Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

The end of the focus puller is nigh?

There's been some discussion here of plenoptics. Pretty awesome stuff, but I feel like it's something we wont see in cameras for a decade or so, simply because the data needed to process the images probably isn't very practical at this point. Who knows though, a 5k Epic Plenoptic sensor sure would be sweet!!
 
When those 28K Monstro sensors come out... it might actually work well!

The fact that you can extract stereo data too is awesome. The depth info you can get from that could be a lot better than a normal stereo disparity map - because you can gradually see around the image. It's not 2 views spaced on the x axis - it's an infinite range of views - left, right, up, down (well within the range of what can be seen within a single lens). This sets us up nicely for multi-view autostereoscopic.

At the moment, the resolution tradeoffs seem a little too great though.

Bruce Allen
www.boacinema.com
 
Yeah, Adobe's plenoptic effort has been online and on here for a while. I love talking about it and wouldn't be opposed to using it.

My one negative thought is this: VFX will become a nightmare if we're talking about multi-view autostereoscopic movies *shudder*

EDIT: Just made some napkin calculations. The lens divides the image into 50 sections. So to get a good plenoptic 4K image you'd need a sensor that's almost 692 megapixels.
The Epic 617 is only (ahem) 261 megapixels and change.

The target's moving Jim, catch up! ;)
 
Eren Ozkural said:
Yeah, Adobe's plenoptic effort has been online and on here for a while. I love talking about it and wouldn't be opposed to using it.

My one negative thought is this: VFX will become a nightmare if we're talking about multi-view autostereoscopic movies *shudder*
Click to expand...

Well you're right in that all of our old tricks won't work.

On the other hand... some of the stuff that has always been broken actually will start working.

For example, right now we can't generate an accurate 3D representation of a scene from a single 24fps camera stream.

But how about high-res plenoptic footage recorded from multiple cameras at 72fps? That might be enough info for a computer to reconstruct a filmed scene as high-resolution globs of voxels moving through 3D space.

That opens up a ton of fun IMHO. Difficult but exciting.

Bruce Allen
www.boacinema.com
 
Thanks Bruce, I don't know why I didn't think of that..Just had a hard time keying...less than favourable footage and doing a whole bunch of camera projection mapping. I'm REALLY looking forward to the day when automation steps up enough to let cg work happen closer to the speed of thought.
 
Worse case scenario Eren you just treat it like standard stereo fx shots with an initial processing stage to generate the stereo plates from the 50 source images.

You wouldn't need to process all 50 shots you would only need to add FX to the two final generated shots. I assume the DP and director would be setting the DOF and passing along the locked off shots anyway. And yeah with 50 images you could definitely do some automatic matte extraction using disparity.
 
You must log in or register to reply here.
Back
Top