Roll20 uses cookies to improve your experience on our site. Cookies enable you to enjoy certain features, social sharing functionality, and tailor message and display ads to your interests on our site and others. They also help us understand how our site is being used. By continuing to use our site, you consent to our use of cookies. Update your cookie preferences .
×
Create a free account

High Processor Usage(Chrome)

Same here. Please resolve it. I'm a subscriber but if I can't play a game...
the problem still exist
It's still occuring here. Made tonight's game almost unbearable. It only spike the CPU if Roll20 is the active tab, drops down otherwise. turning off video helps a little but still runs CPU high
Ok thanks to everyone for letting us know. We're probably going to have to take a step back and look at more wholistic options. I'm not sure exactly what those might be yet, but we'll keep you posted. I really appreciate everyone's help with this. It lets me know you all truly care and want it to get better.
I hope you'll find a solution. Thank you ^^.
exactly as Matt J, says: It only spike the CPU if Roll20 is the active tab, drops down otherwise... i cant play anymore, we thinking to go to foundry with my groups
Still struggling on my end.  last week we had 5 players and my feed constantly had video/voice lag and dice rolls from D&DB would take many seconds to pop up on my screen.  Other players got it instantly.  Got to find different computer or something as my play is getting worse.
1685557549
keithcurtis
Forum Champion
Marketplace Creator
API Scripter
Do you have the same dice lag issues without Beyond20?
keithcurtis said: Do you have the same dice lag issues without Beyond20? Hello keithcurtis, I just tested and R20 dice are instant from menu or command line.  D&DB is a 12 second lag.
Oh boy, don't know if this actually affected it or by how much but, I opted into the new UI because I was all *caution to the wind, what could happen* and my CPU/GPU which were both running a constant 25-30% usage while roll20 was open jumped to 50-55%.
1685758788
keithcurtis
Forum Champion
Marketplace Creator
API Scripter
dgrossjr said: keithcurtis said: Do you have the same dice lag issues without Beyond20? Hello keithcurtis, I just tested and R20 dice are instant from menu or command line.  D&DB is a 12 second lag. I haven't seen recent reports of Beyond20 lag, but I don't follow the extension, and it's not often reported here. Do they have a support discord or similar?
I used to use beyond20 but since I carried out my full computer clean and reboot, in an effort to resolve this CPU/GPU usage. I haven't used it. since and t he issue still persists, relentlessly. Trying to type some simple text onto a map is a real drain and ridiculously slow. Having to have a separate machine (my work laptop) to purely run Roll20 is not a viable solution (and not one available to all).
Roll20 has always been a bit of a dog, but this problem has made my last two play sessions borderline not worth it.   Last night was the worst ever.  I have a recent Mac with 32GB ram,  reboot before starting the play session,  turn off video (why is video so heavy weight), minimal tabs open.   Still can barely move tokens around, Roll20 is almost unusable.  Happens in Chrome, Firefox AND Safari.  At this point I'm seriously considering dropping my pro account and looking for a new product.  I keep seeing new features but no performance improvements. Friday afternoon I was using my Mac to debug a performance issue on a high through put web service with thousands of threads and Friday night I can't use Roll20 without shutting down down literally everything on my Mac except Chrome.
Andrew Searles said: Ok thanks to everyone for letting us know. We're probably going to have to take a step back and look at more wholistic options. I'm not sure exactly what those might be yet, but we'll keep you posted. I really appreciate everyone's help with this. It lets me know you all truly care and want it to get better. Just want to chime in here ... last night I almost cancelled the session.  Last session I spent 45 minutes trying to get things running so I was prepared this time (reboot, everything else turned off, no other tabs open).  We still ended up using theatre of the mind a lot and video, a big part of the VTT experience, has to be turned off. Things just keep getting worse, month after month after month, and last night was the worst ever.    I was literally angry when there was yet another user interface update and performance was even worse, why am I angry on game night? Roll20 is running out of time.  We seriously discussed just using D&D Beyond and Google Meets last night.
