Are you a gamer? So you probably know that latency is the sworn enemy, putting us one step behind our opponents and impacting the game experience just as much as the immersion. Besides pinging online multiplayer games, there are plenty of other things that go into the equation – now is the time to review them!
In video games, it is often necessary for us to find the right balance between image quality and performance in order to achieve the best possible gaming experience. The scale diverges quite widely according to the requirements and materials of each, but also according to the type of game targeted. As a general rule, responsiveness and extreme fluidity are preferred by competitive players, sometimes even at the cost of reducing image quality, or even sacrificing it. In the opposite case, we are satisfied with little in terms of performance when we seek to push the quality and other graphic details to the maximum, to take advantage of a high budget game, ray-tracing, or even a monitor or a 4K TV!
How to evaluate performance in a video game? There is still a lot of confusion about what fluidity, responsiveness, response time, input lag, etc. are, as well as the overriding factors that can affect the gaming experience, whether they are on your machine's side. game, peripherals (mouse, keyboard, controller), connection, or display. The time has come for us to take stock of this subject, with the key to a better understanding of all these elements, some advice to optimize your system and ultimately understand why the players in the gaming market are putting the package on performance. , without reliably and accurately portraying the entire picture.
Latency: what are we talking about?
When you hear the word latency, you usually think first of ping and network latency: the famous lag in the middle of a session on an online game. However, it is not this network latency that we want to talk about here, but rather system latency. In both cases, the term refers to the same concept, namely the time required for your system to generate a rendering and for your actions to be reflected in it, or the time for data transmission between your PC and the server in the case of network latency.
Expressed in milliseconds, latency is found everywhere in the hardware world and is particularly emphasized when it comes to offering components and peripherals to players. We are talking about latency in terms of RAM memory and video memory for the GPU, but also with regard to storage with hard disks and SSDs, monitors (and televisions), and of course peripherals.
Screens, peripherals, components: who is responsible for what?
System latency can be divided into three distinct parts: that related to peripherals, to machine components, and finally to the display. To summarize briefly, we are first constrained by the latency of the render pipeline, in other words, the time required for your machine to execute a job and display the result on the screen, and latency which directly refers to the user. with input lag, or input lag, which designates the time between user action and its repercussion on the screen.
Graphics card, processor and RAM are not the only ones responsible for system latency! The operating system and the storage medium are also two determining elements: on the software side, for example, additional latency can be observed with obsolete drivers, while a windowed display mode uses the window manager of the desktop. (Desktop Window Manager), which may add to the latency compared to full screen mode. Storage-related performance has been significantly reduced since the democratization of SSDs. While a mechanical hard drive running at 7200 rpm has an average latency of 18 ms, an SSD sits below the 100 microsecond mark, a wide gap that shows why SSD is so popular.
Many optimizations can be implemented to reduce system latency, but in the end, the observation is always the same when we look at the hardware: the less efficient and faster it is, the more it will be necessary to make concessions to reduce the latency. at least. This means that it will be better to lower the graphics quality of the game in favor of performance; Note also that GPU and CPU overclocking can be useful to gain a few milliseconds and reduce system latency. Logically, the latency generated by the system, as well as that inherent in the game rendering engine, is much more important than what we see with gaming peripherals and monitors. it is necessary to choose its components well,
We won't go into more detail about how a PC works and how the graphics card and processor work together to perform the tasks assigned to them, but take a closer look at what input lag is. and why is it so important. At the end of the article, we will take a step back on the subject to see if there are solutions available to us to reduce latency.
What you need to know about input lag
The first thing to know is stupid and nasty: the input lag cannot be eliminated or brought to zero, it is just possible to reduce it through certain optimizations, by improving its configuration, or by choosing suitable hardware.
Polling rate: Hertz by the thousands
Above, we deliberately distinguished between input lag and system latency, but these two elements are indeed two sides of the same coin. We hear a lot about input lag when it comes to gaming mice and monitors. In the case of a mouse, the time to record and transmit a click is in addition to the time it takes for the system to process that action. The same goes for the screen: it receives and processes the information provided by the GPU in order to set the pixels in motion on the screen. The lower the system latency, the faster these two devices will get their job done. However, they retain a time limit that is specific to them, this is where we talk about the polling rate for a mouse.
