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
This post has been closed. You can still view previous posts, but you can't post any new replies.

WebRTC Video and Voice Chat Issues

Level2Gamers said: Our most recent session Wednesday, was a total $@^* show. Yes, we were on lowest settings, yes we hit reconnect, yes we all had the same output. We ended up having to use Discord to maintain communication. I'm all about new features and what Roll20 is bringing to the table (pun intended), but all of it is moot without proper communication. This will probably be my last pro month on the site if these issues aren't fixed soon. I apologize if this sounds unreasonable, but forcing a buggy and broken service on your clientele is much more so. I was thinking of a game next week.... however wont bother with roll20 at the moment.... might even cancel my account at this rate due to it being so unstable and devs are adamant that this is way to go... when clearly not working..... if they want to move to a new system fine im all for it.... however proper testing is required before forcing it on everyone
Going to add my voice here. I haven't used the video/voice system since the switching. Don't want to risk my sessions cause of it. Will continue to use the table top portion.  However am thinking of backing down my sub level based of the lack of a useable video/voice function. My personal opinion on this is that we need more communication from the Dev team.  That has been sorely lacking in this.  With the state of the system, I would think daily or close to it updates would be appreciated.  I understand it is the holiday times, however the games aren't stopping.  We are getting sporadic responses but will very little information on what is being done to fix the issues or a timeline of improvements.   And honestly, an apology would go along way. Admitting that this was not a good roll out of a change by the smallest of standards would be a good start.  But a timeline on when we will see this fixed would be the best!
Just did a little test here since our group is on break.  Things look like they are getting better.  Thanks Devs! Some bug observations: 1. Firefox video quality is poor.  (Chrome is crystal clear) 2. Firefox does not play well with the pop-out.  The feed freezes on the last frame at the time of the popout (Chrome no problem.)
1483157758

Edited 1483158885
Was getting a similar whack-a-mole problem tonight. &nbsp;I could hear everyone then refreshed and could only hear the GM. &nbsp;We have 6 total ppl with multiple voice and video receiving settings. &nbsp;Here's my console logs --&nbsp; <a href="http://pastebin.com/xzwppEP2" rel="nofollow">http://pastebin.com/xzwppEP2</a> and for some of the other players: <a href="http://pastebin.com/sExPQ4pa" rel="nofollow">http://pastebin.com/sExPQ4pa</a> <a href="http://pastebin.com/q5PTV9cR" rel="nofollow">http://pastebin.com/q5PTV9cR</a> <a href="http://pastebin.com/e76pzPiD" rel="nofollow">http://pastebin.com/e76pzPiD</a> &nbsp;(DM's perspective) also, here's my latest testRTC results:&nbsp;<a href="https://test.webrtc.org/report/AMIfv96svu9IKix-DI13OIs2D0FpDkobMVBHmMY1vNDAld1RuuJ3u0n0jSiX3R_QnnH6u9WKY-BJpE1WQaomilbsQ0uEpeNvOiDFg8vWJ7-jAxKKGPZYQ28K2RIz0QHmlyRyZ6AnDqkkp5SXkGEfF-CgX_X-1W75ow" rel="nofollow">https://test.webrtc.org/report/AMIfv96svu9IKix-DI13OIs2D0FpDkobMVBHmMY1vNDAld1RuuJ3u0n0jSiX3R_QnnH6u9WKY-BJpE1WQaomilbsQ0uEpeNvOiDFg8vWJ7-jAxKKGPZYQ28K2RIz0QHmlyRyZ6AnDqkkp5SXkGEfF-CgX_X-1W75ow</a>
Guess I'll share the results of my Thursday night session. I dont have any logs to share. Bandwidth was set to Medium, had myself and 3 other players. Our game was seemingly solid for a hallf hour, until suddenly a players video feed went black and lost his audio. Upon reconnecting, it became whack a mole and we couldnt get it to work again. Upon receiving the footage I did capture (recording), the sound did not come in at all for one of my players that broadcasts only voice. Basically, a 4th ruined session in the span of 5 weeks now. What a complete joke. Look, I get the desire to want new features, but all of it means nothing if the only video system you do have is complete garbage and doesnt work. Being a software developer myself, I can also understand the difficulties of releasing new software. But this is getting to the point of just complete nonsense. So openTok was no longer viable for roll20 to be profitable. Well, from what I see, and the amount of people getting pissed off over this, I'd say it's not viable to NOT use openTok for roll20 to remain profitable. At least that WORKED. Get some VC or additional investor capital. Sell your soul. Whatever. Do something to bring back the system that worked.&nbsp; At this point, I'm really regretting becoming &nbsp;pro member as right now, I feel it's a waste of my $100. Just use skype? Just use discord or appear.in? No, those arent great options as i need the video embedded into the application to save on screen real estate. Plus, the reason I subbed at pro level was because the video chat was one thing that made roll20 awesome. Now....its just meh. What a complete letdown.
