[8:13]

Overview

Resolution is defined before framerate speed. You define the cell resolution then the cells framerate displays video information matching that resolution.

2x resolution = 2x information framerate

1x resolution = 1x information framerate


You make may two distinct cells that are different from each other, or clone one cell so two cells are made but they are the same.

Displaying two distinct cells means you may either displaying them once or multiple times.

Displaying cloned cells means you show them multiple times.


Displaying information twice means the resolution + framerate is not equal to itself, the framerate does not match the resolution.

Parallax method

Clones the resolution on one cell, the cell framerate shows information at one times.

This means the resolution does not match the framerate: the resolution is at 2x (cloned) and the framerate of information is at 1x.


When shown as 48fps content in a 120Hz 3D TV


Left Left, Left, Left

Right, Right, Right, Right


So 4x resolution with 1x framerate of information.


Or with real stereoscopic pair shown as 48fps content in a 120Hz 3D TV


Left, Left

Right, Right


So twice the resolution of framerate of information.

3D effect

Is created by resolution at framerate speed with no artifact from resolution and framerate being different.


Simply having time sequential 3D: left left, right right, is not creating the framerate seen in time parallel 3D.

Time parallel 3D is twice the resolution and framerate of one 2d cell: resolution and framerate are the same, 2x resolution = 2x information framerate.


Until you have seen Time Parallel 3D properly you have not seen 3D effectively. Time Sequential 3D does not qualify as the same quality.

Edit. adding following replies I made to the first post for easier reading so you can understand my idea better:


When I say Resolution I'm using in context of a stereo pair of pictures: having two different pictures increases the resolution seen.


Why? Because the orientation seen in a stereo pair is from two different pictures that are pointing in the direction of what is being filmed. Two different pictures means twice the resolution.


Each picture carries it's own frame rate when considering a stereo pair. So the left side is 25 fps and the right side which is a slightly different picture is 25 fps.


(2x resolution = 2x framerate) = (left at 25fps + right at 25fps)


I know that the Time Parallel method can use multiples of a frame to fit into 120Hz TV's and still be called Time Parallel, but it's not what I call Classic 3D.


When the frame is doubled and the fps is made out of sync due to this (as described in the first post), then what happens is what cakefoo described, desynchronization.


So to fix this the left eye has: a single 25 fps shown a single time, not multiple times. Then the right eye shows 25 fps the same way, so there is no desynchronization.

This results in the 3D being more immersive.


|L1,R1|L2,R2|L3,R3|

 

[BlackShark]

You are making it waaay too complicated.


Time parallel 3D = left and right eye views are presented at the same time

There are no considerations for display technology, picture scaling, resolution or framerate in the time parallel/sequential concepts.

 

[8:13]

When I say Resolution I'm using in context of a stereo pair of pictures: having two different pictures increases the resolution seen.


Why? Because the orientation seen in a stereo pair is from two different pictures that are pointing in the direction of what is being filmed. Two different pictures means twice the resolution.


Each picture carries it's own frame rate when considering a stereo pair. So the left side is 25 fps and the right side which is a slightly different picture is 25 fps.


(2x resolution = 2x framerate) = (left at 25fps + right at 25fps)

 

[BlackShark]

You are just making it too restrictive by inserting additional technical restrictions that are not required.

I see perfectly what you are talking about and why you believe one is superior to the other, but the reason for that is that you think about better hardware for one than for the other.


The reason for time parallel being better than time sequential presentation, is not the amount of x+y+time resolution, it's the presentation of this resolution to the eyes and the effect it produces inside your brain.

 

[Lee Stewart]

Should be pretty simple to test. Just go see a 3D movie at a Digital IMAX 3D theater (Time P) then go see the same movie at a RealD theater (Time S)

 

[cakefoo]

The problem with time-sequential is the motion artifacts, not the refresh rate:


Your brain would ideally be piecing together frames like this:

|L1,R1|L1,R1|L2,R2|L2,R2|L3,R3|L3,R3|


But you brain is also going to piece together the bolded consecutive frames:

|L1,R1|L1,R1|L2,R2|L2,R2|L3,R3|L3,R3|


..resulting in a slight hint of desynchronization.

 

[8:13]

I know that the Time Parallel method can use multiples of a frame to fit into 120Hz TV's and still be called Time Parallel, but it's not what I call Classic 3D.


When the frame is doubled and the fps is made out of sync due to this (as described in the first post), then what happens is what cakefoo described, desynchronization.


So to fix this the left eye has: a single 25 fps shown a single time, not multiple times. Then the right eye shows 25 fps the same way, so there is no desynchronization.

This results in the 3D being more immersive.


|L1,R1|L2,R2|L3,R3|

 

[cakefoo]



Quote:
Originally Posted by 8:13 /forum/post/19517052


I know that the Time Parallel method can use multiples of a frame to fit into 120Hz TV's and still be called Time Parallel, but it's not what I call Classic 3D.


When the frame is doubled and the fps is made out of sync due to this (as described in the first post), then what happens is what cakefoo described, desynchronization.


So to fix this the left eye has: a single 25 fps shown a single time, not multiple times. Then the right eye shows 25 fps the same way, so there is no desynchronization.