It's still pretty bad for me and a few of my players. :/
1685935694
keithcurtis
Forum Champion
Marketplace Creator
API Scripter
Robert O. said: Andrew Searles said: Ok thanks to everyone for letting us know. We're probably going to have to take a step back and look at more wholistic options. I'm not sure exactly what those might be yet, but we'll keep you posted. I really appreciate everyone's help with this. It lets me know you all truly care and want it to get better. Just want to chime in here ... last night I almost cancelled the session.  Last session I spent 45 minutes trying to get things running so I was prepared this time (reboot, everything else turned off, no other tabs open).  We still ended up using theatre of the mind a lot and video, a big part of the VTT experience, has to be turned off. Things just keep getting worse, month after month after month, and last night was the worst ever.    I was literally angry when there was yet another user interface update and performance was even worse, why am I angry on game night? Roll20 is running out of time.  We seriously discussed just using D&D Beyond and Google Meets last night. Robert O, if you would like, I would volunteer to join your game for the purposes of seeing if I can replicate this behavior. I am also on Mac/Chrome, with less RAM. If I experience the lag you are seeing , that might help confine the error to something about the game itself (not saying this is a local fault, just that the game instance itself might have some problem). Before you decide one way or another, do all of your players experience the same lag, or is this confined to your local experience? If it's the latter, there's probably not much point. If you do decide to, please PM the invite and you can boot me after the test. I'd like to help. I've never experienced the lag or the CPU drag many others have reported, but I know it's got to be exasperating.
keithcurtis said: Robert O. said: Andrew Searles said: Ok thanks to everyone for letting us know. We're probably going to have to take a step back and look at more wholistic options. I'm not sure exactly what those might be yet, but we'll keep you posted. I really appreciate everyone's help with this. It lets me know you all truly care and want it to get better. Just want to chime in here ... last night I almost cancelled the session.  Last session I spent 45 minutes trying to get things running so I was prepared this time (reboot, everything else turned off, no other tabs open).  We still ended up using theatre of the mind a lot and video, a big part of the VTT experience, has to be turned off. Things just keep getting worse, month after month after month, and last night was the worst ever.    I was literally angry when there was yet another user interface update and performance was even worse, why am I angry on game night? Roll20 is running out of time.  We seriously discussed just using D&D Beyond and Google Meets last night. Robert O, if you would like, I would volunteer to join your game for the purposes of seeing if I can replicate this behavior. I am also on Mac/Chrome, with less RAM. If I experience the lag you are seeing , that might help confine the error to something about the game itself (not saying this is a local fault, just that the game instance itself might have some problem). Before you decide one way or another, do all of your players experience the same lag, or is this confined to your local experience? If it's the latter, there's probably not much point. If you do decide to, please PM the invite and you can boot me after the test. I'd like to help. I've never experienced the lag or the CPU drag many others have reported, but I know it's got to be exasperating. One easy way to see how much lag is coming from where it to make a new empty game and see how much the lag is with just a few random tokens and map assets. Sadly for me it lags abit regardless if it's mostly empty or not. Of course there's abit more load time when starting a larger game with more sheets and larger maps, that's to be expected, but once it's all loaded it about the same after that.
1686071247

Edited 1686076090
@KeithCurtis or R20 Dev team, I'm open to share/test with you. However, joining game on your end may likely show little as compared to jumping on my laptop.  I can't figure it out, but I have 3 laptops, all very similar.  cleared the GPU processor.  it is maxed too.  Have all drivers confirmed, directx, etc... etc...   R20 works fine on work PC, but not on the 2 that I use for game day.  Still having major issue and gaming is almost impossible.  btw... stats on memory...  this is just me loaded.  it gets worse with other folks in game.
1686071576
keithcurtis
Forum Champion
Marketplace Creator
API Scripter
I was afraid that might be the case. :(
I have the same problem with CPU/GPU utilization as some here. For me, much of the load occurs when I currently have any map selected, regardless of whether it is empty or not. If I select a map with the size of 0x0 fields, the unnatural load on my device disappears completely, maybe this will help someone else. Not that maps aren't an essential part of the game, however I always switch to my "0x0 pause screen" when I can to lighten the load on my laptop. For my players with the same problem, this "solution" also seems to help with performance.
I recently got a new laptop (great deal came along on a ROG Strix, it's my birthday, etc)... and yeah. Problems instantly fixed.  The Acer Spin 5 comes with a integrated Intel UHD Graphics 620 GPU, while the ROG comes with a GeForce RTX 4060.  I feel like this might be a case of either the base requirements for Roll20 creeping up (that Acer was fine for a long, LONG time), or it not getting along with certain graphics cards. 
So a review on my side, T470, new Win11 build with drivers, firmware, etc... all up to speed.  Last night I played with virtual memory settings going a bit old school and saw no difference at all.  I even turned off all the Visual Effects to Adjust for best performance .  I have 24GB on this and ran many different settings all yielded similar results.  Very frustrating that R20 and the Intel Integrated HD5xx or HD6xx chips are not working well.  In Line 5 I have opened R20 and launched video and CPU/GPU goes 90+.  This is of course with no other players connected. it goes higher with each additional and with normal 3 guy game, it peaks at 100.  Line 7 shows what happens once I load D&DB and then switch back to R20.  CPU/GPU peak out at/near 100 most of time.  This was tested in Chrome and I have memory savor turned on so when tab isn't active it frees memory.
dgrossjr said:  R20 works fine on work PC, but not on the 2 that I use for game day.  Still having major issue and gaming is almost impossible.  btw... stats on memory...  this is just me loaded.  it gets worse with other folks in game Number of players is the key for me.  It works very well with just me or even one or two players.  When I have 5 players it grinds to a halt.