1483345775
Karl V.
Plus
Translator
Not working at all. Playing whack-a-mole like everyone else. This is a very serious downgrade of the functionallity roll20 offers. I have been a supporter and fan of this service for a very long time, but unless the video and voice chat is brought back very soon I will have to start reviewing my options.
1483360035

Edited 1483360454
When will this be fixed? &nbsp;It's been ongoing since October now. Had my regular session tonight and the same annoying issues occurred forcing us to delay our start and swap to hangouts given my party is tired of beta testing during play sessions. &nbsp; We record these and prior to this issue, Roll20 was the easiest solution for capturing online gameplay and video/audio but no longer. Similar to others, I've began reviewing alternatives to my Roll30 sub as I'm disappointed in the response by Roll20 and exposure of how poorly they QA new releases which makes me question what core features may get hosed next.
1483399228
Stephen Koontz
Forum Champion
Marketplace Creator
Sheet Author
API Scripter
Compendium Curator
Hi rollers, We hear your frustration. There are some users that had no issues with the previous chat system that are experiencing chronic issues with the new system. While our metrics show that the vast majority of users are now using the integrated Video and Voice chat without serious issues, that doesn't matter if you're one of those being affected. Our development team has run out of optimizations and fixes that are possible based on the self-reported data we're getting back. We know that our new system uses more upload data and that can cause problems for users with limited upload data plans. Without access to hardware acceleration or prohibitively expensive servers, this is a reality we won't be able to change. Instead, we'd like to focus on issues that are caused by combinations of operating systems, browsers, geo-location, ISP, and more. Combinations of factors there is no way for our small development team to test. That's where we're hoping you'll help us help you. If your group is experiencing chat issues and you're interested in receiving direct assistance, let us know. Send me,&nbsp; Steve K. ,&nbsp;a private message with the name of the game and time your group will next meet. If we're available, we'll have a developer come to collect data on the specific set of conditions caused by the web of peer-to-peer connections created by your group and provide whatever assistance we can.&nbsp;
Once again unable to use Video or Audio. A different group of players. None able connect.&nbsp; Console Log:&nbsp;<a href="https://www.dropbox.com/s/i2u564cz4vi7ywz/command%20log%20r20.txt?dl=0" rel="nofollow">https://www.dropbox.com/s/i2u564cz4vi7ywz/command%20log%20r20.txt?dl=0</a>
Steve K. said: Hi rollers, We hear your frustration. There are some users that had no issues with the previous chat system that are experiencing chronic issues with the new system. While our metrics show that the vast majority of users are now using the integrated Video and Voice chat without serious issues, that doesn't matter if you're one of those being affected. Our development team has run out of optimizations and fixes that are possible based on the self-reported data we're getting back. We know that our new system uses more upload data and that can cause problems for users with limited upload data plans. Without access to hardware acceleration or prohibitively expensive servers, this is a reality we won't be able to change. Instead, we'd like to focus on issues that are caused by combinations of operating systems, browsers, geo-location, ISP, and more. Combinations of factors there is no way for our small development team to test. That's where we're hoping you'll help us help you. If your group is experiencing chat issues and you're interested in receiving direct assistance, let us know. Send me,&nbsp; Steve K. ,&nbsp;a private message with the name of the game and time your group will next meet. If we're available, we'll have a developer come to collect data on the specific set of conditions caused by the web of peer-to-peer connections created by your group and provide whatever assistance we can.&nbsp; Sounds fine that your willing to offer such help with is a step... My concern is your alienating people if there is no fix...Will you offer them refunds due to not getting the service they paid for?
1483441150
Jens A.
Pro
API Scripter
Is there a specific reason you can't reinstate OpenTok as a fallback solution?
At this point the new video chat solution is what we're sticking with. We had to make a decision based on the long-term success of keeping video chat in Roll20; while I can't discuss the specifics of our deal with our previous provider, it wasn't going to be something we could keep doing long-term and keep offering video chat to the entire Roll20 community like we have been. And that's in addition to the other reasons that we felt it was better long term, which we've already covered in detail; such as the ability to add new chat features (like the pop-out video chat and the video chat whispering that we've already been able to roll out on the new system, and we couldn't do with the old one). As I've already said on here previously, we cannot keep the old system around as merely a fallback, as that's not how the agreement with the provider worked. At this point, as Steve said, the vast majority of our user base, tens of thousands of people every day, have transitioned successfully to the new system. There are still edge cases that exist which is why we are offering to help specific groups if they need that help. At the same time, if you feel like the experience on the site is no longer working for you and want a refund, email <a href="mailto:team@roll20.net" rel="nofollow">team@roll20.net</a> and they will get you one.
