it's been over 5 years since I posted here - (back in the day I was an early adopter of edimensional glasses using a Diamondtron 2070SB CRT).
After a long break from S3D gaming, I am happy to be back in the game as of last week with an Asus VG278H monitor to accompany my current PC setup (Core i7 - GTX670 - Win7 Ultimate 64bit).
My question:
Has the issue I am highlighting below ever been addressed, an official statement released, or a workaround/fix found?
The background:
My initial impressions of S3D with the Asus screen were (and still are) very positive but one thing quickly started to bug me (as detailed later in this post) and after doing some googling, my observations were confirmed in the below forum topic:
https://forums.geforce.com/default/topi ... ch-bundle/
What surprised me is that this thread runs from Feb 2009 to Feb 2011 and nowhere within the thread is there a response from any representative of Nvidia.
I am using the very latest Geforce Driver64bit 313.96 beta (from 29/01/2013) but I have also tested with the last WHQL release 310.90 (05/01/2013) and observed the same problem with both driver sets.
The issue:
Note that my findings are all covered in the linked topic, albeit 4 years ago and on Samsung 120Hz displays.
When in 120Hz 3D mode over Dual Link DVI, I am observing that each of the separate images when viewed without glasses in 120Hz 3D mode looks somehow grainy (almost like a checkerboard effect.)
I have observed this in games such as Trine 2 and Batman Arkham City (although in Arkham City, it's less noticeable.)
When viewed through the glasses in 3D mode, although it's not a game changer, it is clear that there is considerable image degradation compared to 2D mode, and I'm not just talking about loss of brightness, or distorted colour representation compared with 2D, but actually a loss of render quality at the pixel level.
Switching between 2D and 120Hz 3D mode and inspecting without the glasses on shows a clear difference in the quality of rendering.
Repeating this switching/2D 3D comparison exercise When output is set to 3D Vision Discover mode (anaglyph) (with the display still set at 120Hz), while colour is obviously lost because of the anyaglyph mode the 2D image in this mode is identical to it's 3D equivalent.
(ie. looking at the game image through the red lens while in 2D mode, compared to the 'left' stereo image when in 3D mode through the red lens shows an identical quality image.)
While I am sure 120Hz active is better for gaming than a passive gaming display where I would clearly lose some vertical resolution, the resulting 120Hz 3D is not a perfect representation of its 2D equivalent, while the 3D Discover anaglyph mode IS (other than the colour) and anaglyph doesn't require the expensive display/glasses combination!
The end result is that while I have purchased the current 'high end' 3D display for 3D Vision with the purpose of ensuring full frame 1080p to each eye, (as opposed to half vertical resolution per eye on a cheaper passive panel) this 'full quality frame' is not actually displayed for each eye in 3D mode...quality is lost, and along with it some of the reason for going with the 120Hz shutter display in the first place.
Final thoughts:
Has there ever been any confirmation that this is a limitation of the 3D Vision implementation, the current display hardware, or Dual Link DVI transport?
One comment in the thread I have linked suggested that Nvidia may be using some trick to save on shader work in 3D mode...only doing shader rendering at half resolution, or some other trick to reduce GPU load and keep the framerate up which results in the image degradation but in anaglyph 3D mode, the output perfectly matches what is rendered in 2D mode and in 120Hz 2D mode through Dual Link DVI, there is no issue with the image as far as I my observations can tell so I don't think that the problem is in the GPU or with the data transport to the display.
So 1920x1080p@120Hz 2D is the same data output with regard to DVI as 1920x1080p@120Hz 3D (60Hz per eye) - the only difference is that the display has to perform switching in 3D mode, and doesn't have to perform switching in 2D mode.
Could the issue be with the display itself?
Many thanks in advance to anyone who takes the time to read this post
Robin.