I have been away for some time but I did see my system now uses 80-100% processor for Role 20. Which is 6 cores / 12 thread is quite a bit of grunt. My system was lagin out with poor UI and menu performace so a bit of concern. It is like all the graphic processing is being done on the main processor and not the Graphic card or something crasy like that. It did not have this issue in january the last time i played.
Aspin B. said: It is like all the graphic processing is being done on the main processor and not the Graphic card or something crasy like that. TBH, a few of us have this suspicion as well. 
Testing today in an attempt to see if we can play.  Running Chrome with a new profile with zero extensions, no bookmarks etc., nothing else running, archived some old maps, cleaned up as much as possible.  Disconnected my 4K monitor to run as lean as possible. Using another laptop for everything else. Fan came on immediately as soon as my players started connecting and Roll 20 started to chug but it stayed playable.  There is something that happens when my players connect that just kills my computer. I've got three processes fighting for top of the CPU charts: Google Chrome (GPU) (50-60%) Google Chrome Helper (Renderer) (200-300%) Window Server (45-50%) The Renderer didn't hit 1000% like it was hitting last week. I'm running a 2019 MacBook Pro with 32G Ram and an 8GB Graphics card.   Since this machine is more than powerful enough for my day job as a software developer,  I'm loath to upgrade just to run Roll20.
1686363334

Edited 1686363387
Is there just a system requirement for roll20 now with all these updates? :/ It (at least to me)Feels* like it's kinda getting that way. (although with this problem even gaming computer rigs run badly depending) I want my players to be able to enjoy my games even if they're using a basic laptop used for nothing more than watching youtube, checking emails and using word. I really hope this gets sorted out soon and back to being a smooth playing experience for the mass majority of users.
1686368680
keithcurtis
Forum Champion
Marketplace Creator
API Scripter
Anecdotally, it seems like people with more powerful computers are the ones most affected by this phenomenon. i can run Roll20 on a Chromebook, for instance. Some have theorized that the issue might be some graphics cards, but there apparently have not been enough dev-reproducible incidents to diagnose. For those who have not seen the special response form: Drespar &nbsp;said: Hi everyone! For folks still experiencing high CPU or GPU usage, could you fill out this google form here: <a href="https://forms.gle/H4WWn6Nd9rqTexX69" rel="nofollow">https://forms.gle/H4WWn6Nd9rqTexX69</a> Please forgive me for the length, but performance related issues tend to be one of the most difficult to properly diagnose. There are often many many places where a system can differ slightly, or have an out of date driver, etc. Thank you in advance! This will help us greatly in deciphering what common threads might be or what the root issue may be. And of course, tests to try include: Toggling Hardware Acceleration on or off (most likely on) Ruling out extension issues by trying a private browsing window Ruling out a browser issue by testing on a different browser Ruling out bad data by deleting cache and cookies These (especially the Acceleration) showed early promise, but it seems like the ones most consistently affected are the ones for whom this has made no difference.
keithcurtis said: Anecdotally, it seems like people with more powerful computers are the ones most affected by this phenomenon. i can run Roll20 on a Chromebook, for instance. Some have theorized that the issue might be some graphics cards, but there apparently have not been enough dev-reproducible incidents to diagnose. This statement gave me an idea, and I just tested it. I've also been experiencing high CPU/GPU usage just to open the Roll20 VTT for the past few months, and what I just thought of is that those with more powerful/gaming systems probably also have nicer displays. Displays with a lot of pixels. I have a Lenovo ThinkPad T14 notebook with an 11th gen i7 processor, 16GB of RAM, integrated Intel Iris Xe video, and a high DPI display . Here's what my quick test found: 3140 x 2160 (default resolution): GPU immediately jumps to 100% usage and CPU goes to 50%+ before any players are added. 1920 x 1080: GPU now sits at ~50% usage, and the CPU dropped to ~10%. 1440 x 900: GPU went down to ~20% usage, and the CPU was pretty close to idle. So it would appear that Roll20 is scaling up resource usage directly related to the display resolution. Not sure what a solution is, though, because everything else I use for running a game would be very negatively affected by dropping my resolution down so low.