glad there is so many that know how to help roll20 debug their system.&nbsp; we have problem with microphone does not allow one of 4/5 of us to be able to communicate (video works fine ) we dont need VIDEO we NEED microphones to work the make changes and reconnect hasnt seem to work , i would like to know /find the step by step process of what are we supposed to be sharing as our group so that you can fix this.&nbsp;
1483484462
Gid
Roll20 Team
We've created a survey to help us better pinpoint devices, location, bandwidth rates, and other technical specs of those having issues. This will hopefully give us more clues and potentially a pattern to track for those still struggling with Roll20 WebRTC. If you have a few minutes, please fill out this survey:&nbsp; <a href="https://www.surveymonkey.com/r/Roll20WebRTC" rel="nofollow">https://www.surveymonkey.com/r/Roll20WebRTC</a>
Kristin C. said: We've created a survey to help us better pinpoint devices, location, bandwidth rates, and other technical specs of those having issues. This will hopefully give us more clues and potentially a pattern to track for those still struggling with Roll20 WebRTC. If you have a few minutes, please fill out this survey:&nbsp; <a href="https://www.surveymonkey.com/r/Roll20WebRTC" rel="nofollow">https://www.surveymonkey.com/r/Roll20WebRTC</a> I took the survey in hopes that it will help. However, please keep in mind, that my game changed its m.o. because of the faulty AV. Both the groups I lead and the group I play in have moved to alternate solutions for AV. We are still able to play, but having the A/V integrated into one system was a selling point; it isn't anymore. &nbsp;Frankly, the value proposition went way down in my book, but since I have free solutions I can adapt and still play at a better price than the primary competitor. But, that doesn't mean I'm not disappointed. I am a little perplexed at the idea is related to end users systems, O/S, ISP combo's because my experience is that three other AV solutions (Google Hangouts, Skype, and Discord) run error free. I haven't seen the metrics (of course), but the symptoms I see and the ones other report in this thread don't really point that way to me (especially when Steve suggests above that higher end servers would make a difference). That said, I appreciate R20 offering to help individual groups. However, I am concerned that this approach will miss a swath of users having issues because folks just want to game at this point and can do so using other AV solutions (just as my group has done). Besides, this seems to be a bit of a crap shoot for gathering data. Why not a more systematic approach...why not put together a few test games/groups and ask for volunteers with different configurations to join. I don't think it would be hard to get a diverse group to create specific use cases. I'd happily jump into a dev run sandbox "game" and run diags and use cases for a few hours if a test slot fit into my schedule. I'd bet a lot of other folks would too...especially if it wasn't during their pre-scheduled game time. &nbsp;Just a thought.
Took survey. Everyone in our group of six people have been playing wack-a-mole, countless hours, for over a month, we use audio only. Hoping for better... finally having to switch to any other VoIP... :(
We've switched to Discord permanently, we only ever used voice anyway. It's stable and works over VPN for my player in China. I couldn't get Roll20 to work this weekend even with only two of us in the same timezone, and I don't have time to muck with it.
My group spends enough time in game dealing with tech issues from at least one of the players, that dealing with the AV issues isnt even worth the attempt. We havent used WebRTC for a long time and just put up with using Hangouts. We have so little time together, that everyone would rather just play than try to deal with or fix other issues that dont show up elsewhere. I love the idea of all of the 'features' WebRTC is supposed to have, but it never worked for us before the big switch and upgrade, and the few times we've played with it afterwards havent worked at all. Like Scoundrel said, I think the Dev team is missing all of the people that gave up using WebRTC and moved to other AV solutions. I imagine the vast majority are in the boat of: "lets play, and not act as trouble shooters". And while having a support email to work with individual groups is nice, I know my group wouldnt spend the time to work through issues, when we can just use something else.&nbsp;
Guys just a thought but when you try to test for the bugs we users are having are you all remote or same location as &nbsp;most users are not site based.
1483567223

Edited 1483568585
While the refund option is much appreciated, I'm confused on one item here.&nbsp; Riley D. said At this point, as Steve said, the vast majority of our user base, tens of thousands of people every day, have transitioned successfully to the new system. There are still edge cases that exist which is why we are offering to help specific groups if they need that help.&nbsp; Is Roll20 stating they have analytics that confirm that tens of thousands of Roll20 Users are having zero issues with the new A/V but &nbsp;require those having &nbsp;issues to self-report and fill out&nbsp; a survey? Curious what metric you're using to signal "successful" use? I'd agree that my assumption here is that because Roll20 has tens of thousands of users and only a handful have took the time to self report, that means tens of thousands are having a good A/V experience. People get on to play a game with limited time and if your A/V is buggy, most will do what we did the first few times it happened is get annoyed and open Hangouts/Skype/Slack/Discord and play on. Took us 3-4 sessions (and my own private tests with them outside of session time) of A/V instability to swap completely to Skype going forward and actually turn off the AV options completely in Roll20. &nbsp;Wonder then if my tally would count towards the success or failure metric for A/V? It's just disappointing for me as I really enjoy Roll20 compared to Maptools or FG but the A/V was the real clincher. &nbsp;Having that separate again makes it feel like a letdown as the one reason I pushed my party into transferring the campaign to Roll20 from our old VTT was the integration which we got a good single session out of before it went south.
