stereoscopynews-logo

button-3d-blu-ray  3D Movies Trailers    Follow our 3D News RSS feed Follow StereoscopyNews on Twitter Follow StereoscopyNews on Facebook

Access over 3,500 articles in 70+ categories in the HotNews menu here above.

2D + Depth = 3D of the future

3D Fusion (New-York, USA) sells not only autostereoscopic 3D displays, but also a whole new way to capture, transmit, and display stereoscopic images. This way is called "2D+Depth", a not-so-new concept, that not yet widely implemented by anybody else. For many insiders, the future of 3D TV is coming...

2Ddepth

The idea is to transmit one 2D image and one "depth map" carrying the Z information of each pixel. Once that info is transmitted, the receiving 3DFusion-equiped decoder creates the left and right images for your 3D TV. This means that you can easily adjust the 3D effect from 0% to 200% just by turning a knob until it looks good. It is also easy to generate not only 2 but as many as 8 or 9 views in real time and to feed them to an autostereoscopic display with adujstable depth! And on top of those advantages, you get a transmission bandwidth smaller than what is needed for a good stereo pair... Notze that "2D+Depth" is already included in standards such as HDMI1.4 and the SCTE specification.

Read an in-depth analysis of the 2D+Depth and 3DFusion concepts by Rodolfo La Maestra in HDTV magazine (part of his multi-page series on autostereoscopic displays).

We discovered 3D Fusion in StereoscopyNews in October 2010.

A reader's comment :

I think this format is very EASY and very compatible with current bandwidth and encoding options, however it has great limitations when it comes to:
-          Artefacts in edges for ALL views you render from it, even with stereo (unless you accept very very little 3D effect).
-          Missing background information (there’s a so called declipse info addition for that, but most people don’t use it, AND it still doesn’t work if you have ‘multiple’ layers of backgrounds as in real life…)
-          Not able to process any 3D transparencies  (also for this there ought to be some experimental layers/info etc but it’s messy)
I have good friends that work with this format and I respect that. But we did many many tests in the past years with this format and we can say without any doubt it has not and will never have the qualities needed for the future. Even people very close to the development of this format I know personally agreed about the limitations and that all efforts to make it better are creating new problems due to the unchangeable “core-thought” of this specific format.
On the the other hand, again, I have to see it also has specific advantages, and there are some applications where we use it do as part of some processes, such as in conversions. So it’s not only bad!