Gadgetory

this is the most popular game in the world right now this look closely it's the most played so between the rubber banding issues texture pop-in static and dynamic mesh scaling issues and low frame rate overall the Xbox version of the game has a long way to go but it's still leading the charge in sales on PC alone just for perspective player unknowns battlegrounds has now sold 30 million copies with peak daily player count upwards of 2 million on PC today we're benchmarking pubsey on the Xbox including a detailed look at graphics behavior and overall performance this is done using GM's new in-house console benchmarking utilities that we've improved specifically as a result of this game before that this video is brought to you by thermal grizzly makers of the conductor hot liquid metal that we recently used to drop 20 degrees off of our temperatures Thermal Grizzly also makes traditional thermal compounds we use on top of the IHS like cryo not and hydronaut pastes learn more at the link below and for its help in improving our benchmarking utilities we really only have good things to say about pub G on the Xbox by which I mean it's it's so flawed and has such bad performance that it really helped us fine-tune the test resolution of our tools that we use so that's good I guess but we're going to break this video into a few sections we'll talk about graphics issues and shortcomings there some gameplay issues rubber banding things like that frame rate performance frame time performance on the Xbox one X which has frame time spikes sometimes upwards of 100 to 130 milliseconds when for 30 fps you should be closer to 33 milliseconds quite disparate performance so a brief explanation of the tools here if you want more details on our benchmarking tools or not talking about how all of it works it's all internal but we have some information top-level in the article linked in the description below for the very basics we have some new utilities in here now that will allow us to compare frame to frame where are the new pixels by creating contrast maps with black and white pixels we can see which pixels are new versus which are old which is important because there are no framerate our tools for the Xbox you have to make your own which we did and when recording framerate what we're basically doing is analyzing the Delta between the pixel values frame to frame and sometimes that Delta gets a little muddy because if you have things like aliasing or a game like pub G which just has a mixture of really odd attempts at optimization poor optimization overall poor performance overall and weird frame rates sometimes when you have all those problems we need debugging tools which is basically what the black and white mapping is so that we can see what's going on frames a frame and determine is that some kind of compression artifact ting from capture is it aliasing of some kind or is it genuinely just an issue that we need to talk about so more of that in the article but let's get into the basics here we're going to start with graphics and then we'll get into frame rate and frame time performance graphics are easy to start with from the instant the game was launched it was clear what we were in for the player character is displayed in the main menu screen but it takes a second or so after their appearance for clothing and textures to follow even once everything is settled down when new items are equipped like the belts or shoes they also take time to load in and then more time still for higher levels of detail to load in this texture pop-in is constant and every change of location the starting area of the plane wherever the player walks it temporarily turns the world into a muddy brown mess more than it's supposed to be even so poppin issues are a recurring theme for the Xbox one X version of pub G one oddity with the game is that poppin and LOD scaling often seem to draw completely new or incorrectly scaled objects in place of the lower LOD objects you can see this in the trees in the distance for example where sometimes it looks like there's a completely different mesh as we approach the tree and get closer to it this is also highly visible in a scene where we're crashing through fences with a buggy if we slow down time which isn't hard seeing as we're running below 30fps you'll see that it appears that the static mesh fence has been replaced with a dynamic particle match version of the fence and this would be done typically for Destruction as tracking dynamic particle meshes for the object across the entire game would likely cost too much in system resources so swapping out on impact could save on resource cost it appears as if the dynamic particle mesh version is scaled incorrectly though if we freeze frame you'll see that the replacement is larger than the static mesh the first one and that we then rubberband through it as the fence breaks apart texture pop-in is the most common issue the parachute is an easy first seen example dropping it to the map the parachute takes several seconds to load its textures fully and is exhibiting a normal map issue on the underside of the parachute this seems to happen mostly when the camera is at Glanton angles to the parachute revealing the normal map artifacting underneath you can see that in kind of the teal color the specific set of issues is particularly embarrassing for pub G seen as the parachute is an object that's only seen once in the game at the very beginning and then it's gone for the rest of the match texture pop-in issues can basically be seen everywhere we see them on cars like landing for example and seeing a car we see them on the ground at the start of the game you can see the texture resolution kind of loading and unloading as the player count changes bridges while skydiving also have texture poppin and some kind of