Elliot Wright
- SeerUK
- seeruk
- Europe/Stockholm
- Yes
- No
MC-161807
MCPE-53723
It seems like this issue has come up before as
MC-137537which has been marked as resolved.As illustrated by this video, when walking between chunks there is a lag spike. I've used Optifine in the video, but this does occur in vanilla 1.14.4, I just wasn't sure how to make it obvious it was happening in a video.
It then occurred to me that the pie chart would also include these spikes, the impact is just not as apparent as in the other video, regardless, here is a completely vanilla video. You can see, when crossing a chunk boundary there's a spike in "scheduled executables". This can be felt as a stutter.
I'm not sure if it's mainly noticeable because I'm playing on 144Hz monitors. You can't see it in the video, and I'm assuming that's because they've been recorded at either 30 or 60 FPS. Either way - this is noticeable.
Just to add to this - I also tested this in 1.13.2, again with and without Optifine, both versions perform perfectly, there's no stutter at all. Here's another video showing that.
It seems like this issue has come up before as
MC-137537(though, it looks far more severe there) which has been marked as resolved.As illustrated by this video, when walking between chunks there is a lag spike. I've used Optifine in the video, but this does occur in vanilla 1.14.4, I just wasn't sure how to make it obvious it was happening in a video.
It then occurred to me that the pie chart would also include these spikes, the impact is just not as apparent as in the other video, regardless, here is a completely vanilla video. You can see, when crossing a chunk boundary there's a spike in "scheduled executables". This can be felt as a stutter.
I'm not sure if it's mainly noticeable because I'm playing on 144Hz monitors. You can't see it in the video, and I'm assuming that's because they've been recorded at either 30 or 60 FPS. Either way - this is noticeable.
Just to add to this - I also tested this in 1.13.2, again with and without Optifine, both versions perform perfectly, there's no stutter at all. Here's another video showing that.
It seems like this issue has come up before as
MC-137537(though, it looks far more severe there) which has been marked as resolved.As illustrated by
thisvideo, when walking between chunks there is a lag spike.I've used Optifine in the video,but this does occur in vanilla 1.14.4, I just wasn't sure how to make it obvious it was happening in a video.It then occurred to me that the pie chart would also include these spikes, the impact is just not as apparent as in the other video, regardless, here is a completely vanilla video. You can see, when crossing a chunk boundary there's a spike in "scheduled executables". This can be felt as a stutter.
I'm not sure if it's mainly noticeable because I'm playing on 144Hz monitors. You can't see it in the video, and I'm assuming that's because they've been recorded at either 30 or 60 FPS. Either way - this is noticeable.
Just to add to this - I also tested this in 1.13.2, again with and without Optifine, both versions perform perfectly, there's no stutter at all. Here's another video showing that.
It seems like this issue has come up before as
MC-137537(though, it looks far more severe there) which has been marked as resolved.As illustrated by this vanilla video, when walking between chunks there is a lag spike. The pie chart include these spikes, the impact is somewhat difficult to see in these videos because it's in the format of a pie chart, not a bar chart for example, nevertheless, you can see, when crossing a chunk boundary there's a spike in "scheduled executables". This can be felt as a stutter.
I'm not sure if it's mainly noticeable because I'm playing on 144Hz monitors. You can't see it in the video, and I'm assuming that's because they've been recorded at either 30 or 60 FPS. Either way - this is noticeable.
Just to add to this - I also tested this in 1.13.2 and it performs perfectly, there's no stutter at all.
It seems like this issue has come up before as
MC-137537(though, it looks far more severe there) which has been marked as resolved.As illustrated by this vanilla video, when walking between chunks there is a lag spike. The pie chart include these spikes, the impact is somewhat difficult to see in these video
s because it's in the format of a pie chart, not a bar chart for example, nevertheless, you can see, when crossing a chunk boundary there's a spike in "scheduled executables". This can be felt as a stutter.I'm not sure if it's mainly noticeable because I'm playing on 144Hz monitors. You can't see it in the video, and I'm assuming that's because they've been recorded at either 30 or 60 FPS. Either way - this is noticeable.
Just to add to this - I also tested this in 1.13.2 and it performs perfectly, there's no stutter at all.
It seems like this issue has come up before as
MC-137537(though, it looks far more severe there) which has been marked as resolved.As illustrated by this vanilla video, when walking between chunks there is a lag spike. The debug pie chart include these spikes, but the impact is somewhat difficult to see in these video itself. Nevertheless, you can see, when crossing a chunk boundary there's a spike in "scheduled executables". This can be felt as a stutter.
I'm not sure if it's mainly noticeable because I'm playing on 144Hz monitors. You can't see it in the video, and I'm assuming that's because they've been recorded at either 30 or 60 FPS. Either way - this is noticeable.
Just to add to this - I also tested this in 1.13.2 and it performs perfectly, there's no stutter at all.
It seems like this issue has come up before as
MC-137537(though, it looks far more severe there) which has been marked as resolved.As illustrated by this vanilla video, when walking between chunks there is a lag spike. The debug pie chart include these spikes, but the impact is somewhat difficult to see in these video itself. Nevertheless, you can see, when crossing a chunk boundary there's a spike in "scheduled executables". This can be felt as a stutter.