Expressed in Hertz, it indicates the polling rate (frequency) per second at which the mouse transmits its position to the CPU. The higher this rate, the shorter the input lag. Office mice do not generally go up to more than 125 Hz, while the standard for a gaming mouse is now set at 1000 Hz: they are polled no less than 1000 times per second or every millisecond!
Such a mouse requires additional CPU resources, something which could pose problems of overconsumption when this high speed appeared on the market at the time. Today, manufacturers tend to always go higher with a rate of 2000 Hz for some, see 8000 Hz for the Razer Viper! Needless to say, the processor will be all the more stressed with this interrogation rate of 1 / 8th millisecond. Something less obvious: a mouse at 1000 Hz is far from being the most impacting element in the system latency chain, in all cases the gain between a classic gaming mouse and that of Razer, for example, is less than one. millisecond, at best it will prevent micro-stuttering effects. Let's take the opportunity to engage with the question of the instructor: the vast majority of gaming monitors have a frame rate of 144 Hz, which corresponds to a display refreshes every 7 milliseconds. Things to keep in mind to better understand the marketing discourse surrounding “gaming” peripherals.
Framerate, refresh rate, response time, overdrive: performance at all costs
The avalanche of terms, technologies and other solutions can be overwhelming for first-timers. It's about knowing what it is all about and why it matters. The PC screen, like the television, is decisive for players: it is the last link in the chain, the one that is directly in front of our eyes.
The monitor is subject to input lag, its main component being the delay between the sending of an image by the graphics card and its display on the screen. It is nevertheless on the lag related to the monitor itself that we will focus on, starting with its famous response time.
What is the difference between input lag and response time?
Because it is necessary to make the difference between the input lag and the delay related to the response time. The first is perceived latency (delay between a click and its repercussion on the screen) which can affect our responsiveness. It can come from the processing time of the image sent by the PC or the console by the monitor, in particular televisions with a lot of electronics. The second is more visual latency that we see in the sense of fluidity and our perception of motion blur and related artifacts. More precisely, the response time is the delay required for the pixels to go from one state to another, for example to go from black to white, then from white to black.
Response time: a race for the millisecond
Manufacturers do not hesitate to display very short response times for their monitors, at 1 ms or even less. Without a specialized measurement tool, the consumer cannot verify the relevance of these claims, only tests and technical reviews allow to know more precisely what to expect for a given model.
So we have manufacturers looking to put the lowest possible number on product datasheets, but metrics that in reality are generally less flattering. The reason is simple: there is currently no measurement standard for manufacturers, some people, therefore evoke a GtG response time (gray-to-gray), in other words the time for a pixel to go from a shade of gray to another, where others display a Motion Picture Response Time (MPRT) value that measures the persistence of a pixel on the screen.
Neither of these two values is ideal or representative of real use, why?
The MPRT response time is closer to what the end user will be able to obtain, taking into account the refresh rate of the screen, as well as its technologies: the different types of backlighting of the LCD panels or the insertion of ‘black images. However, it does not make it possible to determine the performance of the panel without all of its technologies, which cannot always be deactivated.
The GtG response time is just as far from reality, it only considers transitions from gray to gray and not from black (pixel off) to white or vice versa. The displayed value is generally the most flattering: the one measured with overdrive, overdrive which certainly eliminates ghosting (the persistence of pixels which creates a ghost image), but often induces reverse ghosting. Clearly, not all pixel transitions occur at the same speed, while some transitions between different gray levels probably reach the figure indicated by the manufacturer, others are generally slower.
To summarize, let's say that we can trust the figures announced by the brand knowing what to expect: an imprecise value, which lacks representativeness. If the gap is narrowing more and more as technologies evolve, the choice of a monitor should always fall first on its type of panel: IPS, VA, TN, or even OLED. There are indeed strong disparities between these panels, in terms of contrast, brightness, viewing angles, but also in terms of speed and pixel transition.
The interest in overdrive
It is difficult to close this paragraph on response time without mentioning the overdrive, now present on a large majority of monitors labeled gaming. It makes it possible to reduce the response time and therefore the afterglow of a screen thanks to a system that applies an electrical surge to the liquid crystals in order to make them change state more quickly. Accessible within the OSD of monitors equipped with it, it can generally be adjusted on several levels.
