Graphical settings in games often come with an inadequate explanation, and it can be hard to know what’s best for your setup. Here I’ll attempt to explain various graphical settings, I’ll describe their effect, how they work and what sort of impact they’ll have on performance.
This may be a bit obvious, I think it’s important to have a solid foundation when you’re building a tower of graphical knowledge.
A ‘pixel’ refers both to the physical, colour changing dot that populates your screen in the millions and the computational representation of that pixel.
A display resolution is the dimensions of a display in pixels, a game’s resolution is the dimensions of the game’s window. In Fullscreen mode a game will stretch to fit the physical screen, which is rather inexpensive. Games can run much faster rendered at a smaller resolution and then stretched back to your screen’s resolution, although the decrease in quality can be rather apparent.
Resolutions are written “width x height“, As an example, the most common resolution is 1920×1080, which is 1920 pixels wide and 1080 high.
An aspect ratio is the ratio between the width and height in a resolution. For example, a resolution of 1920×1080 has an aspect ratio of 16:9, 1280×1024 5:4 and 1440×900 16:10
Any resolution with an aspect of 16:9 will be the same shape as a 1920×1080 display.
Anti-aliasing involves combining neighbouring pixels to reduce jaggies.
Here is a list of Anti-aliasing types and their relevant performance impact:
Off: No impact
TXAA + SSAA: 10
Computers update the pixels on a screen one at a time, typically as fast as 60 times a second to keep up with what could be on the screen(On sensible platforms, at least…)
V-it stops new information from being sent to the screen while the screen displays it. Because (most, if not all) software updates the screen from the top and then to the bottom(vertically), changes made to the visual content during a frame will only affect the part of that frame that hasn’t been drawn yet.
As an example. one place this often manifests when in a first person game, there is a horizontal pole and the player looks up sharply. By the time the monitor has rendered the whole pole, it’s now actually a few pixels lower than it used to be, so it’ll render parts of that line again. And if that the pole becomes 1 pixel lower half-way through rendering a line, only the second half of that line will be rendered at all, leaving an unsightly jagged line.
Vsync basically means that the visual content won’t change while the monitor is being updated. One way to do this is called “Double buffering”, where basically the visual data is written to a separate location(A buffer), and the monitor takes the data from that buffer and displays it as soon as it’s finished.
Small problem is, Double buffering adds latency because while the buffer is being copied to the screen, any attempts to draw to it would defeat the purpose and produce screen artefacts. Another method, Triple buffering, is effectively double buffering with 1 extra buffer. While one buffer is busy being read from, the other will be written to, so it doesn’t need to wait for the first buffer to finish being read. The buffers will switch roles and the Double buffering latency will no longer be an issue.
Science! I’ll update this with more graphical features in the future because my articles are like Hogwarts.