This results in the 3D being more immersive.


|L1,R1|L2,R2|L3,R3|

But I believe you are getting wronger with this reply. That would make it worse. Shutterglass tech relies on our brains not minding the de-synced frames. We're supposed to see


|L1,R1|L1,R1|L2,R2|L2,R2|L3,R3|L3,R3|


but since our brains are trying to put together adjacent frames, we piece together adjacent frames that weren't meant to be put together.


|L1,R1|L1,R1|L2,R2|L2,R2|L3,R3|L3,R3|


You want the bad frames to be unnoticeable, and to do that you need to decrease the amount of time a sequence like |R1,L2| is visible. In a sequential system a bad frame will always rear its ugly head 60 times per second in 60fps content, but we can control the duration they're visible, by increasing the refresh rate of the lens and panel like this:


|L1,R1,L1,R1,L1,R1,L1,R1|L2,R2,L2,R2,L2,R2,L2,R2|L3,R3,L3,R3 ,L3,R3,L3,R3|


8hz per frame at 60fps = 480hz or ~2ms per refresh. A set of refreshes (like |R1,L2|) would last 4ms, down from 8ms of current tech, and down from 16ms the system you proposed would do.

 

[8:13]

For shutter glasses tech I suppose your right, but then I was talking about 3D in general and by that I assumed autostereoscopic displays where desyncing wouldn't be a consideration.


Thank you for correcting me.

 

[BlackShark]

On the contrary time sequential 3D would be an excellent consideration for auto-stereoscopic displays. There are so many different pictures to draw for so many viewpoints. That's way too many pixels to cramp into such a small space, and the precision requirements of the lenticular array would quickly become very expensive due to the difficulty to maintain high precision optics on such a scale.


By using a high refresh rate FullHD pixel array and an moving optical device that restrict the light path into a single targetable direction, you could produce a full resolution auto-stereoscopic picture by scanning the light path trough the whole viewing space and update the picture for each different viewing zone. I think that's what Microsoft is working on in their lab.

 

[gain3]

 http://www.siggraph.org/publications...tting-it-right


heres time sequential depicted, how it can apparently "break up". I play with pretty fast action games at fix 120fps ( no stutter, glitch). Never saw this happening, therefore I consider it overrated. Oh and I cant see rainbow either

.


Will try time parallel DLP in a few weeks , ( mainly because I want better colors/brightness - Brilliantcolor™) I'll report back.

 

[obveron]



Quote:
Originally Posted by gain3 /forum/post/19538642

http://www.siggraph.org/publications...tting-it-right

I enjoyed this read, thanks.

 

[8:13]

 synced-3d-vs-field-sequential-3d-seeing-is-believing: Link to video by RogerMaddy

 

[mkoss]

Just go see a 3D movie at a Digital IMAX 3D theater (Time P) then go see the same movie at a RealD theater (Time S)


Is this true? I thought that both were time parallel but differed in that I max in linear polarized and real D is circular polarized.

 

[icester]



Quote:
Originally Posted by mkoss /forum/post/19875837


Just go see a 3D movie at a Digital IMAX 3D theater (Time P) then go see the same movie at a RealD theater (Time S)


Is this true? I thought that both were time parallel but differed in that I max in linear polarized and real D is circular polarized.

RealD is a single projector system and as such it is only able to display left and right images sequentially.


Mathew Orman

 

[mkoss]

RealD is a single projector system and as such it is only able to display left and right images sequentially.


Mathew Orman


Ok this is from the theatre perspective of presenting RealD. However they chose not to go the 2 projector route(time parallel), for whatever reason, that would have still allowed circular polarization. In terms of doing a time parallel presentation linear or circular polarization can be used based with each having its caveats.

 

[BlackShark]

It's a bit more complicated : the fact that it's a single projector alone does not prevent time parallel display.

For instance the Sony 4K 3D process and the LG CF3D projector are single projector systems that provide time parallel projection.

The RealD system is time sequential because there is a device (the Z-screen) put in front of the projector that creates the alternating the polarisation in sync with the projector.


All three of these systems use circular polarisation compatible with RealD 3D glasses.

At the moment, almost all the theatres that distribute realD glasses use the RealD time sequential system but as technology advances and more products are available to cinemas operators, it will become more and more difficult to identify which type of presentation is used just by looking at the glasses.


Even Dolby3D (which is time sequential too) can be made in a time parallel way.

 

[mkoss]

I haven't researched theatre presentations much but knew that I had the option to use either circular or linear polarization with 2 projectors using either HTPC or 2 3D-Xl's and a silver screen to maintain polarization for time parallel presentations which would be my ultimate choice. I already have a high gain silver torus and 2 Optoma HD7100's to go this route. Only need to 3D-XL's and polarization filters/glasses to go passive.

 

[icester]



Quote:
Originally Posted by mkoss /forum/post/19877188


I haven't researched theatre presentations much but knew that I had the option to use either circular or linear polarization with 2 projectors using either HTPC or 2 3D-Xl's and a silver screen to maintain polarization for time parallel presentations which would be my ultimate choice. I already have a high gain silver torus and 2 Optoma HD7100's to go this route. Only need to 3D-XL's and polarization filters/glasses to go passive.

That is only 720p.

For 1/2th of the price one can get full HD 3D passive system.


Mathew Orman



Mathew Orman

 

[mkoss]



Quote:
Originally Posted by icester /forum/post/19877234


That is only 720p.

For 1/2th of the price one can get full HD 3D passive system.


Mathew Orman



Mathew Orman

I know it's only 720P but I already had the projectors for a while. The HD7100's

have a very good picture quality. Bought them back when I got the scaler with one of them in the 7300 package. I have only used one to date and have a fresh bulb for the other, so trying to go 1080P would be more expensive for me.

I can always upgrade down the line. The only thing I'm worried about is the loss of Lumens with the filters.