However, the management of the overdrive does not reach the same quality depending on the brands and monitors. The increase in electrical voltage sometimes involves an overshoot, this is a situation where the pixel exceeds the desired state and ends up in a position that is too far away. This phenomenon results in a more or less significant reverse ghosting depending on the applied overvoltage, resulting in artifacts that are sometimes more annoying than the ghosting itself.
The important thing to remember here is that the higher the overdrive is adjusted, the more likely you are to see reverse ghosting. In most cases it is better to set the overdrive a little lower to avoid any artifacts.
TN, OLED, IPS, VA: what is the fastest?
TN panels are still the fastest in terms of response time without overdrive. IPS and VA panels manage to catch up by applying an overdrive that is sometimes too aggressive. OLED panels have an almost instantaneous response time thanks to self-emitting pixels, but monitors of this type are still very rare and expensive, for which many players, consoles as PC, are turning to OLED televisions.
A word about televisions
Regarding response time, televisions are subject to the same problems as PC monitors and we, therefore, find the same weaknesses and qualities depending on the type of tiles used. But when it comes to input lag, that's another story. The input lag on television can sometimes be ten times greater than with a monitor, for the simple reason that it applies many treatments to the image such as motion compensation and scaling engines, and all other image enhancement features. In addition to this problem, that of the refresh rate also intervenes, with a majority of televisions which are satisfied with 60 or 50 Hz.
The processors that power televisions provide important arguments for viewing video content over a simple monitor, but many console gamers continue to favor the monitor in order to achieve a better ratio between performance, price and diagonal. The thing is less and less true, however, so much so that even some PC gamers are now favoring the television.
The response of the big names in television to attract more gamers first consisted of the deployment of a game mode. This disables all or part of the image processing, which has the result of drastically dropping the input lag.
The arrival of HDMI 2.1 and its 48 Gbps bandwidth has also changed the situation: this new standard allows you to enjoy 4K definition at 120 Hz and VRR, or even ALLM (Auto Low Latency Mode) which detects consoles and PC to automatically switch to game mode. There is always a certain gap between television and monitor in the context of competitive use, but it is clear that a lot of efforts have been made, as much on the platform side with the PlayStation 5 and Xbox Series X | S, only the television itself.
Unfortunately, many inexpensive TVs lack these arguments and provide a poor gaming experience. We will not hide that the most important problem today remains the price: the best gaming televisions, OLED or not, generally require a significant investment!
Refresh rate and number of frames per second
We're not done with latency and input lag; last elements to enter and undoubtedly the most important: the refresh rate, and the number of frames per second, which we especially know under the abbreviation FPS for Frame Per Seconds.
In absolute terms, things seem very simple: a screen with a frequency of 60 Hz can display a maximum of 60 frames per second. It's true, but there are subtleties to be aware of.
Like the mouse polling rate, the screen refresh rate is expressed in Hertz and results in a latency inherent in this rate. For example, a 60 Hz screen has a latency of 16.66 ms, a figure that drops to 6.94 ms for a 144 Hz panel, 4.16 ms for 240 Hz, or even 2.77 ms for a 360 screen. Hz.
The higher the Hertz value, the more capable the monitor is of providing a smooth display. Is it still necessary that the PC or the console connected to the monitor manages to keep pace, hence the interest in choosing its screen according to its material and vice versa. To take full advantage of the 240 Hz refresh rate of a gaming monitor, the graphics card must be able to calculate and transmit 240 frames per second. Suffice to say that depending on the games and the defined graphics quality, most graphics cards will struggle to reach this rate. So, why choose such a monitor?
The more hertz the better?
Quite simply, because by adjusting the monitor's refresh rate, the input lag associated with it is reduced. This means that even if your game isn't running over 100 FPS, you get a shorter input lag with a monitor that has higher capabilities. A high refresh rate, called HFR (High Frame Rate), is therefore not only of interest to be able to push the number of frames per second even further, but also to reduce the input lag. We can see it with the latest consoles from Sony and Microsoft: while the vast majority of games do not run at more than 60 fps, we still benefit from an HDMI output at 120 Hz.
