Tridef's Virtual 3D mode is just as good as 3D Vision

[Exposed]

Tridef's Virtual 3D mode is just as good as 3D Vision.

Is that a valid statement? If so, why? If not, why?

I'm specifically referring to overall depth quality. Tri-def uses the Z-buffer for Virtual 3D mode, just like Crysis 2. 3D Vision integrates into the Direct3D pipeline and creates two unique perspective views, one for each eye. Yet some people claim Tri-def's Virtual 3D mode is every bit as good as 3D Vision regarding depth quality.

Thoughts?

[cybereality]

Virtual 3D is in no way as good as true dual-render 3D. Although it may look acceptable at times, its clearly flawed in many ways (as you have probably seen if you played Crysis 2). The advantage is that it will work with more titles and the performance cost is very low. But it terms of pure visual quality, it can't compare to real 3D.

[Likay]

Crank up the convergence using virtual 3d-mode and then everything should be clear.

[Exposed]

Thanks for input. It seems to be that in order for someone to actually believe Tridef's Virtual 3d would look the same as 3D Vision or anything else, they'd have to use some pretty low stereoscopic settings in the first place.

I've used Tridef extensively and I really can't see how anyone can compare it to the other "true" modes. Virtual 3D mode just doesn't capture that "ultra realistic" depth perception on objects. My only explanation for someone's rationale that Virtual 3D is "just as good" as true 3D is that they're not using an adequate amount of stereoscopic settings in the first place, especially since 3D Vision and others defaults to a really low depth value.

[Likay]

Tridefs virtual 3d is almost most accurate described as a 2d-3d "conversionprocess". As you say the engine uses z-buffer values which then roughly said applies a depthmap to an already rendered 2d-image. This is probably the best way to produce 3d from 2d on the fly and in my opinion it completely sucks compared to dual rendering.
One of the problems are occluded zones (an area which is visible in one eye but not the other which adds a great part to the stereoscopic experience). Illustrate this by holding a tennisball 1 foot from your face. It's easy to imagine that the left eye sees an area of the ball which the right eye doesn't and vice versa. This phenomena gets less when increasing the distance but then the actual stereoscopic effect of course also decrease. The "2d+depth"-process needs to either suppress some of the visible areas for one eye or try to reproduce what it doesn't "see" from the 2d-image to create 3d. This is why the tridef virtual 3d, crysis 2 native 3d etc only works well at a distance but then also with an unsatisfying degree of 3d-experience.
There are however examples where virtual 3d is better then dual rendering. An example is when there are too many or too big anomalies using dual rendering that destroys a good experience. Personally i don't play using any kind of 3d at all in those occasions. It's each of their own to decide of course.