Sounds like a very cool idea. I like the concept. Good luck working out all the kinks!
Risky is the new safe
Risky is the new safe
Published by Vlasta on March 23rd 2012.
RealWorld Icon Editor has always been about icons, images and 3D models. Over the years, the 3D editor received countless minor updates, but nothing substantial. I have been avoiding 3D for a long time and I cannot do that anymore. The time has come for a serious change.
The main problem: 3D is too complicated. Both the Unicorn3D editor and the Rendered Image editor are hard to use for new users.
Let's look at the current workflow. A user
- creates a 3D model (.u3d) in the Unicorn3D editor,
- creates a rendered image (.r3i) by connecting a 3D model, lights, camera and renderer settings,
- converts the image to icon.
I believe the most unintuitive is step #2. People have hard time guessing what the term rendered image is supposed to represent. The connection between 3D and image is unclear.
3D layers in images
Let's replace the term rendered image with the term 3D layer. While the concept is exactly the same, 3D layer brings the right associations. An image has given width, height and background color and the image is allowed to contain a 3D layer.
Embedding the current rendered image inside a layered image is not hard. Even the current version can be reconfigured to allow this. But a simple name change is not enough.
Redesigning the user interface
User interface is what makes or breaks a software application today. Let's look at this concept screenshot.
That's right, I want the raster layer and the 3D layer to have the same look and feel. Users should feel comfortable when editing the 3D layer.
As you can see, there is one raster layer with a photograph and one 3D layer with a 3D model of glasses. The glasses are currently selected (indicated by the blue outline and hatching. The "Object parameters" panel allows me to set parameters of the selected 3D object - the color in this case.
The 3D layer should
- be able to contain multiple objects,
- allow positioning of objects in 3D space, rotating and zooming them and setting their parameters,
- allow application of layer styles to 3D layers.
3D objects are not created or edited in the 3D layer. The 3D layer only references external 3D objects and takes care of their arrangement and composition with raster images.
Why it is risky?
Although what I have just described sounds trivial, there are some obstacles that need to be dealt with.
For example the user interface. My goal is to make the controls in the 3D layer so intuitive that even users, who have never worked with 3D graphics would be able to feel comfortable and in control. 3D layer should become the gate to the 3D world for many users.
Another problem is speed. Converting 3D scene to a high quality image may take minutes, which is not acceptable when editing. I am unsure how to deal with that at this moment.
While working on the 3D layers, I am running into unexpected problems on daily basis. It may take me several months to implement all the needed features and I may even fail. But I am going to try anyway, RW needs something new. Wish me luck.
Recent comments
Anonymous