LOD scaling in general that sometimes seems a bit odd buildings have their own popin issues tree alpha textures change and plenty more there are also problems with the environment poppin like when we observed parts of the world failing to draw which seems like some kind of kollene bug or something like that because you can pan around from inside the plane and see that there's just a bit of almost calling lag as it's drawing the world underneath where the plane was previously occluding that world this isn't really a big deal as it's not part of the game where it actually matter you're not competing yet but it just further illustrates the incomplete nature of pub G particularly on the Xbox one X where it's still lagging behind in version number it's not just textures of course playing on a fiber connection even we've noticed severe rubber banding that would indicate a server-side network or synchronization issue this is partly visible when driving through destructible objects as the static meshes again are swapped with particle meshes but it's even more visible in occasional combat scenarios in this example we punched a lagged out link that player to death and then attempted to end the garage with a motorcycle in it the kill counter updated as zero kills we rubber-banded in and out of the garage for a moment and then it ticked over to one kill although less significant because it's pre combat we also experienced rubber banding when approaching the picnic tables in the pre-loaded match area of the game it seemed to become more of an issue as players joined of course and similarly in our full 12 minute recording pass for one of the tests we also noticed the fluctuating player count impacted even the ground texture resolution with the ground textures higher LOD unloading at one point there are all kinds of other examples and oddities like this split second for example where the wall disappears in one of the houses which happens again a second later then there's moments before when the shadows were drawing in correctly while parachuting down and just generally graphics and visual issues that are either inconsistent with reality or just poor performance overall before starting with some of the on-screen graphics we have a new data presentation method to show you what's happening with an Xbox in terms of framerate so what's going to happen is on the left of the screen we will have a series of red and green bars we'll also have ones and zeros so it's a binary is this a new frame is it not a new frame and then next to that we have the frame count the number of frames as the benchmark progresses so every time you see a red bar what that means is that we have dropped a frame we can determine this by checking whether the pixels constitute a new frame or if there is no Delta between pixels which would demonstrate a repeated frame so when you see a lot of red in a row that means we are stuttering we're replaying the most recent unique new frame over and over all the way through the red until we can finally get enough system resources to render a new frame represented by green and if any of this confuses you just check the article below we'll try and explain it in more detail there getting into the performance discussion you can again find our testing overview in the article linked below but let's start out with a pre-match frame time chart just because it's a bit extreme it's one of the cases so for this frame time chart performance is disproportionately bad for the first few minutes of every match which is illustrated here as a reminder we're looking at frame times versus time and at 30fps that means our target frame time should be 33 point 3 milliseconds in other words the frame to frame render interval to create a new frame is about 33 milliseconds the clip starts around 33 0.3 milliseconds as expected or 30fps but has frequent spikes up to 100 milliseconds and 133 millisecond frame times that means in the worst case you're waiting a full 133 milliseconds for the next unique frame to appear in other words you're staring at a redrawn version of the same frame for 1/10 of a second while waiting for the next one that's a long time later in the chart we encounter frame times that equate a rough 26 to 27 FPS average never quite 30 shown by the rapid bouncing up and down of the frame time plot as it's between our measurement resolution and the actual intervals and here's some pre-match footage if there's any place in the game that should really overwhelm the xbox it is the pre-match free-for-all where 100 players are scrambling to fire as many weapons as possible in close proximity assets load in here players all leap around and jump on mass and then explosions go off everywhere every time a Molotov explodes the frame rate tanks with a good fireball in view the frame rate can dip as low as 12 FPS and seen an explosion during gameplay generally means that an enemy is within throwing distance which is exactly the wrong time for perforins to drop to 12 again 12 FPS watching smoke grenades go off in real combat had barely any impact whatsoever which means that this scene is more like a Molotov specific issue for whatever's happening when that fire is being drawn you'd have to find one outside of the pre-match and throw it to see the effect in real gameplay of course but after a countdown everyone is abruptly teleported out of the pre-match area and into a featureless brown tube which slowly resolves into a plane for gameplay performance doesn't matter until boots hit the ground but it's still hard to accept the amount of time it takes to load trees houses and even the plane that's right in front of you as you dive towards the ground at 209 km/h in our first skydiving run in a frame time plot here frame rate drops to around 15 fps between parachute deployment and hitting the ground