Guild Wars 2

Cogent, clear answer, thanks.

SolarNova.1052:

Habib Loew … i got a question.

please excuse any odities or wrongfull ‘tone’ of my text, im dyslexic

How are you going to be able to allow WvW to go without culling on your minimum spec machines when even with current culling and minimum settings ..and OC’s ..a minimum spec machine will run under 10 fps in WvW battles???
..hell even top of the line machiens with OC’s can barely keep minimum FPS over 30.

Your minimum specs are wrong …plain and simple ..you cant ‘play’ GW2 on those minimum specs. it will ‘run’ ..but ‘play’ ? ..not a chance.

I can only asume ur ‘Engine programmers’ have some “super awesome special” modifications coming that increase GW2 performance by over 100% ..becouse thats what its going to take, from the ‘players’ perspective, for GW2 to be playable on ur current minimum spec machine.
And if they do have these optimizations/modifcatiosn coming ..i hope it isnt done by removing eye candy and/or features of the game.

I am glad that ur superiors seem to be ok with increasing urservers bandwidth, i know that isnt cheap.

Anyone who’s been playing computer games for a while knows that minimum specs is never a guide line that should be followed when considering purchasing a game or computer. That is what it takes to quite simply run the game and play it a bit. By following the minimum specs and playing WvW you are asking to have performance issues. It is the same with almost every single other game out there put in a similar situation. With that being said, because most consumers don’t understand why it’s called minimum requirements the industry should probably remove that and just keep recommended specs. Even then, recommended specs in most cases means you can run most of the game medium settings and more intense situations requires a drop in quality and fps.

I know people don’t always have tons of money to spend on a gaming rig, but sometimes you have to accept the fact that you have to spend more money on one. You don’t even have to spend tons of money to get an adequate computer. If you ask anyone I know fairly well if they need help with figuring out their best possible option for a computer I’m always happy to help them. However, the world isn’t always fair you can’t expect to play every game on a cheap machine.

GW2 is easily one of the flashier mmos out there when it comes to particle effects and animations. Toning some of these down would help lower end machines client side. Even quite simply allow them to set it. Many other mmos have this option why not GW2.

Now, increasing the servers bandwidth usage or anything along the lines of network and server upgrades is the responsible thing for a business to do. If the service in inadequate it needs to be changed to something that can perform. A mmo not willing to do this deserves to lose business. I am glad to see that they are willing to do this. However, why is this an issue to begin with? Why go in knowing it can’t handle it and putting a ridiculous mechanic in place. Simply pony up and buy what is needed. This will put out the product that was promised and make for much more happy players.

I can understand the time needed for new content and knowing what game doesn’t have from the beginning. I along with many other people can not tolerate an inherently broken system like culling. I acknowledge the fact that wvw participation may not have been anticipated to have been so popular. However, if culling was not in to begin with, it would have been much easier to fix the problem.

And thank you habib for the response.

Habib Loew.6239:

3) Client performance issues related to rendering (potentially) all the players on a map at once. (Note that we base our performance requirements in this case on min-spec clients. We don’t want to stop anybody being able to play the game after all.)

I appreciate the post but this is dumb. Many people who own a decent computer aren’t playing the game now because it is unplayable. Let them turn down the rendering or something. I came back during the holidays to try the game out and it’s still invisi-wvw.

I’ve heard alot of people say that they don’t wvw cause they’re computers can’t handle it, so why would you base it off those systems, when they get 5-10 fps anyway in a big fight where culling matters the most?

Anyway, great news none the less.

Wonderful insight! Thanks for the answer and I look forward to more updates in the future!

Habib Loew.6239:

As you know we’ve been working on this problem for a while but what I think we haven’t ever said before is that our goal is to remove culling completely from WvW.

Thank you for the update. I’m sure this will spread quickly and positively. Communication always goes a long way.

Roe

I think it is sort of fun. I like to play ‘How many degrees to the left do I have to change my fov to make the 40 man group charging me disappear.?’ I like to mix it up with my other favorite game ‘I wonder what everyone is shooting at’.

But moving up the list fast is, ‘Why does my arrow stop there?’

Anywho easily the worst part of WvW cant wait for something.

So that leaves us with issue #3: client performance. Some time ago the WvW team acquired an engine programmer who is focused 100% on this issue (and he is being assisted by another engine programmer who isn’t officially on the WvW team). They’re working on some really fantastic optimizations and engine modifications which we hope will allow even min-spec clients to render all the players on a WvW map. We’ll be talking in more detail about the specific changes they’re making when things get just a little more nailed down, but I can say right now that I’m very impressed with the work they’ve done already.