1686439712

Edited 1686439918
keithcurtis said: Anecdotally, it seems like people with more powerful computers are the ones most affected by this phenomenon. i can run Roll20 on a Chromebook, for instance. Some have theorized that the issue might be some graphics cards, but there apparently have not been enough dev-reproducible incidents to diagnose. For reference, I overcame this issue by upgrading to a more powerful device, so it is certainly not universal. Since we know that the issue did not exist on the DEV server for affected users, it seems pretty clear it is something on Roll20's end.&nbsp; Assuming Roll20 follows best practices, they should have multiple servers and past versions of Roll20 code saved.&nbsp; Why not spin up the version that was on the DEV server and have Morgan (since she was able to reproduce the issue) test to see if the the issue exists.&nbsp; If not, advance to the next version, test again.&nbsp; And so on until the Roll20 devs isolate the specific version of Roll20 that causes the problem. Then they would know exactly what version, and the associated features, that need to be analyzed further to try to solve the issue. -Adam
Retyping all of this because Roll20 logged me out somehow, so posting gave me a "you must be logged in to post" and also cleared this form so the back button didn't allow me to recover the text because OFC that's how Roll20 would implement this. My performance issues are entirely related to integrated video I spent almost an hour and a half testing this morning and have the following reasoning: Performance without integrated video (IV) is fine. Regardless of map, resolution, browser, etc. etc. Performance degrades with IV on and NO players (just mirroring myself). Performance further degrades with each new player with IV enabled added. Performance further degrades with each video player size increase (small to medium to large) Performance degradation caused by # players and size of video player COMPOUNDS. Performance with Firefox is significantly worse than Chrome (1.5x CPU) Performance for the GM is impacted even if they have [Names Only] as a video size (?!?!) We switched to integrated quite a while ago so this increasingly poor performance has been ongoing. Although difficult to test because of how flakey legacy video is: Legacy video has little impact on performance with just the GM mirroring myself. Adding other players as a small performance impact. Video size seems to have little impact. Legacy is pretty much unusable now as far as I can tell. Other observations: I believe my session Friday was playable because I was using small video size and two of my players had no cameras on Friday (just by coincidence). Firefox was still unusable in this scenario.
1686504156
keithcurtis
Forum Champion
Marketplace Creator
API Scripter
Was the performance degradation experiential or as measured by cpu usage? I.e. I play with four other people all using V&amp;V and we don't suffer lag. Although I suppose it's possible that when I say "we don't see lag" is that I just have different expectations of response time. Roll20 can be a bit slow at times, but it's very, very rare (in my personal experience) for it to become unplayable. Still, that's good observational data.
1686519435
Andrew R.
Pro
Sheet Author
I have the issue, as I've reported before, and I've never used Integrated Video or Legacy Video in any of my games. I've always used Google Hangouts (then) and Discord (now).&nbsp;
What I don't understand is why when I started using Roll20 and at certain stages performance has been fine. Is it the OS upgrades have affected how Roll20 is processed? Or is it things that Roll20 does now that it didn't before? Why when so many people are having similar issues that the devs are seemingly unable to recreate the problem? What is different for them than us? Honestly, I have not DM'd for a long while now because I cannot stand the noise my laptop makes with the fans going max speed for what might as well just be a static image on screen - it really makes no sense. But then I am ignorant of the goings on. It would be good to know that the devs are taking this seriously and looking into how the issue can be resolved for the many users here. The lack of communication does make one think that it's not considered an issue, but it is. I already cancelled my subscription but was persuaded to take a month for 99c in the hope that something might change in that time. I don't like the idea of my laptop having to run at max capacity for what is/should be a non-intensive application (GFX wise at least). &lt;grumpy face&gt;
Adago said: What I don't understand is why when I started using Roll20 and at certain stages performance has been fine. Roll20 makes changes to their software.&nbsp; When they make changes they test their changes but are unable to test on all of the different types of devices / browsers / etc that Roll20 users use to access the software.&nbsp; Roll20 then releases new features and is unaware of problems until a user notifies them that something is not working properly.&nbsp; Once it becomes apparent to Roll20 that a problem is not with a single user and is instead more widespread, they have a decision to make.&nbsp; Try to figure out the problem and fix it, or roll back to the previous version of the software. I would guess that as this problem is not universal they have made the business decision that the new features they have provided to the majority of the userbase is more important than the problems they have created for a minority of the userbase. At the end of the day, this problem has not cost them enough money in cancelled subscriptions to make it a top priority to resolve. -Adam
Adam Caramon said: At the end of the day, this problem has not cost them enough money in cancelled subscriptions to make it a top priority to resolve. Sigh.&nbsp;
I believe the problem is related to graphics cards. I have 2 pc one with 7th gen i7 and the other with 8th gen i7. In both I notice an excessive use of the integrated graphics cards Intel HD 620 (95-100%) and intel UHD 620 (65-75%) which totally collapses to 1-3% when I switch to another browser tab and then roll20 doesn't it is more in the foreground.