1483571986
Stephen Koontz
Forum Champion
Marketplace Creator
Sheet Author
API Scripter
Compendium Curator
Brandon H said: Is Roll20 stating they have analytics that confirm that tens of thousands of Roll20 Users are having zero issues with the new A/V but &nbsp;require those having &nbsp;issues to self-report and fill out&nbsp; a survey? That's correct. We have metrics on active users participating successfully in chat. And despite the snarkiness of your question, the answer is yes, we need users still having chronic issues to self-report. Knowing who is connecting and for how long doesn't help us understand when users are not ever successfully connecting. Or, more frustratingly when they connect but are disconnected in a way that doesn't throw an error or is otherwise indiscernible for our system from them voluntarily leaving the chat.&nbsp;
1483585705
David S.
Pro
Sheet Author
My weekly Tue. group hasn't had working AV since the switch to the new system. Yesterday was our first session of a new campaign, after finishing the old one. Same problems, randomly can't hear various people while others can hear them just fine. Doesn't matter if we switch to audio only or anything. We've used Discord since the switch. Here's what I grabbed for the console logs, right before we went to Discord:&nbsp; <a href="http://pastebin.com/Ji5jT2Jb" rel="nofollow">http://pastebin.com/Ji5jT2Jb</a> Before the switch we did mixed video and audio only. I'd love to get back into having video or at least stable audio once again.
Steve K. said: Brandon H said: Is Roll20 stating they have analytics that confirm that tens of thousands of Roll20 Users are having zero issues with the new A/V but &nbsp;require those having &nbsp;issues to self-report and fill out&nbsp; a survey? That's correct. We have metrics on active users participating successfully in chat. And despite the snarkiness of your question, the answer is yes, we need users still having chronic issues to self-report. &nbsp; Not a very professional way to answer your community. A simple yes would have done.&nbsp; how can you tell when there are connections but discrepancies between people not connect to each other if not reported. I see most people not wanting to bug report and just move on
I filled out the survey. &nbsp;My groups have been using discord since the swap. &nbsp;One was using it already for reasons that have nothing to do with the swap and i moved the other onto discord as the new system wasn't working at all. &nbsp;Even when it was in BETA it barely worked and i had to switch all my games back to OpenTok. &nbsp;I've been waiting for the forum posts to calm down before attempting to switch back. That hasn't happened yet. &nbsp;I only get to play with my group for 3 hours once week on Sunday night and spending even 15-30min of that limited time to trouble shoot Roll20's voice software makes no sense when there are 3rd party options that do the job. &nbsp;As has been suggested by other posters I'd be willing to make myself available to trouble shoot if it helps speed up the process.
1483653706

Edited 1483654067
Stephen Koontz
Forum Champion
Marketplace Creator
Sheet Author
API Scripter
Compendium Curator
David S. said: My weekly Tue. group hasn't had working AV since the switch to the new system. Yesterday was our first session of a new campaign, after finishing the old one. Same problems, randomly can't hear various people while others can hear them just fine. Doesn't matter if we switch to audio only or anything. We've used Discord since the switch. Here's what I grabbed for the console logs, right before we went to Discord:&nbsp; <a href="http://pastebin.com/Ji5jT2Jb" rel="nofollow">http://pastebin.com/Ji5jT2Jb</a> Before the switch we did mixed video and audio only. I'd love to get back into having video or at least stable audio once again. Thanks for the console log, David. It was very helpful in spotting an issue we haven't seen before. The log shows you attempting to connect to the 4 other people in the call but never receiving a response from any of them.&nbsp; If you're able to provide the "Video + Voice" settings and "Video Bandwidth" setting you and as many of your players used as you can provide, it would be very useful. Also, if you could run through this WebRTC connectivity test that information in conjunction with your log would be helpful. To test your WebRTC connectivity, please go to <a href="https://test.webrtc.org/" rel="nofollow">https://test.webrtc.org/</a> for a webRTC diagnostic test. Once there: 1. Click on the Start button. 2. Allow the test to run itself through the various subtasks (Microphone, Camera, Network, Connectivity, and lastly Throughput). 3. If the test runs into any issues, click on the bug icon at the top, this brings up a new window to create a bug report. 4. Describe what issues you had with the diagnostic test and then click on “Upload Report” 5. This will pull up a new window that provides a link to the report (“The *link* to the report is now available”). This link will be available for the next 90 days. Copy the link and paste it here in this thread.