Felt and visual latency come together here and this is the reason why refresh rate and the number of frames per second are very important for gamers. The point is that the refresh rate of a screen is a fixed value, while the number of frames per second varies depending on many factors: your configuration, on the one hand, the graphics that the GPU needs to calculate in real time on the other (textures, shadows, details, etc.). In other words: you just have to move the view of your character from left to right and vice versa so that the number of images varies more or less strongly. Be careful, however, not to confuse this completely normal phenomenon with drops in FPS, which are more the result of overload at the CPU / GPU level or other optimization problems.
This variation between the refresh rate and the number of frames per second can be problematic, both for visual comfort and gameplay. The variable and irregular volume of data processed by the GPU can produce artifacts due to the steady, steady scan of the monitor.
The best-known example is undoubtedly tearing, an image tearing that results from the gap between the GPU calculation time and the refresh rate. Several solutions have been deployed to counter these undesirable phenomena, one thinks first of the V-Sync which allows the graphics card to bait its calculations according to the screen. If it prevents tearing, it nevertheless creates the stuttering or stammering, with a purpose ultimately be far from ideal.
The VRR or variable frame rates, with G-Sync for Nvidia and FreeSync for AMD, changed everything: it is now the screen that adapts in real time to the number of frames per second issued by the GPU, a feature that allows you to obtain a fluid display in all circumstances and that can be found even on the latest home consoles!
Solutions to reduce latency?
As we have seen through this article, it is possible to reduce the perceived latency (input lag) and visual latency by choosing the right hardware to play: the right CPU / GPU pair, an appropriate screen definition, and peripherals designed for video games.
On the other hand, many solutions have emerged and continue to arrive on the market. We notably mentioned the increase in the refresh rate of monitors, which goes hand in hand with the rise of GPUs, but also technologies such as G-Sync and FreeSync. Besides that, we can also mention the ALLM with HDMI 2.1 for those who play on televisions.
On the peripheral side, keyboards and controllers have always had a greater lag than gaming mice, something which tends to be erased, for example with the DLI “Dynamic Latency Input” for the controllers which accompany the Xbox Series X and which benefit from ‘an input lag of only 2 ms, against 8 ms previously.
Finally, Nvidia is particularly active when it comes to developing technologies targeting gamers. We have mentioned them many times on Frandroid, for example with DLSS, an artificial intelligence rendering upscaling technology that makes it possible to enjoy ray-tracing much more comfortably, something AMD is also working on with its FidelityFX SR.
The latest technology to directly reduce perceived latency, however, is Nvidia Reflex. Accessible with RTX 3000 cards, Reflex needs to be implemented directly in the game by the developers; naturally, the optimizations it benefits from are therefore currently supported by only a handful of games, mostly competitive titles like Apex Legends, Fortnite, Valorant, Overwatch and other Call Of Duty: Black Ops – Cold War.
Nvidia has also thought of those who are not equipped with the latest generation GPU, cards affected by a severe shortage when they are released. It is thus possible to activate the Ultra Low Latency option from the Nvidia control panel, an option similar to Reflex, but which influences the render queue from the driver and not directly from the game.
To reduce latency, both perceived and visual, here are some avenues to explore:
- Choose devices with a polling rate of 1000Hz, or higher if you find this relevant, but keep in mind that the device is very rarely the weak link in the latency chain.
- launch your games in full screen mode in order to bypass the Windows window manager
- Activate the features that allow you to play on perceived latency, Ultra Low Latency and Nvidia Reflex for example, and on visual latency: G-Sync and FreeSync. Enabling game mode within Windows 10 can also help.
- Choose a monitor with a high refresh rate that has been tested in tests for response time. The higher the refresh rate, the smoother the game will feel, even if it does not exceed 60 fps.
Source: Gaming – Frandroid by www.frandroid.com.
The article has been translated based on the content of Gaming – Frandroid by www.frandroid.com.
We stand by the rights of the original author of the post, no matter what. We always respect and prioritize the copyright of the content and always include the original link of the source article. If the author of the original article has any issue with it, just leave a report below, we’ll edit it or delete it. Whatever it takes. We will make it right as quickly as possible to protect the rights of the author.
Thank you very much! Best regards!