3D Vision Blog

A normal user's look into the world of 3D Stereo Technologies

3D Vision Blog header image 2

Changing the Convergence Level in 3D Vision

June 23rd, 2009 · 10 Comments · GeForce 3D Vision

3dvision-convergence-option


Normally when you are using 3D Vision you have control only on the level of depth by either turning the scroll wheel on the IR transmitter or by using CRTL + F3 or CTRL + F4 to decrease or increase it. And this is normal as in most games you don’t have to play with convergence, because everything is usually being set in the available profile, but what happens if you run a game that has not been profiled by Nvidia? If you run a new game, or some old and not very popular one you might notice that there is nothing impressive when you run it with 3D Vision. Here you can try to change the convergence in order to improve the situation and it actually helps… usually, but there is no guarantee. But even if the game looks good by default you can achieve even better results by changing the convergence and thus improving the level of depth, the pop-out of the screen effect or even both…

You should know that by default the keys for changing the convergence level are CRTL + F5 and CTL + F6, but they are not active so even when pressing them nothing will change. In order to be able to control the level of convergence you must first “Enable the advanced in-game settings” from Nvidia’s Control Panel by going in to “Stereoscopic 3D”, choosing “Set up stereoscopic 3D” and then opening the “Set Keyboard Shortcuts”. When you activate the advanced settings you would be able to use the CTRL + F7 key combination in order to save the custom settings you set by changing the level of convergence so that you will not have to make them each time you run a given game. However you should know that there is still no graphic representation showing you the level of convergence, so when you are changing it you must carefully watch the changes on the screen. Nvidia is probably going to build additional functionality into a later version of the driver that will show you visually the level of convergence such as the one already available for the level of depth.

Tags: ··········


10 responses so far ↓

  • 1 applejack // Jul 6, 2009 at 04:44

    when using iZ3D driver for anaglyph red/cyan, its wizard suggest to set up convergence so the closest object on screen becomes one image.

    im confused though. I think the closest object on screen should use maximum pop out effect, but does a “one image” object pops out the most ?, or maybe its depth become the same as the monitor’s ?

    do you agree with iZ3D instructions, or do you have some other method for setting up convergence for best results ?

  • 2 Bloody // Jul 6, 2009 at 11:08

    I’m setting the convergence on per application basis, not the same for every game. Sometimes it is not possible to make a game having a great pop-out effect. What iZ3D suggests it for getting only a depth effect an not pop-out, this is what happens when you set the closest objects to appear as single image.

    To get a pop-out and depth effect at the same time your convergence point should not be too far or too close, but somewhere in the middle of the scene in the game. Meaning “the double image” should intersect (become single) somewhere in the middle of the scene of the game when you are viewing objects that are close and ones that are far. And as the detail goes backwards the image should start to go double again as it is in front (double when looking without the glasses on)…

  • 3 applejack // Jul 6, 2009 at 14:18

    I get it, so the “single image” object is some kind of a reference depth point where closer objects start popping out while further objects goes in depth..

    thanks! (correct me if im mistaken :) )

  • 4 Bloody // Jul 6, 2009 at 14:42

    Yes, it is like that… but it might not be a single object, but just a part of an object where the two “images” just cross to kind of create that point ;)

  • 5 jacob pederson // Jan 24, 2010 at 20:26

    I find the convergence setting has no effect whatsoever. Does this only work in some games? I can see the effect in WOW, where convergance is part of the ingame controll panel.

  • 6 jacob pederson // Jan 24, 2010 at 23:49

    Figured it out. You have to HOLD ctrl-f6 for a good 10-15 seconds before you start to see an effect. It must adjust in really small increments. Anyhow, the difference this makes in top down games like Titan Quest and Torchlight is absolutely incredible: from hi down there, to holy crap that’s awesome in 15 seconds :)

  • 7 jacob pederson // Jan 25, 2010 at 00:17

    Noticed a couple games where convergence effects something other than the render window. In Company of Heroes it effects the gui only. Even stranger, In Mass Effect it effects the light blooms only.

  • 8 Bloody // Jan 25, 2010 at 10:36

    It does not provide better results every time, but for some games tweaking the convergence can bring much better results in terms of depth perception. That is usually if they do not have a profile already made for them in the 3D Vision driver…

  • 9 Thomas Sturm // May 26, 2010 at 02:19

    How about another wheel on the transmitter for convergence? I often find myself fiddling with it to a small extend – but since I am playing more than one game, CRTL + F5 and CTL + F6 (like “quicksave” is often on CRTL + F5) might already be used – so you either have to change the keyboard settings for every single game, or change the convergence control to some wicked key combination no self respecting game would ever use for anything. The wheel and the meter would make me even buy a new transmitter. It’s not new either: when I started playing 3D, I had a GeForce 256 (the one with DDR-ram already!), and drivers where in the 70s, version-wise. The Elsa Revelator 3D shutter glasses (that I used until last week) came with a modified nVidia stereo 3D driver.

    With Elsa’s solution, you got a proper OSD when you pressed CRTL + F5, displaying a kind of object histogram of the current scene. You saw perfectly well how many objects were distributed at which depth. 3D area was marked by 2 vertical lines on that histogram: increase depth, and the lines move away from each other, increase convergence and both lines move to the left in parallel.

    The screen was a red line a bit off to the left in that histogram: if you had the leftmost line left of it, you had objects “in front of” the screen. Sheesh folks! the driver from Elsa was based off the 1998 nVidia stereo driver. Nothing much happened to it since then.

    Sorry for the rant. But nVidia had the edge all along with stereo 3D, only nothing was made of that potential. And now I have to google a blog just to find out how to access the convergence settings for a software I have been using for more than 10 years. Just to find out: they haven’t even changed, they are just disabled by default….sorry, I am at it again, aren’t I?

    Oh well, nVidia only could have made a few hundred millions from this in the last 10 years, so I guess so it’s not worth me going on about it.

    But I LOVE that depth-wheel on the transmitter, though!

  • 10 Manuel // Apr 15, 2013 at 04:29

    From my prospective it is not good to adjust the convergence to a point where objects pop out of the screen and here is why:
    If you are very close to a door or a wall (as close as you can get) you don’t want to perceive the door closer than your screen level\plane because, on the edges of the screen your brain goes crazy. The frame of your screen would be further back compared to the wall or door, which in theory should always be over your screen.
    Adjusting your conversion in order to have the closest thing at the level of the screen is a more realistic set up. Your screen should be a sort of window though which you see the world.
    Therefore anything you can see should be beyond the window.
    Well, this is my opinion.

Leave a Comment