I'm not sure if it's mainly noticeable because I'm playing on 144Hz monitors. You can't see it in the video, and I'm assuming that's because they've been recorded at either 30 or 60 FPS.
Either way - this is noticeable.Just to add to this - I also tested this in 1.13.2 and it performs perfectly, there's no stutter at all.
Hey [Mod] violine1101, I have also submitted evidence in a Vanilla installation - please see the video labelled "vanilla video" where the issue is highlighted using the debugging pie chart.
I'll update the description to remove those videos to avoid any issues with this.
Please let me know if I should resubmit this issue instead.
As illustrated in my issue (
MC-161807), this isn't limited to AMD GPUs, or CPUs - I'm experiencing this with an Intel CPU, and an NVIDIA GPU - it's not a very severe stutter I'm experiencing, but it's definitely noticeable.Here's a video of this: https://gfycat.com/mellowcolorfulgull
I'm also experiencing this. I've tested it on Realms, and I've been running the alpha dedicated server files on a VPS and it lags from there too in exactly the same way.
Here's a video from on my Realm: https://gfycat.com/scentedeminentethiopianwolf
Here's a video from on my VPS: https://gfycat.com/wickedarcticeyelashpitviper
My VPS is hosted in Nuremberg by Hetzner, I live in the UK. My latency to that server is around 37ms, pretty consistently. I also host Java edition (which I had shut down, only the bedrock server was running in the above video), on that same VPS and don't have any issues with that. It has access to 4 vCPUs, and 16GB of RAM and a local SSD, so definitely not a performance issue caused by the network, or the machine the server is running on there. Seems like there needs to be some sort of client-side smoothing of movement.
I'm playing on Windows 10, 1.12.1. The Bedrock server version is 1.12.1.1.
For me, this issue affects all versions of bedrock that I've played, that's iOS, Switch, and Windows 10. Switch was the first I've played more often and it was much more noticeable there I thought - but yeah, this isn't just limited to Windows. It looks like the animations are capped at 30FPS or less which contrasts heavily against the rest of the game running much smoother.
Make sure you've not accidentally enabled 5.1 audio or some similar surround sound configuration that doesn't match your speaker count.
Hey [Mod] Michael Wobst, I can no longer reproduce this issue. Though unfortunately I don't have the same world, I've tested this in a few different chunks in similar circumstances and it's been totally fine.
E: Apologies, look like me responding has re-opened this issue!
This is still an issue (at least in 1.14.x, 1.15.x and 1.16.x), and has been for a long time. If you watch Minecraft YouTubers (e.g. HermitCraft) then you'll hear them talk a lot about how when other players are online they can't AFK at their mob farms because they don't get any spawns. They play on a Vanilla server with some datapacks, but this affects all Minecraft servers, except those who use Paper with per-player-mob-spawns turned on.
On the one hand, you can start to think the Vanilla behaviour here makes sense. Your join a brand new, empty server (i.e. new world), mob cap is 70 for monsters, they spawn around you - all is well. You move (far) away, and another player joins, the mob cap doubles, and mobs are spread fairly evenly. The problems come in when you start lighting places up, or doing things like creating mob farms. If one of those 2 players lit up all of the caves and surface around them, and no mobs could spawn around them at all then the mob cap would remain the same globally, and the other poor player would get all of the mobs spawning around them instead.
Mob farms are where I personally have the biggest issue with this, if you play on an active multiplayer server it just makes them completely useless. You need to find a time when nobody is online at all other than you for them to work.
Clearly, given that Paper have made a working solution for this that feels right when you play it, it means that Vanilla could also do this.
I think the logic would go something like this: mob caps behaviour remains the same, but the game tries to spawn <mob cap> / <players online> mobs around each player. Mobs shouldn't spawn too close to each other unless they spawn in packs up to a harder limit (i.e. so if you had a single-layer mob farm then you wouldn't get 70 mobs spawning on that one layer).
For now, I'll continue to play on a Paper server, but to be honest I'd much rather play on a Vanilla or possibly Fabric server, as Spigot and Paper modify a LOT of stuff about the game, and I'd rather have Vanilla behaviour - except for this - this is just a biggie!
It seems I'm playing devils advocate here, with an increase in FPS:
FPS in 1.16.5: 238
FPS in the latest 1.17 release (1.17): 279
Your operating system: Windows 10 (20H2, 19042.1052)
Your CPU: AMD Ryzen 9 5900X
Your Graphics card and driver version, as reported by the game: NVIDIA GeForce RTX 3090 (466.63)
Your profiling zip file: 2021-07-04_19_59_13-PerformanceTest1_17-1_17.zip
Honestly, the main reason I've found this issue is because of the dramatic drop in shader performance with OptiFine. I did the above tests with the Vanilla game though of course! It looks like Vanilla performance has actually improved for me. I've seen quite a few people complaining about drops in performance on older, perhaps now unsupported hardware though. I guess for me there's nothing wrong with the base game, so I'll go ahead and hope that the final release of OptiFine improves the performance again significantly.
Wasn't able to log in last night, still not able to log in now with my main account. I've swapped account to a different account of mine and was able to log in successfully, and all of my friends were able to log in to my server too. I can't see any reason for me to not be able to log in otherwise - very strange!