate.

The model's parts are completed and the base colours, now I have to add details!

It helped me think around what exactly I have to draw, and I think I will only have to do two models, one seen from the front that is able to turn, and one that is entirely in profile, although only a bust, as I don't need to animate the legs for this story in profile.

Needed to fiddle around with some settings, but I didn't need to alter any code at all which was refreshingly unexpected. Not going to post a preview at this time as I used the Koharu sample model to make sure things would work before getting into content production in case I'd have to change my approach, but it looks like I'll be able to do things exactly as imagined.

Solved my SDK/JDK issues and managed to get simple AR working in a simple Android apk!

Started sketching out a simple story for the project.

Looked around for cheap/free tools for AR that works with Unity and found ARToolkit, an open source toolkit for AR that works with all larger platforms and has a unity package that seems simple enough to use.

https://artoolkit.org/

Today I had a meeting with my teachers to make sure that my ideas would work for this part of the course and it was approved. The main points of my project will be:

  • It uses Live2D, optionally Euclid if I get time over
  • I will attempt to make a book with parts that are enhanced in AR
  • I won't have time to make an advanced and fleshed out story, so it will simply be a cute small thing about the character becoming sentient and coming to life on the paper
If possible I want to avoid making a separate app for users to have to download but this might not be feasible for me in such a short time and limited (read: no) budget, so if I can't make it work as a web application I'll make it a mobile app for Android.

Not sure how far I can take this, but I think ideally I'd like to make a digital popup-book where part of the story is on printed paper, and part of it is with digital pop-up characters, that could be really fun. Maybe also a mix of physical and digital popup content? Either way I don't think I'll have the time to learn the technology and make a fully fledged story and printed material all for this project, so it'd probably just be a short proof of concept book.

Today we kicked off the final projects for Experimental Media Production. We are basically free to do whatever we want as long as we have an experimental approach. We're also free to define experimental to some extent as long as we can argue convincingly about it.

I haven't clearly defined what I'm going to do quite yet. We have individual meetings with teachers on Friday to discuss our ideas, so I should have something more solid by then, but as a starting point I'd like to work with Live2D and/or Euclid during this project, and then we'll see how far I take it. It depends on how quickly I get the hang of things. I did some loose brainstorming, though, and I want to do something AR-related if I can get things to work.