that was also just about enough time for the parachutes texture to fully load right before it unloads and is never seen again the first couple minutes after landing are the most dangerous our average frame rate stayed around 25 fps in this time frame and that potentially makes parachuting into popular areas for competing resources more difficult after that frame rates generally stayed around 30 FPS if we stayed in one place or walked but dropped if we drove too quickly and entered new areas on a vehicle this also happened if we smashed fences as seen previously part of this is resource loading and unloading from adjacent cells as parts of the environment are dropped in and out of the players viewport but that might not seem too bad but again most console games do their utmost to never drop below their framerate goal for even a fraction of a second especially when that goal is only 30 FPS this frame time chart shows the FPS averaging out to 26 to 27 or in actual gameplay which is certainly subpar and sometimes jarring overall and then we do have occasional spikes a fixed frame rate is the whole point of a console here and we're losing that in pop G surprisingly one on one shooting punching and firefights had little impact on performance compared to the dips seen while driving a vehicle LOD distance is obnoxiously low on foliage and trees and there's clearly a visible line of grass that advances before the player network problems in general Genki nests were far more pressing problems than frames per second during our time with the game on more than one occasion it was possible to walk straight up to an enemy punched them and they were either frozen helpless by lag or incapable of shooting that closed right interesting control it really didn't seem to matter well I can do this rubber banding happened multiple times on our highest-scoring run where we really weren't trying we only encountered two other players both of whom chose to just drive away from us rather than deal with a fight eventually we hit a rock at 90 miles per hour and ran ourselves over with our own motorcycle but that can't really be planes on the game watching this death however frame by frame shows our character hitting the rock flipping upside down teleporting back towards the rock and a standing pose without bike and then equipping his gun sliding back towards the river again and a roughly dying the camera then pans up to an invisible hill and plummets again as the ragdoll falls through the bottom of the river on this note we've observed several general collision issues and ragdoll physics defects when people die in the game and just to make sure this wasn't some issue with the Xbox one axe trying to run at 4k we also manually force 1080p through the Xbox settings in this instance we ended up with the frame time plot as seen on screen now and frame rates still dropped below 30 fps in the early parts of the game we made do more precise comparisons in the future if we do a PC benchmark with this but for now we can say that running at 1080p definitely doesn't solve the problems we've described on the Xbox one X in its current state pub G on the Xbox one X is technically playable you can play it it's not really that enjoyable of an experience though there's a lot of ground here for the developers to improve things and it's hard to build an argument for the Xbox version versus the PC version especially with the issues the PC versions about one version number ahead at least at 1.0 plus now so that's a bit of a difference the only real upside is that on the Xbox you don't really encounter hackers which is somewhat of a rampant problem on PC but none of those things really matter because if we're looking at performance of the Xbox the big thing to take away here in the current version of the game is that other than all the kind of weird visual oddities texture pop-in meshes things like that collision there is objectively an inconsistent framerate that can make it hard to play at times or in the very least unenjoyable to a point of in worst cases hitting as low as 12 fps that's pretty damn bad and it's not just 12 FPS either because the idea of FPS or frames per second is is almost sort of silly because it inherently is an average as a metric frames per second but what's really happening when we say 12 FPS it doesn't mean you're just getting 12 frames per second it means that after you calculate out the huge frame time spikes it happens to mathematically come out to 12 but you're not getting the updates in an that as a player would actually be any measure of fluid because when we're talking about a frame time spike 133 milliseconds dragging our average down to 12 fps after it's calculated which is just a function of a thousand divided by the frame time basically you're talking about that ultimately what's happening is you're playing the game and at some point during play you hit something that causes the same frame to be redrawn at every interval for the next 1/10 of a second or slightly more and that is absolutely noticeable but some people call it stutter or hitching or things like that the point is you're seeing the frame for a tenth of a second which is a long time so yeah that's a that's Pope G on the Xbox it's all the problems with pub G on the PC when it was pretty one point no I think it saw some problems but not as bad compounded with being an Xbox so it was definitely fun to test though and it was worth improving our Xbox testing tools because of the problems well look at the PC version next those two subscribe for that as always you know help us out directly on patreon.com slash gamers and exes if you like this type of content and go to store like Garren's nexus net to pick up a shirt like this one or we did just restock our G on mugs so thank you for watching I'll see you all next time