1686606852
keithcurtis
Forum Champion
Marketplace Creator
API Scripter
Could be. Everything I own has integrated graphics.
Alex S. said: I believe the problem is related to graphics cards. I have 2 pc one with 7th gen i7 and the other with 8th gen i7. In both I notice an excessive use of the integrated graphics cards Intel HD 620 (95-100%) and intel UHD 620 (65-75%) which totally collapses to 1-3% when I switch to another browser tab and then roll20 doesn't it is more in the foreground. For me it was the moment they introduced the window functionality to dynamic lighting. Everything worked absolutely fine, my top cpu usage in game was about 10% with a bunch of people on huge maps full of dynamic lighting. Then once they added windows I was at 90% cpu usage without hardware acceleration and 25% cpu/45% gpu with it on in empty games with no maps or players. But at the time they had the development servers which was a few versions behind, and while the productions server was running so poorly, the same game copied over to the development servers ran perfectly.
Robert O. said: Retyping all of this because Roll20 logged me out somehow, so posting gave me a "you must be logged in to post" and also cleared this form so the back button didn't allow me to recover the text because OFC that's how Roll20 would implement this. My performance issues are entirely related to integrated video I spent almost an hour and a half testing this morning and have the following reasoning: Performance without integrated video (IV) is fine. Regardless of map, resolution, browser, etc. etc. Performance degrades with IV on and NO players (just mirroring myself). Performance further degrades with each new player with IV enabled added. Performance further degrades with each video player size increase (small to medium to large) Performance degradation caused by # players and size of video player COMPOUNDS. Performance with Firefox is significantly worse than Chrome (1.5x CPU) Performance for the GM is impacted even if they have [Names Only] as a video size (?!?!) We switched to integrated quite a while ago so this increasingly poor performance has been ongoing. Although difficult to test because of how flakey legacy video is: Legacy video has little impact on performance with just the GM mirroring myself. Adding other players as a small performance impact. Video size seems to have little impact. Legacy is pretty much unusable now as far as I can tell. Other observations: I believe my session Friday was playable because I was using small video size and two of my players had no cameras on Friday (just by coincidence). Firefox was still unusable in this scenario. WebRTC takes quite a lot of resources, specially if coded poorly which your point 7 tells us it is. I'm guessing they're let our browsers handle the transcoding, since it would be to costly to handle it on the server side. If that is not done with a lot of experience, this is exactly what happens. -- On another note regarding lag. I'm quite new here and just started a campaign a few weeks ago. The lag for some of us is SUPER bad, and everyone is sitting on gaming computers deluxe...&nbsp;
I'm going to throw a curve ball out here - I've been having this issue for about 3 months now. Looked at all the solutions etc - we are talking literally a blank, brand new map - nothing on it, no other people, and just opening and moving around the interface of roll20 instantly makes my fans ramp up to the max speed due to how much CPU it's using. It never used to do this, been a roll20 user for a way long time and it's always been stable. Then I did a crazy thing - I dragged Chrome over to my 2nd monitor. I moved around the roll20 page....and no fan ramping. Usage was still - in my opinion - higher than you would hope an optimised site/system would be...but no fan noise for certain. I then dragged it back to my main monitor...instant fan ramp. Both monitors are using the same graphics card, the same type of connector - the only difference is my main monitor is 1440p. Could, and its a long shot, it be that the site doesn't function well with specific resolution settings, or is this all just tinfoil hat theory by a man desperate to find a fix so I - and all the community - can play our fun dnd games in peace?
Hey Folks, I'm sorry for the problems you all have been experiencing. I wince every new post I read. Troubleshooting and diagnosing performance problems can be VERY difficult. Often times some symptoms can lead you down one direction while others take you in a different direction. At the end, you realize you've been tracking two different unrelated problems. I want you all to know that I take performance problems very seriously; we are taking them seriously. So seriously that we've spent the last couple weeks thinking through the logistics of how we can fully address these problems. I don't have any updates right now other than to simply let you know that this is important, it requires attention, and it needs to improve. The way we fix this problem and the many others that are listed here is by making more systemic changes. That fix will take us a while but will produce results. When I know more specifics and it makes sense, I'll let you all know. Until then, know that I have this thread followed and I read each and every post. Thank you all again for your hours of debugging, your patience with a frustrating game experience, and your tenacity to see this fixed.
