Phenomenon 2008 expectations

#21

Interesting considering the big Phenomenon question…

http://www.macnn.com/blogs/?p=468
"On November 20, the US Patent & Trademark Office published three of Apple’s newly granted patents. The most notable amongst the three is titled Depth ordering of planes and displaying interconnects having an appearance indicating data characteristics . This particular patent relates to Apple’s Shake application in respect to depth ordering of planes and displaying interconnects having an appearance indicating data characteristics.

Patent 1: Depth Ordering of Planes and Displaying Interconnects having an Appearance Indicating Data Characteristics

Patent Background: Compositing and effects applications (such as Shake by Apple Computer, Inc.) receive and process multimedia items (e.g., text, image, video, audio, etc.) to produce a multi-layered composited output. Such applications can import and composite multimedia items and allow transformations of multimedia items that change characteristics (e.g., position, size, shape, and rotation) using transformations such as pan/move, rotate, scale, etc. A graphical user interface (GUI) used in these applications display multimedia items in a three dimensional (3D) workspace and allows a user to select and manipulate the displayed multimedia items. Keyboards or cursor controlling devices (e.g., computer mouse) are used to interact with the GUI to select and manipulate multimedia items in the 3D workspace.

Apple’s Abstract: A method for determining a depth ordering between two planes in a 3D workspace of an application is provided. The method includes projecting vectors from a reference position through points of interest on the planes, determining corresponding points on the other plane, determining a number of points of interest on a first plane having a depth lower than its corresponding point and a number of points of interest on a second plane having a depth lower than its corresponding point, and setting the depth ordering based on these numbers. A method for displaying interconnects between nodes of a tree displayed in an application is provided. A node represents a multimedia item or a function and an interconnect represents data passed between two nodes. The method includes displaying an interconnect in the tree having an appearance indicating at least one characteristic of the data being passed between two nodes.

Apple’s patent covers the following subject matter in depth: Process Tree and Multi-Plane Nodes, Multi-Pane Interface, Transforms and Controls, Hot Key Selection of Transforms and Activation of Onscreen Control Guidelines, Multi-Plane Depth Ordering, Identification of Data Characteristics through Interconnect Appearance and Transformation of Screen Space Cursor Movements.

Apple lists Mitchell Scott Middler (Hermosa Beach, CA), Ron Brinkmann (Hermosa Beach, CA) and Peter Warner (Paris, FR) as the inventors of this patent which was initially filed in April 2005. For further details, see patent 7,298,370.

Apple’s second patent win of the day titled Apparatus and method for indicating password quality and variety, generally relates to apparatuses and methods for indicating the quality and variety of a password used for computer system authentication.

Apple’s third patent win of the day titled Wire bonding method and apparatus for integrated circuit, relates to packages for integrated circuits. More particularly, Apple’s invention relates to package arrangement to reduce loop inductance."

So what might this mean for us?

#22

Interesting considering the big Phenomenon question…

http://www.macnn.com/blogs/?p=468
"On November 20, the US Patent & Trademark Office published three of Apple’s newly granted patents. The most notable amongst the three is titled Depth ordering of planes and displaying interconnects having an appearance indicating data characteristics . This particular patent relates to Apple’s Shake application in respect to depth ordering of planes and displaying interconnects having an appearance indicating data characteristics.

Patent 1: Depth Ordering of Planes and Displaying Interconnects having an Appearance Indicating Data Characteristics

Patent Background: Compositing and effects applications (such as Shake by Apple Computer, Inc.) receive and process multimedia items (e.g., text, image, video, audio, etc.) to produce a multi-layered composited output. Such applications can import and composite multimedia items and allow transformations of multimedia items that change characteristics (e.g., position, size, shape, and rotation) using transformations such as pan/move, rotate, scale, etc. A graphical user interface (GUI) used in these applications display multimedia items in a three dimensional (3D) workspace and allows a user to select and manipulate the displayed multimedia items. Keyboards or cursor controlling devices (e.g., computer mouse) are used to interact with the GUI to select and manipulate multimedia items in the 3D workspace.

Apple’s Abstract: A method for determining a depth ordering between two planes in a 3D workspace of an application is provided. The method includes projecting vectors from a reference position through points of interest on the planes, determining corresponding points on the other plane, determining a number of points of interest on a first plane having a depth lower than its corresponding point and a number of points of interest on a second plane having a depth lower than its corresponding point, and setting the depth ordering based on these numbers. A method for displaying interconnects between nodes of a tree displayed in an application is provided. A node represents a multimedia item or a function and an interconnect represents data passed between two nodes. The method includes displaying an interconnect in the tree having an appearance indicating at least one characteristic of the data being passed between two nodes.

Apple’s patent covers the following subject matter in depth: Process Tree and Multi-Plane Nodes, Multi-Pane Interface, Transforms and Controls, Hot Key Selection of Transforms and Activation of Onscreen Control Guidelines, Multi-Plane Depth Ordering, Identification of Data Characteristics through Interconnect Appearance and Transformation of Screen Space Cursor Movements.

