Why does a computer run a game poorly, but is able to show a video of it perfectly fine
In: 0
To run a video it is a simple playback of what has been recorded previously. Playing a game the computer has to dedicate multiple operations to running individual things. So if 1 thing plays the video is the same thing that displays what is on screen it would be 1 to 1. But then you move your character. That is a 2nd operation. Then you move your camera, a 3rd operation and so on.
Basically you are comparing running a single movie to running 100s of movies at once.
If you are ten: Running a game requires you to have enough processor speed to keep up with the information that is constantly being entered. Your GPU needs to then actively render all the images associated with what you are currently doing. Every input into a game changes what needs to be rendered and the GPU and Processor need to keep up with those inputs. Where as watching a video of a game is playing back a recording, may take some initial processor usage but will quickly turn into being simple playback.
Rendering a game means doing all of the computations to build the scene in real time. For example if the game has complex physics simulations, it has to do the simulation before/during every single frame. Whereas a video is just a file which contains the scene which is already finished and all computations are done. Essentially just switching between images (but a lot more complicated than that)
By show a video, do you mean like a Youtube video of a let’s play, or an opening cinematic when you start the game?
When running the game, the computer needs to compute the appearance of all the objects that are on screen, along with lighting and particle effects, and any physics that might be applicable as well.
A video is much easier to run. You can sort of think of it as a series of screenshots shown in rapid succession with an audio track accompanying it. All of the processing to create the video was already completed when the file was produced, so your computer just needs to play it back.
The video is already rendered, meaning the computer knows what pixel has to be what color ahead of time.
For a video game, the computer has to render the image on the spot, taking into account where objects are, how far away they are what their texture is, any lighting, and if there are brother objects in front of it.
This is why GPUs are so important, because doing this math is very hard work, and GPUs are specialized to do that work. The same is true for crypto mining, 3D modeling, video rendering (ironically), and basically anything with vectors.
In the same way that you can hand in your homework to the teacher in 30 seconds but not finish the same question in an exam in 30 seconds.
Games are complicated. They have to constantly calculate where you are, where the enemy are, is anyone running into a wall and need to be stopped, is anyone hitting a bullet and need to take damage, is that patch of the sky still in view so the mountain need to be drawn on screen etc. A video is just exactly what you see. There is nothing behind the player, there is no calculation to see if that bullet hits (that doesn’t matter) etc.
If a game runs poorly on a computer, the most likely reason is because the game is requiring the computer to do more math all it once than it can handle. By contrast, playing a video requires no math and very little effort.
As an analogy, consider how long it would take you to do all of the problems on a math test versus read your answers out loud after you finish taking the test. Doing the math takes much more time and effort than simply reading off your answers.
Think of it as a question on a math test. Student A goes through the complicated math problem to arrive at the answer “42”. Student B has the answer key, so answers “42”.
They look the same to the teacher grading, but the student representing a live game had to do all the calculations to get to the answer. The student representing the video got to simply look up the answer from someone else’s hard work. It takes the same hard work to make a video, but it’s super easy to play a video once it’s been made.
When you watch a video, all you are doing is looking at a series of still images in order accompanied by an audio recording.
To actually play the game, your computer must perform all of the computations and mathematics required to run the game and, more importantly, to draw all of the graphics, including lighting, particle effects, etc.
To put it in more ELI5 terms, imagine making a flipbook animation and then flipping through it. Now imagine that instead of drawing the flipbook ahead of time, someone else was telling you what all the characters should be doing and you had to draw it *as* it was being played back.
Running a game is like this: Imagine I give you a bunch of paper and a pencil. I tell you draw me a rabbit and give me the picture when you are done. When you are done, I tell you move the rabbit a little bit in the next picture, or change the angle a little bit, or maybe both, or maybe neither, and you draw me the next picture. Maybe I tell you I don’t care so much about how the rabbit looks and you can draw faster (low settings in a game), or maybe I tell you I want a super good looking rabbit and it takes you a really long time per picture (high settings). Maybe you are a fast drawer and can get them to me pretty quickly (good hardware), and maybe you aren’t so fast, or even aren’t capable of drawing rabbits at all (bad hardware). Even if you are a great drawer AND I’m okay with a mediocre rabbit, it takes time to draw them.
Playing a video is like this: Imagine I give you a bunch of rabbit drawings already completed and sorted, and I tell you to give the next one in order to me, once per second. It’s much faster, because you don’t have to draw the rabbit or worry about what it’s doing in the next picture – that’s already been done for you. Even though you might be pretty horrendous at rabbit drawing, it doesn’t matter – somebody else already did it, all you gotta do is hand them over on time.
That’s the difference between running a game or playing a video of it, although a computer will go much more quickly than a person drawing or handing over completed drawings.
Hyper simple: A computer takes longer to draw a video game because it has to draw that image from scratch. A video is just showing you images that were already drawn.
More complex: A video is a series of images. Your computer doesn’t have to know what’s in those images, just what color each pixel is. It has the benefit of being able to load up each image in advance, each frame of the video, because those images aren’t going to change from one replay to the next.
When you’re playing a game, each image is being drawn by the computer for the first time as you play. It has to do this because you’re hitting buttons that change what the next image will look like. Instead of just knowing what pixel is what color, your computer has to know what every object on the screen looks like, what shape it is, how to color it in. It has to know where the light sources are, and how the light bounces off of objects to create highlights and shadows. The computer has to figure out all this stuff, and uses that information to draw the next frame. Then and only then does it know which pixel is supposed to be which color. Then it has to start all over again for the next frame.
Latest Answers