Then I did a crazy thing - I dragged Chrome over to my 2nd monitor. I moved around the roll20 page....and no fan ramping. Usage was still - in my opinion - higher than you would hope an optimised site/system would be...but no fan noise for certain. I then dragged it back to my main monitor...instant fan ramp. Both monitors are using the same graphics card, the same type of connector - the only difference is my main monitor is 1440p. Odd as it sounds, depending on your GPU and CPU, sometimes a lower monitor resolution will cause a CPU bottleneck as the GPU processes the graphics data faster than the CPU can process the frame, causing it to run at high capacity in an effort to keep up.
Andrew Searles said: Troubleshooting and diagnosing performance problems can be VERY difficult. Often times some symptoms can lead you down one direction while others take you in a different direction. At the end, you realize you've been tracking two different unrelated problems. Hey Andrew - I appreciate most of your posts as it is clear you have a better style of dealing with the community than other Roll20 staff have in the past.&nbsp; That said, you aren't providing much in the way of specifics here, and it is hard to give Roll20 the benefit of the doubt when we view their actions in totality.&nbsp; The DEV server proved that this problem wasn't due to something on end-user's PCs.&nbsp; If the problem exists on a production server, but does NOT exist on the DEV server (older version of Roll20 code) when accessed with the same device, it is obvious that something Roll20 changed caused the problem.&nbsp; The fact that Roll20 decommissioned the DEV server shortly after this was pointed out is troubling. Now, you can certainly state that the changes Roll20 have made SHOULD not cause these problems, but can we at least get a baseline admission from Roll20 that something that Roll20 changed on their end (most likely the Windows/Door update) caused the current problem?&nbsp; I feel like that would go along way to building good will / understanding between end-users affected by the problem and Roll20 staff. I hope there is not some kind of Roll20 employee policy that prohibits you from publicly acknowledging fault lies with Roll20.&nbsp; This lack of accepting responsibility, coupled with other less-than transparent policies/rules/practices Roll20 has in place makes it difficult to support Roll20 devs / believe they are doing what they can to get to a resolution. Personally, I can remember a time when I stalwartly defended Roll20 and would highly recommend it to all of my players / friend groups.&nbsp; In a perfect world, I'd like to still be there, but the lack of transparency on Roll20's part makes that impossible. -Adam
1687307129

Edited 1687308237
Andrew Searles
Roll20 Team
Hey Adam Thanks for your comment! I wanted to respond directly. I can see how my message wasn't clear. Thank you for pointing that out and keeping us honest! you aren't providing much in the way of specifics here, and it is hard to give Roll20 the benefit of the doubt when we view their actions in totality. You're right. I didn't give specifics when I said, "Troubleshooting and diagnosing performance problems can be VERY difficult." Here is why it's hard for us to diagnose performance problems. (1) It is difficult to get steps to reliably reproduce the specific problems, (2) symptoms look similar but can be caused by something radically different, (3) there are a multitude of variables that affect performance, and (4) cycle time can be time-consuming. Let's take each one of these individually. (1) It is difficult to get steps to reliably reproduce the specific problems Getting reliable steps to reproduce a problem is the gold standard for which we shoot. These steps also make sure that we can test the solution as well. When we have these steps, the true fix to the problem is close behind. But, due to the number of variables that affect a performance problem, it can be nearly impossible to pinpoint all of them. (2) Symptoms look similar but can be caused by something radically different Often times lag, especially frame rate lag, can be caused by any number of things. This is often why so many people describe different ways they get into a laggy state OR it's why some people can't put their finger on exactly what happened right before it got laggy. Grouping these symptoms can lead us down multiple paths. For example, the fix that we posted didn't work for a number of people on this thread. For us, in our testing and what we were reproducing, it worked and is still working. This doesn't mean we didn't fix a problem, it just means we didn't fix all the problems. (3) There are a multitude of variables that affect performance Performance of the VTT is tied to the hardware and software that is used to display it. The number of variables that affect performance are equal to the number of computer types, operating systems, graphics cards, browsers, and rendering options that are out there. These variables can make it hard to get to the bottom of what is causing problems. (4) Cycle time can be time-consuming Because of the large number of variables that are out there, the number of things we need to check can get daunting. I'm sure those of you that have messaged support have experienced this frustration. I know I've read a number of comments on this thread that indicate