Anyone up for some testing????? This is not Roll 20 certified, but maybe we can take some time, all pull logs, and maybe get this hashed out. <a href="https://app.roll20.net/lfg/listing/65814/gm-roll-call" rel="nofollow">https://app.roll20.net/lfg/listing/65814/gm-roll-call</a>
1483660028

Edited 1483660078
Kathy
Plus
I will be honest. If your metrics show tens of thousands of users not having problems with the system, it's time to look at the validity of your measurements. (I apologize that this is so cranky, but it's honest.)&nbsp;
Numtini said: I will be honest. If your metrics show tens of thousands of users not having problems with the system, it's time to look at the validity of your measurements. (I apologize that this is so cranky, but it's honest.)&nbsp; I'm sorry, but that logic makes no sense. This thread represents the vocal minority of people that this doesn't work for. Why would the tens of thousands of people that this works for have a need to post on this thread when they have no issues. I had WebRTC problems a few weeks back with a session, so I posted a report. But yesterday I had a session with no issues whatsoever, so I had no need to send one this time. Just because a number of people have made reports and had problems, does not make them the majority. Most people on roll20 have never posted a forum post beside LFG so why would like say something if, for them, it just works?
1483668015
David S.
Pro
Sheet Author
Thanks for the console log, David. It was very helpful in spotting an issue we haven't seen before. The log shows you attempting to connect to the 4 other people in the call but never receiving a response from any of them.&nbsp; If you're able to provide the "Video + Voice" settings and "Video Bandwidth" setting you and as many of your players used as you can provide, it would be very useful. Also, if you could run through this WebRTC connectivity test that information in conjunction with your log would be helpful. No response from any of them? Huh, odd. I could hear a couple of the folks there, just not all of them. We were all voice only for sending, for that time. I'll ask about the bandwidth and what the receiving was set at, but it may be a bit (I'm offline for a week starting this weekend). It was working great for us before the switch, on voice and video. Here's not one, but four different WebRTC links, each with different results. These were taken moments apart, and I changed nothing in between the tests. UDP randomness I'd expect, but not so much TCP. And the Throughput section didn't even seem to run, one of the times. This was all on the same hardware, in the same session of Chrome. I just opened up different tabs, and ran the next a little while after the previous finished. That last test showed no real issues that I could see, while the previous showed issues that I would guess would cause problems. <a href="https://test.webrtc.org/report/AMIfv94QCNfs2iJdgO4" rel="nofollow">https://test.webrtc.org/report/AMIfv94QCNfs2iJdgO4</a>... <a href="https://test.webrtc.org/report/AMIfv97PMpbUElw06P1" rel="nofollow">https://test.webrtc.org/report/AMIfv97PMpbUElw06P1</a>... <a href="https://test.webrtc.org/report/AMIfv95QTqR5gpFQMQq" rel="nofollow">https://test.webrtc.org/report/AMIfv95QTqR5gpFQMQq</a>... <a href="https://test.webrtc.org/report/AMIfv95hRQ1TQgA-bro" rel="nofollow">https://test.webrtc.org/report/AMIfv95hRQ1TQgA-bro</a>...
I think what Scoundrel said about having the devs set up a test game just for diagnostics sounds like a great idea. I don't want to take time away from my weekly game to troubleshoot when we've always used Mumble anyway (for the push to talk), but I'd be up for hopping into a room with a dev and some others to see how things go. I prefer using just Roll20 for my one-shots, which I normally run once or twice a month, but I'm putting them off in the hopes that this gets fixed relatively soon. Also, I took the survey, but the devs might want to consider adding in an option for other browsers. I use Vivaldi, but I said I used Chrome since it runs Chrome in the backend... But then again, so does Opera, which was also on the list. Keep plugging away at it!
Here's my WebRTC test: <a href="https://test.webrtc.org/report/AMIfv96JpgdCEkavPve" rel="nofollow">https://test.webrtc.org/report/AMIfv96JpgdCEkavPve</a>... The Reflexive Connectivity gave me an issue, everything else was fine.
Bad news and (potentially) good news (Bad) Last night our session started with everything working, but it went to whack-a-mole after about 10 minutes and we had to switch to Hangouts.&nbsp; (Possibly Good) I went digging around into WebRTC a little and found some indications that the 'Reflexive Connectivity' is actually the best way to connect, so the fact that so many people (including me) who are having problems say they have Reflexive Connectivity failures seems suspect. So I dug around some more and found a blog somewhere that indicated that freeing up UDP (inbound and outbound) could help. I went into Windows Firewall, and created new Inbound and Outbound rules to allow all UDP traffic from my browser of choice. Now, I retried the webRTC diagnostic, and Reflexive Connectivity succeeds! I haven't tested with my remote group yet, but at least I can now establish stable connections between the PC's in my house, which I could NOT do successfully before either. So, fingers crossed and I'll report back next Thursday at our next group session.