Since client-side is your largest hurdle, why not add “culling” as a configurable slider under Graphic Settings? Ultimately FPS is at stake, so let the players decide if they want more players visible or better performance. You should, of course, first optimize to the best you can for minimum-performance machines, but in the event you can’t remove culling completely and are being held back by minimum-performance machines, make it a Graphic Setting like any other Graphic Setting that minimum-performance machines don’t have the requirements for. It’s not like they can’t play the game still, they just can’t see everyone, but don’t let them hold back the rest of your players from enjoying the game.

Call it “Limit Visible Players” or something.

The culling is seriously worse than what it was a month ago. Tired of zergs popping up out of thin air.

enji.7459:

Habib Loew.6239:

3) Client performance issues related to rendering (potentially) all the players on a map at once. (Note that we base our performance requirements in this case on min-spec clients. We don’t want to stop anybody being able to play the game after all.)

I appreciate the post but this is dumb. Many people who own a decent computer aren’t playing the game now because it is unplayable. Let them turn down the rendering or something. I came back during the holidays to try the game out and it’s still invisi-wvw.

Agreed it is a stupid concept. This is not catering to all specs. You have essentially added a horribly executed mechanic to cater to low specs machines while punishing higher end players.

Cactus.2710:

OK … I’ve scoffed harder than anyone at all the end-of-the world prophecies that have been floating around the last several months, but I’ve just witnessed the most conclusive sign ever that armageddon is imminent. ANet in the form of Habib just quickly responded with a candid, comprehensively descriptive, insightful, and encouraging post on a significant issue for GW2. I truly never thought I would see this day, but if I could buy Habib a case of Negra Modelo I most certainly would do so.

I have to admit I am just shocked as well. The information is pertinent and informative and gives the standing/rate of development.Well done.

I would also like to see the option of turning it on if possible and leave it up to player what is more desirable FPS vs Full Rendering

i don’t think the culling got worse but the servers got more organized

Lopez.7369:

Habib, will the fixes for culling in WvW work for PvE events too? Culling can be pretty annoying in the PvE world too.

The root cause of the problem has almost nothing to do with the content, so yes … I would expect whatever fix they have for culling to affect both PvE and PvP.

Habib,

Appreciate the detailed response, but it seems like it doesn’t address some important questions.

1) Why is there a 2-4 second delay between destealth and rendering in WvW, even under optimal circumstances (i.e. 1v1 out in the middle of nowhere, no other players present) This shouldn’t be ‘model loading’, because I’m talking about repeated stealth/destealth cycles in the same fight, when presumably all the assets for both characters are resident already. I’m playing on a high-end machine and I get this, and so does everybody I talk to. How do the changes you outline address this behavior?

2) Why is enemy player culling vastly less well-behaved in large fights than friendly player culling?

3) Can you say with a reasonable degree of confidence that the culling / visibility system is currently free of major bugs? Because it seems to me like it isn’t. I can be more specific, but it really looks to me like the model loading gets delayed or ‘stuck’ in some cases, and sometimes gets unstuck due to user input such as changing the view direction. This kind of behavior, combined with the seemingly random selection of enemies I see popping in and out during large fights, strongly suggests to me that the system is currently bugged. Are you doing anything to make sure the system really works the way it is supposed to work?

Thanks for looking at it, and here’s to hoping it gets fixed.

Habib Loew.6239:

Thank you, Grit, for posting that link. It points to a post in which I describe the culling issue in some detail and discuss a few of the issues involved with changing culling. An even more in depth explanation of what culling is can be found here https://forum-en.gw2archive.eu/forum/pvp/wuvwuv/The-real-problem-here-is-invisible-enemies-Give-their-algorithms-time-to-match-servers-properly/page/4#post356817 (be sure to read both that post and the next – my description exceeded the post size limit).

As you know we’ve been working on this problem for a while but what I think we haven’t ever said before is that our goal is to remove culling completely from WvW. In order to remove culling completely we have to address three issues:

1) Bandwidth out of our servers/datacenter (traffic would increase without culling)
2) Bandwidth in to each client (traffic to each client would also increase without culling)
3) Client performance issues related to rendering (potentially) all the players on a map at once. (Note that we base our performance requirements in this case on min-spec clients. We don’t want to stop anybody being able to play the game after all.)

Until all three of those issues have been dealt with we can’t turn culling off because doing so would cause something to break or perform poorly.

Ok, now let’s talk about what progress we’ve made!