the frustration. It can be time consuming for everyone involved. — I hope this explains a little more as to why it's hard to troubleshoot performance tickets. I'll restate what I said in my previous message to emphasize it. I want you all to know that I take performance problems very seriously; we are taking them seriously. So seriously that we've spent the last couple weeks thinking through the logistics of how we can fully address these problems. I don't have any updates right now other than to simply let you know that this is important, it requires attention, and it needs to improve. The way we fix this problem and the many others that are listed here is by making more systemic changes. That fix will take us a while but will produce results. The DEV server proved that this problem wasn't due to something on end-user's PCs.&nbsp; If the problem exists on a production server, but does NOT exist on the DEV server (older version of Roll20 code) when accessed with the same device, it is obvious that something Roll20 changed caused the problem.&nbsp; The fact that Roll20 decommissioned the DEV server shortly after this was pointed out is troubling. You are right, performance hasn't always been a problem for everyone and it has been getting worse for some people. It's something that we've been monitoring and it's why we're taking it seriously. I do agree that the timing of decommissioning the dev server might seem weird given your narrative but I can assure you that we didn't decommission the dev server to hide anything. In your next comment, I'll address what I think you might be talking about more specifically. can we at least get a baseline admission from Roll20 that something that Roll20 changed on their end (most likely the Windows/Door update) caused the current problem? Absolutely! We have made a number of changes that could very well be the source of a number of problems in performance for some people. It's one of the main things we're looking into to try and solve these problems. And, you're right, the Windows/Doors could be the source of the problems. Here's some technical information that might help give you more context. We use a number of different rendering engines to help show your map. The main engine we use currently is called Fabric but we've found that it does not give us everything we need. We've identified a new rendering engine called Babylon that gives us a lot more capabilities into the future. It might also reduce the system requirements and speed up performance overall. In order to replace Fabric with Babylon, we had two options. We could either start a long project to replace it all at once or we could slowly roll out new features using Babylon instead of Fabric, slowly replace it over time. We chose the second option and released the windows and doors update with Babylon on top of Fabric. Our current data tells us that the introduction of Babylon does not account for all the performance problems we've seen. For example, the recent fix had nothing to do with Babylon. We're not certain there is one single thing we can "fix" to solve all performance problems. We don't see a smoking gun. I wish we did, it would make things easier to fix. More likely, what we're seeing is a gradual degradation of performance as we add more and more features that require more processing power. This is why we're taking this seriously. We want to make it easier for everyone to tell their stories regardless of the device with which they play. In order to make that happen, we're going to have to do more than just a few fixes. We need a systemic change. Personally, I can remember a time when I stalwartly defended Roll20 and would highly recommend it to all of my players / friend groups.&nbsp; In a perfect world, I'd like to still be there, but the lack of transparency on Roll20's part makes that impossible. If transparency is what is needed, then I can TOTALLY help with that! That's a problem we can fix! But seriously, I really do appreciate you and your dedication to Roll20. It's an honor. And thank you again for keeping us honest. If you or anyone else have questions, comments, thoughts, or opinions, please let me know. If you're not comfortable leaving them in the thread, feel free to message me directly or schedule a time to chat using this link. <a href="https://calendly.com/acsearles/performance-issues" rel="nofollow">https://calendly.com/acsearles/performance-issues</a>
That was really insightful Andrew, Thank you. I'm glad to see you guys are in fact taking all this seriously and not dodging the issues. This does put most of the worries I've had to rest. I look forward to seeing this get fixed and am excited to see what new things I can enjoy once it does going forward.
Myself and multiple players are having the same issue the last couple of weeks. Following , and for now hoping you guys will find a fix seeing as its made the game enviroment feel downright hostile atm.&nbsp;
1687514069

Edited 1687514101
Last session (yesterday) was simple said a fumble on many sides for some players. High RAM and CPU usage, Crashes and lag. Since i am playing Roll20 there were always some lags - but the last experience was truly horrendous. We used Firefox and Chrome, restarted the sessions, disabled dynamic lighting and some mods but it was not enough to have a good and smooth feeling. I hope it will get better soon.