1483725380

Edited 1483725472
Much like "David S." four tests and slightly different results: <a href="https://test.webrtc.org/report/AMIfv97Qqk2MdWRCxbB" rel="nofollow">https://test.webrtc.org/report/AMIfv97Qqk2MdWRCxbB</a>... <a href="https://test.webrtc.org/report/AMIfv967A9mvzlo9GUP" rel="nofollow">https://test.webrtc.org/report/AMIfv967A9mvzlo9GUP</a>... <a href="https://test.webrtc.org/report/AMIfv976O0EK9z0FG3T" rel="nofollow">https://test.webrtc.org/report/AMIfv976O0EK9z0FG3T</a>... <a href="https://test.webrtc.org/report/AMIfv96A8GVvY0K9Mss" rel="nofollow">https://test.webrtc.org/report/AMIfv96A8GVvY0K9Mss</a>... Unlike Ken H. I am pretty clueless... :) Hope this helps!
I think what Scoundrel said about having the devs set up a test game just for diagnostics sounds like a great idea. I don't want to take time away from my weekly game to troubleshoot when we've always used Mumble anyway (for the push to talk), but I'd be up for hopping into a room with a dev and some others to see how things go. I prefer using just Roll20 for my one-shots, which I normally run once or twice a month, but I'm putting them off in the hopes that this gets fixed relatively soon. I'm obviously one of the whiners here, but I'd be more than happy to get into a game with a GM. Or to have a dev pop into one of the games I run. Anything I can do to help because the system really was great and I'd like to have it be reliable again.
1483726960

Edited 1483727485
Stephen Koontz
Forum Champion
Marketplace Creator
Sheet Author
API Scripter
Compendium Curator
Adam L. said: Here's my WebRTC test: <a href="https://test.webrtc.org/report/AMIfv96JpgdCEkavPve" rel="nofollow">https://test.webrtc.org/report/AMIfv96JpgdCEkavPve</a>... The Reflexive Connectivity gave me an issue, everything else was fine. Adam, it seems like your problem is an upload bandwidth issue. From your test log: {"ts":1483713149580,"name":"test-run","id":20,"args":{"success":"Transmitting at 15.124 kbps."}}, {"ts":1483713151577,"name":"test-run","id":20,"args":{"success":"Transmitting at 8.2 kbps."}}, {"ts":1483713152579,"name":"test-run","id":20,"args":{"success":"Transmitting at 8.176 kbps.”}}, 15 kbps is 0.015 Mbps You'll need access to about 1Mbps per other person in the chat to have a stable experience. What kind of results do you get from using&nbsp;<a href="http://www.speedtest.net/" rel="nofollow">http://www.speedtest.net/</a> for down and upload bandwidth?
My upload speed generally isn't that low; if I go do a speedtest at&nbsp; <a href="http://www.speedtest.net/" rel="nofollow">http://www.speedtest.net/</a> or any other monitoring site I come in at ~100 Mbps download and ~6 Mbps upload. Also, I don't have issues if I use Google Chat or Skype or appear.in (or the old OpenTOK, for that matter). Any idea why that would be happening here?
1483729361
Stephen Koontz
Forum Champion
Marketplace Creator
Sheet Author
API Scripter
Compendium Curator
Adam L. said: My upload speed generally isn't that low; if I go do a speedtest at&nbsp; <a href="http://www.speedtest.net/" rel="nofollow">http://www.speedtest.net/</a> or any other monitoring site I come in at ~100 Mbps download and ~6 Mbps upload. Also, I don't have issues if I use Google Chat or Skype or appear.in (or the old OpenTOK, for that matter). Any idea why that would be happening here? Skype isn't browser based, and therefore has access to hardware acceleration and data compression that Roll20 cannot take advantage of. Moreover, all of the options above use a All to One to All rather than a Peer to Peer connection. That means that when you use Skype, you send your video/voice data one time to the Skype servers and then the Skype servers carry the load of sending that data to each of the other people in the call. With Peer to Peer you are sending the data to each peer connection and thus has a higher upload bandwidth cost. That said, the US average upload speed in 2015 was 34Mbps which is well above WebRTC's needs. I would suggest investigating what your upload bandwidth promised to you by your ISP should be (I'm guessing it's higher than 6mbs). If you're unable to improve your upload bandwidth, I'd suggest using a 3rd party chat software that can take advantage of hardware acceleration and has servers that will carry your upload bandwidth load as part of their service.