Apple lists Mitchell Scott Middler (Hermosa Beach, CA), Ron Brinkmann (Hermosa Beach, CA) and Peter Warner (Paris, FR) as the inventors of this patent which was initially filed in April 2005. For further details, see patent 7,298,370.

Apple’s second patent win of the day titled Apparatus and method for indicating password quality and variety, generally relates to apparatuses and methods for indicating the quality and variety of a password used for computer system authentication.

Apple’s third patent win of the day titled Wire bonding method and apparatus for integrated circuit, relates to packages for integrated circuits. More particularly, Apple’s invention relates to package arrangement to reduce loop inductance."

So what might this mean for us?

#23

Possibly the depth ordering relates to being able to take parts of one image and instead of placing either in front or behind, you can place at some depth in another image. For example, the flower in the field could be in amongst the other flowers and the butterfly can go between them. Possibly it’s a technique for extracting depth information from 2D images.

It might also allow intersecting planes unlike the current Multiplane. Hopefully they’ll allow multiple parenting too for layers.

I hope this means transform concatenation beyond arbitrary intermediate nodes. When you put a composite node between two transforms, it shouldn’t break the concatenation because you can simply move the transform node above the composite node and you get the same result. The software should be able to figure this out by itself.

The mention of 3D workspace suggests hardware acceleration for all items, which is a good thing in some ways because it means fast manipulation of all the source elements. It’s a bad thing in that it will likely require a good graphics card. If you composite too many items or use high resolution images, you might run out of video memory. Plus hardware rendered images don’t always come out the same as software rendered ones.

Generally I’m in favor of hardware acceleration throughout the whole program though as the performance increase far outweighs quality degradation during preview. There would be less need for caching if the viewer preview can be generated quickly enough. You can compare the difference in the current multiplane node. Throw in a few layers in software mode and hit the play button. Then do it in hardware mode and there’s a world of difference. This is lost when you add heavier effects like blurring so complete hardware acceleration and quality downgrading in the interests of performance would be a very good thing IMO.

#24

After looking at Ron Brinkman’s book the Art and Science of Digital Compositing (1999), I notice a mention of this exact thing on p261:

“Research is being done on allowing Z-depth information to be automatically extracted from live-action footage. Stereoscopic photography inherently contains some depth cues and even a single camera view can give some information if it is moving enough to introduce parallax shifts. Sophisticated software algorithms are able to analyze such imagery and can often produce useful depth information for the scene. As the availability and reliability of these tools increases, expect to see more and more use of Z-depth compositing in conjunction with live-action footage.”

#25

I don’t think it has anything to do with extracting depth from an image. It sounds like an automatic depth ordering system for the multiplane node, instead of ordering planes according to a list. It would just determine when more than half of one plane is in front of another plane and draw it on top of (after) the other one.

Another possibility would be a 3d node interface, where you could see the node icons and their connections in a 3d view and spin it around. That’s something I thought would be a cool novelty ever since I saw a node interface, but I doubt they would actually make it since going 3d usually adds a thick layer of confusion to any app. Might explain the “displaying the interconnect” line (aka rendering noodles) in the patent though.

Extracting depth from 2d is an interesting topic though, I’m most interested in its implications for stereoscopic movies. I would guess that ILM is using this kind of technology to do 3d conversions from 2d movies. Would be cool to see the capability in publicly available software.

#26

Zdepth extracted from an image has been around for atleast 6-7 years in plugins for Shake and other compositors(Furnace, etc…). So a patent based on that would be void because of prior art.

#27

It already does that though with the autoOrder toggle. That determines the z-value of the plane, which is well-defined and changes which one goes in front.

That could keep the space used down on some composites and maybe could be used instead of grouping so that you don’t have to open a group and then move all the other nodes to get enough space to see what’s going on. I agree it would likely be confusing if you are trying to manipulate noodles with depth to think about.

I don’t know, I’ve never really taken to stereoscopic movies and I wouldn’t like to see it becoming a popular trend so that we end up with 3D clips everywhere that people can’t watch without 3D glasses. If Phenomenon goes back to the old Shake price, that might not happen though.

For creating 3D elements from 2D, that would certainly be a good tool to have, especially when it comes to tracking and trying to blend CG imagery with film.

I think it’d be useful so that you don’t need matte painting done so much. Just extract the depth information and drop CG elements into the film. Of course then we’ll have no end of videos on youtube along the lines of OMG a new UFO sighting in Tahiti, it looks so real it must be real, post on every forum about it.

But if Brinkman said about the research in 1999 and his name is on the patent, maybe the plugins were based on his research. Could this maybe just be the same thing but built-in? Just like some industry standard keyers are included in Shake now.

Plus it could be a unique algorithm that does it better than the others.

#28

I think Apple needs to buy Nvidia and make a hardware video acceleration product. That would be the Ender.

#29

Hi all,

Is there any news about the new software.

#30

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.