Issue #1 was the easiest to deal with because we can basically just throw money at the problem. When we first started down the road toward removing culling from WvW I took a bunch of bandwidth measurements and then went to the executives and said, essentially, “Hey, if we disable culling our network traffic will increase by X%. Are we ok with that?”. The answer I got was a clear and unambiguous “yes!” So issue #1 isn’t a problem after all.

Issue #2 is a little harder. We need to ensure that folks with a min-spec network connection won’t be overwhelmed by the data we send them and we obviously can’t just buy a better connection for all of our players. So we put our heads together and came up with a plan to reduce the bandwidth required for WvW (and Gw2 in general) as much as possible. Those changes are in testing now and will be rolled out as soon as we’re convinced that they’re solid. Assuming we’re able to get everything working the way we’d like (and I’m fairly confident that we will) then this will address issue #2.

So that leaves us with issue #3: client performance. Some time ago the WvW team acquired an engine programmer who is focused 100% on this issue (and he is being assisted by another engine programmer who isn’t officially on the WvW team). They’re working on some really fantastic optimizations and engine modifications which we hope will allow even min-spec clients to render all the players on a WvW map. We’ll be talking in more detail about the specific changes they’re making when things get just a little more nailed down, but I can say right now that I’m very impressed with the work they’ve done already.

So that’s where we are. Engine programmers are working their magic even now and we’re testing the networking changes that will be required. I believe that our goal of removing culling from WvW is achievable and I’m looking forward to the day that I can announce to you all that we’ve pulled it off!

Good job really

I have a suggestion: Make the culling changes player optional.

e.g. I have 10k$ worth PC configuration and a solid 100 Mb/s international connection while some1 else have PC with lower configuration and slower internet so my PC can handle bigger data processing while the other pc cannot.

Make it so i can choose to enable rendering more players/data and use more Bandwidth while the other pc owner can disable/activate culling to not get his Bandwidth/PC burned out.

Putting an limitation option is a must – let players decided if they need culling or not depending on their PC configuration/Bandwidth capacity.

Good post.

However, since #1 is not a problem and #2 and #3 are entirely down to individual users’ connections and machines being up to scratch, then why not add an option in the game to disable culling while you work on making the rendering and netcode efficient?

You already have many, many performance related options that people can turn on and off, shadows, reflections, FSAA, etc etc. “WvW Player Culling” would simply be another option that people with powerful PCs could use?

Elochai.1280:

enji.7459:

Habib Loew.6239:

3) Client performance issues related to rendering (potentially) all the players on a map at once. (Note that we base our performance requirements in this case on min-spec clients. We don’t want to stop anybody being able to play the game after all.)

I appreciate the post but this is dumb. Many people who own a decent computer aren’t playing the game now because it is unplayable. Let them turn down the rendering or something. I came back during the holidays to try the game out and it’s still invisi-wvw.

Agreed it is a stupid concept. This is not catering to all specs. You have essentially added a horribly executed mechanic to cater to low specs machines while punishing higher end players.

What, something crazy like releasing the game in DX9 only so that Windows XP users could still play? You woudn’t see a AAA PC game doing that in this day and age, surely??

;)

Windows xp is an obsolete os that even microsoft has said they are going to stop supporting.

i’d put it on graphic effects as well… if you wanna play on Ultra details u need a good pc as well.. so you need for no culling a good internet connection simple as that.

No beating around the bush there. As open a response as we could have hoped for. Much appreciated.

i want to protest, because thieves are nerfed to the ground because of your culling problems !

DanH.5879:

i want to protest, because thieves are nerfed to the ground because of your culling problems !

nop. they got nerfed in spvp, not wvw. lol

teg.1340:

DanH.5879:

i want to protest, because thieves are nerfed to the ground because of your culling problems !

nop. they got nerfed in spvp, not wvw. lol

And the changes had nothing to do with culling anyways.

DanH.5879:

i want to protest, because thieves are nerfed to the ground because of your culling problems !

Theives are OP in WvW due to culling.

Go stealth, sneak up on zerg, launch daggerstorm, run away before you are un-culled.

2 theives working as a pair never become visible, the delay between them being un-stealth and un-culled is enough to let them be pretty much permanently invisible.

Chamone.6890:

Go stealth, sneak up on zerg, launch daggerstorm, run away before you are un-culled.

Not to make this thread about thieves as well, but this is a terrible tactic and a bad example of exploiting culling.

And you don’t need two thieves to permastealth.

Tulisin.6945:

Chamone.6890:

Go stealth, sneak up on zerg, launch daggerstorm, run away before you are un-culled.