1687702698

Edited 1687702716
Andrew Searles said: Hey Adam Thanks for your comment! I wanted to respond directly. I can see how my message wasn't clear. Thank you for pointing that out and keeping us honest! you aren't providing much in the way of specifics here, and it is hard to give Roll20 the benefit of the doubt when we view their actions in totality. You're right. I didn't give specifics when I said, "Troubleshooting and diagnosing performance problems can be VERY difficult." Here is why it's hard for us to diagnose performance problems. (1) It is difficult to get steps to reliably reproduce the specific problems, (2) symptoms look similar but can be caused by something radically different, (3) there are a multitude of variables that affect performance, and (4) cycle time can be time-consuming. Let's take each one of these individually. (1) It is difficult to get steps to reliably reproduce the specific problems Getting reliable steps to reproduce a problem is the gold standard for which we shoot. These steps also make sure that we can test the solution as well. When we have these steps, the true fix to the problem is close behind. But, due to the number of variables that affect a performance problem, it can be nearly impossible to pinpoint all of them. (2) Symptoms look similar but can be caused by something radically different Often times lag, especially frame rate lag, can be caused by any number of things. This is often why so many people describe different ways they get into a laggy state OR it's why some people can't put their finger on exactly what happened right before it got laggy. Grouping these symptoms can lead us down multiple paths. For example, the fix that we posted didn't work for a number of people on this thread. For us, in our testing and what we were reproducing, it worked and is still working. This doesn't mean we didn't fix a problem, it just means we didn't fix all the problems. (3) There are a multitude of variables that affect performance Performance of the VTT is tied to the hardware and software that is used to display it. The number of variables that affect performance are equal to the number of computer types, operating systems, graphics cards, browsers, and rendering options that are out there. These variables can make it hard to get to the bottom of what is causing problems. (4) Cycle time can be time-consuming Because of the large number of variables that are out there, the number of things we need to check can get daunting. I'm sure those of you that have messaged support have experienced this frustration. I know I've read a number of comments on this thread that indicate the frustration. It can be time consuming for everyone involved. — I hope this explains a little more as to why it's hard to troubleshoot performance tickets. I'll restate what I said in my previous message to emphasize it. I want you all to know that I take performance problems very seriously; we are taking them seriously. So seriously that we've spent the last couple weeks thinking through the logistics of how we can fully address these problems. I don't have any updates right now other than to simply let you know that this is important, it requires attention, and it needs to improve. The way we fix this problem and the many others that are listed here is by making more systemic changes. That fix will take us a while but will produce results. Hi Andrew, Thanks for the response, and in particular the great amount of detail you have gone into here. Absolutely! We have made a number of changes that could very well be the source of a number of problems in performance for some people. It's one of the main things we're looking into to try and solve these problems. And, you're right, the Windows/Doors could be the source of the problems. Here's some technical information that might help give you more context. "could very well be the source..." This is not as direct as I imagine most of us would like to see.&nbsp; Can you explain how a person's device could work perfectly fine on the DEV server, have issues on the PROD server using the exact same device without any alterations on their side, and it NOT be due to a change on the Roll20 side?&nbsp; This is the part that I cannot understand. Can you further explain why the DEV server would not be used to ferret out these issues?&nbsp; I.E., when it was determined that the problem did not occur on the DEV server but did occur on the PROD server, why was the DEV server decommissioned rather than use it as a troubleshooting tool?&nbsp; If Roll20 had offered a free art pack or something similar to affected users in exchange for them running through a testing script on the DEV server a few times as Roll20 incremented it through higher versions, would this not have provided valuable data to isolate the issue? If transparency is what is needed, then I can TOTALLY help with that! That's a problem we can fix! But seriously, I really do appreciate you and your dedication to Roll20. It's an honor. And thank you again for keeping us honest. If you or anyone else have questions, comments, thoughts, or opinions, please let me know. If you're not comfortable leaving them in the thread, feel free to message me directly or schedule a time to chat using this link. <a href="https://calendly.com/acsearles/performance-issues" rel="nofollow">https://calendly.com/acsearles/performance-issues</a> I'm not sure how to interpret these statements.&nbsp; Are you saying you can't help with transparency?&nbsp; Because I would argue you have provided far more detail here than any other Roll20 staff have, so you are already doing better than what the community is used to receiving. Thank you for taking the time to respond directly.&nbsp; It is appreciated. -Adam
1687771697

Edited 1687941867
So, even in prepping my session for tomorrow (no other players in game) cpu&amp;gpu usage is between 85/100%. Having tried to disable all the settings and UI improving things I pay the big pro sub for and&nbsp; still getting these results in whatever browser I use I'm really dreading my game tomorrow. (which is the opposite of what I hope to feel before a game)&nbsp; All these UI improvements look promising and exciting but feel pretty useless if the players can't see&nbsp; or hear the DM. Which is sort of the bread and butter of an RPG.&nbsp; I've always had "some" problems in roll20 (mainly a/v related)&nbsp; but mostly took them for granted as they were manageable and because the alternative would be to forfeit my money and time investment and start over on a different platform.&nbsp; But if this continues I think I will, as others before me, have to do just that.&nbsp; EDIT: We ended up just using discord, and will keep ending up there untill there is a noticeable fix. Having the fun drained out of game night due to technical difficulties is just not worth it.&nbsp;&nbsp;