Skip to content

How Variable Refresh Rate Works: V-Sync, G-Sync, and FreeSync Explained

Official Doom Eternal gameplay screenshot from the Steam store listing.

Screen tearing and stutter are two names for the same underlying problem: your graphics card produces frames at a rate that changes from moment to moment, while your monitor refreshes at a fixed rate. When the two fall out of step, the image visibly breaks. Variable refresh rate, or VRR, fixes this by letting the display refresh exactly when a new frame is ready instead of on a rigid timer. The technology is sold under several different names, including V-Sync, G-Sync, FreeSync, Adaptive-Sync, and HDMI VRR, and the differences between those names matter more than most marketing copy lets on.

This guide explains how the pieces fit together, using the official specifications and documentation from VESA, AMD, NVIDIA, and the HDMI Forum. By the end you should be able to read a monitor listing, know which flavor of VRR it actually supports, and tell whether it will work with your GPU and your game.

Why a fixed refresh rate breaks moving images

A monitor refreshes its picture on a fixed schedule. A 60 Hz panel redraws the screen 60 times per second, and a 144 Hz panel does it 144 times. The graphics card, by contrast, does not render on a schedule. The time it takes to draw a single frame depends on how complex the scene is, how powerful the GPU is, the resolution, and the target frame rate. The HDMI Forum uses exactly that list when it explains why frame times vary.

When the GPU finishes a frame off schedule, the display has two bad options. It can repeat the frame it already showed, which reads as a stutter or judder. Or it can start drawing the next frame while it is still half-rendered, which is the horizontal tear you see as a line splitting the picture. The HDMI Forum’s Variable Refresh Rate documentation states this directly: if the GPU does not finish the next frame in time, the source must either repeat the current frame or display the partially rendered next frame, and that is what causes judder and tearing.

VESA described the same mechanism when it announced Adaptive-Sync in 2014. Computer monitors normally refresh at a fixed frame rate, but a game’s frame rate varies with the complexity of what is being rendered. When the display’s refresh rate and the GPU’s render rate are not synchronized, the user sees tearing or stuttering. The entire problem reduces to that one mismatch.

V-Sync: the old fix that trades tearing for lag

V-Sync is the traditional answer. It forces the graphics card to wait for the display’s next refresh boundary before handing over a finished frame, so the panel never shows a partially drawn frame and tearing disappears. The cost is that the GPU now waits. If a frame is ready early, it sits idle until the display is ready to accept it, which adds input lag between your input and the result on screen. If the GPU misses a refresh interval entirely, the display repeats the previous frame and the effective frame rate drops in steps, a 60 Hz display falling to 30 frames per second the moment the GPU cannot keep up.

This is why competitive players have historically run games with V-Sync off. The tearing is a visual artifact, but the added lag is a performance problem. V-Sync eliminates tearing by constraining the GPU rather than by changing how the display refreshes. Variable refresh rate flips that approach around.

Adaptive-Sync: the standard that made VRR possible

The fix is to make the display wait for the GPU instead of the other way around. VESA added this to the DisplayPort standard in May 2014 under the name Adaptive-Sync, as an update to DisplayPort 1.2a. With Adaptive-Sync, the display dynamically matches the GPU’s rendering rate on a frame-by-frame basis. The monitor is no longer on a fixed timer; it refreshes when a frame is actually ready.

The idea was not new to DisplayPort. VESA notes that Adaptive-Sync had been part of the embedded DisplayPort (eDP) specification, which connects a laptop’s internal panel to its GPU, since 2009. What the 2014 announcement did was bring the same mechanism to external displays over a standard connector, which is what turned it into a consumer feature. VESA also publishes a compliance test specification, and displays that pass can carry an official Adaptive-Sync logo.

AMD FreeSync: the open ecosystem

AMD built its FreeSync brand on top of the Adaptive-Sync standard. The company’s pitch is direct: no tearing, no stuttering, just gaming, with fluid, artifact-free visuals at virtually any frame rate. AMD says FreeSync works over both HDMI and DisplayPort and that its ecosystem covers more than 4,000 certified displays.

FreeSync is split into three tiers, and the tier names encode real differences in refresh rate and HDR support. AMD’s specification table lists tear-free experience, low latency, and low framerate compensation, or LFC, across all three tiers. The difference is in the refresh rate floor. Baseline FreeSync requires a maximum refresh rate of at least 144 Hz on displays narrower than 3440 pixels. FreeSync Premium raises that floor to 200 Hz for those displays and to 120 Hz for ultrawide displays at or above 3440 pixels, while FreeSync Premium Pro keeps the same refresh requirements and adds HDR. AMD also says every FreeSync display goes through a certification process that tests refresh rate, latency, and HDR color and luminance accuracy.