Not to make this thread about thieves as well, but this is a terrible tactic and a bad example of exploiting culling.

And you don’t need two thieves to permastealth.

There are much worse examples of exploiting culling which I won’t go into here. I was responding to him saying thieves were nerfed due to culling, when in fact the opposite is true.

Sure, you don’t need two thieves, but it makes it a heck of a lot easier. Me and my buddy can run around permastealth, do kitten-hot damage and res one another if we do happen to get downed. We would not be able to do this if there was not a big delay between stealth running out and actually being visible to the enemy.

Perma-stealthing alone requires you to sacrifice far too much damage.

In the meantime, can we go back to the experimental alternative method you had a while back? It was so much better.

Nabrok.9023:

In the meantime, can we go back to the experimental alternative method you had a while back? It was so much better.

I disagree on this completely. I often found I couldn’t see my allies and we couldn’t coordinate anything. We never knew if each other were pushing in or pulling back. Frankly, they both are terrible. I’d rather see all my allies for more coordination and get a few of them to show up first.

Burl.9567:

Nabrok.9023:

In the meantime, can we go back to the experimental alternative method you had a while back? It was so much better.

I disagree on this completely. I often found I couldn’t see my allies and we couldn’t coordinate anything. We never knew if each other were pushing in or pulling back. Frankly, they both are terrible. I’d rather see all my allies for more coordination and get a few of them to show up first.

You could still see the green dots on the mini map for your allies.

It’s way worse when you get the door down, charge into a keep and the only indication there’s any enemy there at all is the red circles that appear at your feet. Then as you’re sitting there dead “Holy kitten, there’s a lot of bad guys here”.

Great post. Well organized, factual, and informative. Can you post more often please, this is the kind of post players are looking for.

GuardianOMS.8067:

Kuldred.2436:

I came from playing WoW, spending a majority of my time doing the open world PvP (Wintergrasp, Tol Barad), and was especially excited about WvW in GW2. I am still to this day absolutely amazed at what I CAN’T see during the battles. No matter what video settings I run with, I can only see half the distance I could in WoW, and the fact that entire groups of enemy players can run right by me either invisible or without ID tags above thier heads is shocking.

I thought WoW was based on a much older engine, yet could handle this much better?

I know it’s only been a few months, but if we have another large content update with added ‘fluff’, new gem store stuff, new class tweaks and even new areas or stuff to do, and the culling STILL isn’t fixed, I’m going to have a hard time spending my free hours in WvW.

You’re really silly if you think comparing a graphically intensive game like GW2 is the same as WoW, a game that can literally run on toasters (LITERALLY). There are many things to consider when trying to explain the culling issue, the fact that WoW is not handling, sending, and receiving the same amount and weight of information to scores of players at once AND in a high resolution (regardless of your settings) is one of them.

Both games are not created equally or by the same minds. Don’t expect it work like it is. Not trying to sound snobby, just saying.

You completely missed my point, so I guess I did a poor job of explaining. You didn’t get culling in the WoW battles because it was done within the constraints of thier engine/system. I’m fully aware of the less graphics, textures, players, blah blah.

What I want to know is why ANET didn’t design WvW within the limits of thier system? What they have set up obviously cannot handle what is actually happening. I ran WoW on a GTX260, now I’m using a 560ti and have culling pretty much everywhere. So this is my computers fault?

Habib Loew.6239:

our goal is to remove culling completely from WvW. In order to remove culling completely we have to address three issues.

Wow! This is great news. Thank you so much for giving us such a detailed status report and also for working towards the best possible solution!

Habib Loew.6239:

Thank you, Grit, for posting that link. It points to a post in which I describe the culling issue in some detail and discuss a few of the issues involved with changing culling. An even more in depth explanation of what culling is can be found here https://forum-en.gw2archive.eu/forum/pvp/wuvwuv/The-real-problem-here-is-invisible-enemies-Give-their-algorithms-time-to-match-servers-properly/page/4#post356817 (be sure to read both that post and the next – my description exceeded the post size limit).

As you know we’ve been working on this problem for a while but what I think we haven’t ever said before is that our goal is to remove culling completely from WvW. In order to remove culling completely we have to address three issues:

1) Bandwidth out of our servers/datacenter (traffic would increase without culling)
2) Bandwidth in to each client (traffic to each client would also increase without culling)
3) Client performance issues related to rendering (potentially) all the players on a map at once. (Note that we base our performance requirements in this case on min-spec clients. We don’t want to stop anybody being able to play the game after all.)

