Testing G-Sync Against V-Sync Disabled
I'm basing this on an unofficial poll of Tom's Hardware writers and friends easily accessible to me over Skype (in other words, the sample size is small), but most everyone who understands what V-sync is and what it compromises appears to turn it off. The only time they go back is when running with V-sync disabled is deemed unbearable due to the tearing you experience when frames coming from your GPU don't match the panel's refresh cycle.
As you might imagine, then, the visual impact of running with V-sync disabled is unmistakeable, though also largely affected by the game you're playing and the detail settings you use.
Take Crysis 3, for example. It's easy to really hammer your graphics subsystem using the taxing Very High preset. And because Crysis is a first-person shooter involving plenty of fast motion, the tears you see can be quite substantial. In the example above, output from FCAT is captured between two frames, and you see branches of the tree completely disjointed.
On the other hand, when we force V-sync off in Skyrim, the tearing isn't nearly as bad. Consider that our frame rate is insanely high, and that multiple frames are showing up on-screen per display scan. Thus, the amount of motion per frame is relatively low. There are still issues with playing Skyrim like this, so it's probably not the optimal configuration. But it just goes to show that even running with V-sync turned off yields a varying experience.
This is just a third example in Tomb Raider, where Lara's shoulder is pretty severely misaligned (also look at her hair and tank top strap). Incidentally, Tomb Raider is one of the only games in our suite that lets you choose between double- and triple-buffering if you use V-sync.
A final chart shows that running Metro: Last Light with G-Sync enabled at 144 Hz basically gives you the same performance as running the game with V-sync turned off. The part you can't see is that there is no tearing. Using the technology on a 60 Hz screen caps you out at 60 FPS, though there is no stuttering or input lag.
At any rate, for those of you (and us) who've spent countless hours watching the same benchmark sequences over and over, this is what we're used to. This is how we measure the absolute performance of graphics cards. So it can be a little jarring to watch the same passages with G-Sync turned on, yielding the fluidity of V-sync enabled without the tearing that accompanies V-sync turned off. Again, I wish it was something I could show you with a video clip, but I'm working on a way to host another event in Bakersfield to allow readers to try G-Sync out for themselves, blindly, to gather more dynamic reactions.