1483730096

Edited 1483730166
Stephen Koontz
Forum Champion
Marketplace Creator
Sheet Author
API Scripter
Compendium Curator
David S. said: Here's not one, but four different WebRTC links, each with different results. These were taken moments apart, and I changed nothing in between the tests. UDP randomness I'd expect, but not so much TCP. And the Throughput section didn't even seem to run, one of the times. This was all on the same hardware, in the same session of Chrome. I just opened up different tabs, and ran the next a little while after the previous finished. That last test showed no real issues that I could see, while the previous showed issues that I would guess would cause problems. <a href="https://test.webrtc.org/report/AMIfv94QCNfs2iJdgO4" rel="nofollow">https://test.webrtc.org/report/AMIfv94QCNfs2iJdgO4</a>... David, your issues may be upload bandwidth related as well: {"ts":1483665212980,"name":"test-run","id":20,"args":{"success":"Transmitting at 105.824 kbps."}}, {"ts":1483665213984,"name":"test-run","id":20,"args":{"success":"Transmitting at 538.518 kbps."}}, {"ts":1483665215030,"name":"test-run","id":20,"args":{"success":"Transmitting at 383.389 kbps."}}, {"ts":1483665216032,"name":"test-run","id":20,"args":{"success":"Transmitting at 466.478 kbps."}}, {"ts":1483665217039,"name":"test-run","id":20,"args":{"success":"Transmitting at 430.73 kbps."}}, {"ts":1483665218253,"name":"test-run","id":20,"args":{"success":"Transmitting at 269.918 kbps."}}, {"ts":1483665219259,"name":"test-run","id":20,"args":{"success":"Transmitting at 627.646 kbps."}}, {"ts":1483665220270,"name":"test-run","id":20,"args":{"success":"Transmitting at 429.452 kbps."}}, Your upload connection is averaging less than half a MB, which could result in inconsistent performance. Could you run&nbsp; <a href="http://www.speedtest.net/" rel="nofollow">http://www.speedtest.net/</a> and let us know what the max upload/download bandwidth your connection can get. Dustin said: Much like "David S." four tests and slightly different results: <a href="https://test.webrtc.org/report/AMIfv976O0EK9z0FG3T" rel="nofollow">https://test.webrtc.org/report/AMIfv976O0EK9z0FG3T</a>... Dustin, your results are similar: {"ts":1483724945058,"name":"test-run","id":20,"args":{"success":"Transmitting at 90.022 kbps."}}, {"ts":1483724946110,"name":"test-run","id":20,"args":{"success":"Transmitting at 389.724 kbps."}}, {"ts":1483724947115,"name":"test-run","id":20,"args":{"success":"Transmitting at 610.736 kbps."}}, {"ts":1483724948126,"name":"test-run","id":20,"args":{"success":"Transmitting at 672.538 kbps."}}, {"ts":1483724949134,"name":"test-run","id":20,"args":{"success":"Transmitting at 593.27 kbps."}}, {"ts":1483724950141,"name":"test-run","id":20,"args":{"success":"Transmitting at 601.994 kbps."}}, Would you be able to provide the speed test for your down/upload bandwidth?
Per speedtest.net: Download = 1.4 Mbps, Upload = 0.8 Mbps testmyspeed.com: Download = 2.15 Mbps, Uploads = 0.87 Mbps
Steve K. said: 15 kbps is 0.015 Mbps You'll need access to about 1Mbps per other person in the chat to have a stable experience. What kind of results do you get from using&nbsp; <a href="http://www.speedtest.net/" rel="nofollow">http://www.speedtest.net/</a> for down and upload bandwidth? So Steve, is this just going to be the state of the video/voice system from now on? At minimum you need 1Mbps per other person you are playing with? &nbsp;Just want to understand expectations? &nbsp;Cause that might be the average in the USA, that isn't the case in all parts of the world. &nbsp;And spending another $20-$30 a month to get those kind of speeds just to use a system that used to work with the speeds I have doesn't make sense to me. Maybe that is what is needed to be posted. &nbsp;"Minimum requirements" for video/voice chat. &nbsp;And maybe that should be on the page where you pick which level of sub you want to do. &nbsp;I guess what many of us are looking for is some transparency here. &nbsp;You had to make the change fine. &nbsp;You rolled out your solution (poorly if you are honest with yourself) and it isn't as powerful. &nbsp;So be upfront with us, what do we need for internet speeds and computers to be able to run what your system needs to have video/voice. &nbsp;That way people can stop banging their head against the wall and just know to find another solution for video/voice and if need me, another solution for playing their RPG.
Steve K. said: Skype isn't browser based, and therefore has access to hardware acceleration and data compression that Roll20 cannot take advantage of. Moreover, all of the options above use a All to One to All rather than a Peer to Peer connection. That means that when you use Skype, you send your video/voice data one time to the Skype servers and then the Skype servers carry the load of sending that data to each of the other people in the call. With Peer to Peer you are sending the data to each peer connection and thus has a higher upload bandwidth cost. this is the best feedback I have seen. Thank you. I did some research a couple days ago and guessed this was what was going on...R20's deployment of WebRTC is actually a different architecture under the hood Skype, Discord, etc. This validates what I suspected. Thanks for the information...that means a lot to me.