The useful detail is LFC. When a game’s frame rate drops below the display’s minimum refresh rate, ordinary variable refresh rate stops helping, because the monitor cannot refresh slowly enough to keep matching. LFC compensates by repeating frames so the panel stays inside a range it can actually refresh at, which keeps the experience smooth even when performance falls. That is why LFC matters for demanding games that dip below a monitor’s lower VRR bound.

NVIDIA G-Sync: hardware modules and validation tiers

NVIDIA took a different route with G-Sync, splitting its offerings into three tiers. G-SYNC ULTIMATE uses the top NVIDIA G-SYNC processors and adds HDR, high contrast, and low-latency gameplay on top of variable refresh rate. Standard G-SYNC also uses a dedicated NVIDIA processor and promises no tearing, stuttering, or input lag, with a full variable refresh rate range and variable overdrive. G-SYNC Compatible is the notable one: it does not use an NVIDIA processor at all. Instead, NVIDIA validates a monitor that supports standard variable refresh rate and certifies that it delivers tear-free, stutter-free gaming after a test suite the company describes as more than 300 tests.

The practical difference is that native G-Sync monitors carry an NVIDIA module that manages the full refresh range and overdrive tuning, while G-Sync Compatible monitors rely on the standard variable-refresh-rate mechanism and pass NVIDIA’s validation. A G-Sync Compatible display is, in effect, a VRR monitor that NVIDIA has tested against its own GPUs. NVIDIA also sells G-Sync Pulsar, which combines variable refresh rate with high-frequency backlight strobing to multiply motion clarity, and Reflex Analyzer, which measures system latency separately from refresh rate.

HDMI VRR: the same idea for consoles and TVs

Consoles and televisions needed their own version of the standard because most of them use HDMI rather than DisplayPort. The HDMI Forum added Variable Refresh Rate to the HDMI specification as one of its enhanced gaming features, alongside Auto Low Latency Mode and Quick Frame Transport. The HDMI Forum’s description is the same core idea: a gaming source such as a console or computer delivers frames as fast as it can, which is often slower than the display’s normal static refresh rate, and VRR reduces or eliminates lag, judder, and frame tearing.

The HDMI Forum’s explanation of the mechanism is worth reading on its own, because it states the frame-timing problem clearly. A GPU needs different amounts of time for each frame, and that time depends on scene complexity, GPU power, resolution, and frame rate. When the GPU is taxed and misses the deadline for the next frame, the source either repeats the current frame or shows the partially rendered one. VRR waits until the next frame is actually ready to send it, which is what produces the smoother result.

What VRR does and does not fix

VRR removes the tearing and the judder that come from a fixed-refresh display meeting a variable frame rate. It does not raise your frame rate, and it does not make a weak GPU suddenly capable. If your frame rate drops below the display’s minimum refresh rate, VRR either stops helping or relies on LFC to fill the gap, and LFC smooths motion but does not create new frames of actual simulation.

VRR is also separate from the other performance technologies that get bundled into the same marketing pages. Upscaling and frame generation, sold as DLSS, FSR, and XeSS, change how many frames or pixels the GPU produces; VRR changes how the display presents those frames. Latency reduction technologies such as NVIDIA Reflex or AMD Anti-Lag are a third, separate layer that shortens the input-to-photon path. A monitor with G-Sync or FreeSync does not automatically include any of these, and a GPU with frame generation does not automatically need VRR. They are complementary pieces that solve different problems.

What to check before you buy

The first number to look for is the refresh rate range. VRR works inside a window, often stated as something like 48 to 144 Hz. Below the minimum, you are back to fixed-refresh behavior unless the display supports LFC. The wider the range, the more of your actual gameplay is covered.

The second thing is the connector and the GPU pairing. FreeSync works over DisplayPort and HDMI, and modern NVIDIA GPUs can drive many FreeSync monitors through G-Sync Compatible, but you should confirm the specific monitor is on the validated list rather than assume it. Native G-Sync monitors use a hardware module and are built for NVIDIA GPUs. On a console, the relevant standard is HDMI VRR on a display that supports it. The rule is that the GPU, the cable, and the display all have to speak the same VRR protocol for any of this to work.

Finally, do not assume VRR is on by default. It usually has to be enabled in the GPU control panel and, on some displays, in the monitor’s own on-screen menu. A monitor that supports VRR but ships with it disabled will tear exactly like a fixed-refresh panel until you turn it on.

Variable refresh rate is not one feature with many names. It is one mechanism, a display that refreshes when a frame is ready, implemented by different standards and branded differently by different companies. Knowing which layer a monitor actually supports tells you what you are buying, and what you are not.

Sources