Here are a few of my logs, all taken seconds apart and seeing completely different results. 1:&nbsp; <a href="https://test.webrtc.org/report/AMIfv95stUNVFPPyXsj" rel="nofollow">https://test.webrtc.org/report/AMIfv95stUNVFPPyXsj</a>... 2:&nbsp; <a href="https://test.webrtc.org/report/AMIfv97cY9503nHDgyr" rel="nofollow">https://test.webrtc.org/report/AMIfv97cY9503nHDgyr</a>... 3:&nbsp; <a href="https://test.webrtc.org/report/AMIfv96MaHuuaeSQozN" rel="nofollow">https://test.webrtc.org/report/AMIfv96MaHuuaeSQozN</a>... And here is the log of a group member who is helping me troubleshoot:&nbsp; <a href="https://test.webrtc.org/report/AMIfv95im7VMFRnge--" rel="nofollow">https://test.webrtc.org/report/AMIfv95im7VMFRnge--</a>... Before this change, he and I were the most stable members of the group, and now we can't even see or hear one another, even when it's just the two of us in the interface. Reconnect button and refreshes have both done nothing to alleviate the issues my entire group has been having since the transition. Just for good measure, I also took that survey. Hopefully something good comes of this.
1483744358
David S.
Pro
Sheet Author
speedtest.net results: 12ms ping, 5.10 Mbps down, 0.63 Mbps up. I'm on a 6/1 connection, so this is a typical result for me. This is enough for Google Hangouts to work for me, and the previous video/voice system here. Is this no longer the case?
I was able to connect to everyone in my group today (4 players and myself), however none of them were able to see or hear each other. Didn't grab a console log for myself but will be getting them all to run the diagnostic and send me their bug link. Used appear.in for the rest of the session with no issues.
Steve K. said: That said, the US average upload speed in 2015 was 34Mbps BTW - I'd be curious as to where you got this stat. Telecom is my industry; not sure this is true for consumer grade. Downloads...yea...uploads...I'd love to see the source report. Any chance you could share?
Scoundrel said: Steve K. said: That said, the US average upload speed in 2015 was 34Mbps BTW - I'd be curious as to where you got this stat. Telecom is my industry; not sure this is true for consumer grade. Downloads...yea...uploads...I'd love to see the source report. Any chance you could share? I just googled it, straight from the site that was suggested to use to test our speeds <a href="http://www.speedtest.net/reports/united-states/" rel="nofollow">http://www.speedtest.net/reports/united-states/</a> The speed above quoted is the average DOWNLOAD speed of Americans. &nbsp;Upload is below 20Mbps. Still good, but that being the average, I am sure there are a lot of people that don't have that much. &nbsp;Again if the metric quoted earlier about over 10 thousand users not having a problem the of course this means the edge cases are likely tied to lower than required upload speeds. &nbsp;So again I say, come out with the minimum requirements, instead of suggesting that might be the issue. Figure it out and just post it. Make it clear so people can move on from this. And as far as those metrics go, I would like see confirmation that just isn't the number of users logging into Roll20 for a session. &nbsp;Cause I know a lot of people that don't and never do bother with the voice/video option in roll20. &nbsp;So, is this metric based on log ins or actually from the WebRTC system counting connections?
<a href="https://test.webrtc.org/report/AMIfv97xJwN9RfBOM5oUZMl5PtSMXGbezp85dih9BOjGKHrBXStQfClbLNjp_FPXkUt3BD_puyXG3nswauI2P-1wB7U3cx_Ke2mVJh1Ldzm6qH8Hq3s_BdBxZ3Un2RIMwUU85Lm4vAD2i1gSkYjBpbSbER9DZshv7Q" rel="nofollow">https://test.webrtc.org/report/AMIfv97xJwN9RfBOM5oUZMl5PtSMXGbezp85dih9BOjGKHrBXStQfClbLNjp_FPXkUt3BD_puyXG3nswauI2P-1wB7U3cx_Ke2mVJh1Ldzm6qH8Hq3s_BdBxZ3Un2RIMwUU85Lm4vAD2i1gSkYjBpbSbER9DZshv7Q</a>
1483808927

Edited 1483809045
Scoundrel said: Steve K. said: That said, the US average upload speed in 2015 was 34Mbps BTW - I'd be curious as to where you got this stat. Telecom is my industry; not sure this is true for consumer grade. Downloads...yea...uploads...I'd love to see the source report. Any chance you could share? Like Scoundrel, I work in the telecom industry, and not for a small company: that "average" speed is nowhere near correct for consumers. Sure, if you are lucky enough to live in the cities where there is fiber (very few) or pay an outrageous amount of money for a MIS, you can. But that is by no means near available for the vast majority (90%+) of the US. Upload speeds average 3-8Mbps in the US. Unfortunately, fiber connections in those small pockets skew "average" results.