Until all three of those issues have been dealt with we can’t turn culling off because doing so would cause something to break or perform poorly.

Ok, now let’s talk about what progress we’ve made!

Issue #1 was the easiest to deal with because we can basically just throw money at the problem. When we first started down the road toward removing culling from WvW I took a bunch of bandwidth measurements and then went to the executives and said, essentially, “Hey, if we disable culling our network traffic will increase by X%. Are we ok with that?”. The answer I got was a clear and unambiguous “yes!” So issue #1 isn’t a problem after all.

Issue #2 is a little harder. We need to ensure that folks with a min-spec network connection won’t be overwhelmed by the data we send them and we obviously can’t just buy a better connection for all of our players. So we put our heads together and came up with a plan to reduce the bandwidth required for WvW (and Gw2 in general) as much as possible. Those changes are in testing now and will be rolled out as soon as we’re convinced that they’re solid. Assuming we’re able to get everything working the way we’d like (and I’m fairly confident that we will) then this will address issue #2.

So that leaves us with issue #3: client performance. Some time ago the WvW team acquired an engine programmer who is focused 100% on this issue (and he is being assisted by another engine programmer who isn’t officially on the WvW team). They’re working on some really fantastic optimizations and engine modifications which we hope will allow even min-spec clients to render all the players on a WvW map. We’ll be talking in more detail about the specific changes they’re making when things get just a little more nailed down, but I can say right now that I’m very impressed with the work they’ve done already.

So that’s where we are. Engine programmers are working their magic even now and we’re testing the networking changes that will be required. I believe that our goal of removing culling from WvW is achievable and I’m looking forward to the day that I can announce to you all that we’ve pulled it off!

Wow, that was a great post! Just one question though on number 3. You talk about having the requirement of the new culling system to be able to run on min spec hardware; is there any way you could implement a “culling slider” that would adjust the amount of culling that happens on the clients side? So for users that have capable hardware to render every player can max out that slider but for users running less than optimal hardware can just adjust that slider downwards until they find a capable trade-off. This seems rather reasonable to me because I find it unfair that those with high performance hardware would have to suffer because of min spec users.

With that said though, if you find a way to allow min spec users to have absolutely no culling, hence no slider being necessary(and it seems like you’re almost there!) then that is obviously the best solution for everyone.

Krakah.3582:

2 – Should be moot the the vast majority of users are on broadband circuits, very few are playing on dial-up. Hell cell phone data plans are faster than dial-up.

3 – Note this is a personal preference, if I had a mid to low end rig, I’d trade hands down the overwhelming amount of particle effects to have player models drawn, even if it’s just colored Gumby avatars. I would love to see options to lower or selectively disable enemy or friendly particle effects to some super base level or not show some at all.

To your preference at point 3:
Personally I only encounter culling issues when I smash into zergs of 35+ and my laptop is 10 years old and below minimum hardware requirements. What’s most annoying to me is that, due to playing on absolutely the lowest settings and without proper shader support, I can’t see important things like combo field boundaries, targeting circles, red rings, the particle clouds that denote pathing for some personal stories (example: Captain Weyandt’s spectral trail doesn’t render for me), etc.

I personally would rather have a version of skill effects for crap-level systems that utilizes minimalist resources, as it’s only Guild Wars 2 that this computer has those types of issues with. I’m sure that tournament gamers would prefer such minimalist features as well, and it could potentially sort some of the culling issues out.

I think the answer is providing super low quality options/support as well as the super high quality support rather than altering the game to only appeal to computers closer to my end of the spectrum.

Nabrok.9023:

Burl.9567:

Nabrok.9023:

In the meantime, can we go back to the experimental alternative method you had a while back? It was so much better.

I disagree on this completely. I often found I couldn’t see my allies and we couldn’t coordinate anything. We never knew if each other were pushing in or pulling back. Frankly, they both are terrible. I’d rather see all my allies for more coordination and get a few of them to show up first.

You could still see the green dots on the mini map for your allies.

It’s way worse when you get the door down, charge into a keep and the only indication there’s any enemy there at all is the red circles that appear at your feet. Then as you’re sitting there dead “Holy kitten, there’s a lot of bad guys here”.

If you are balls to the walls rushing into a keep, you are doing it wrong.

Gilgamesh.2561:

So that leaves us with issue #3: client performance. Some time ago the WvW team acquired an engine programmer who is focused 100% on this issue (and he is being assisted by another engine programmer who isn’t officially on the WvW team). They’re working on some really fantastic optimizations and engine modifications which we hope will allow even min-spec clients to render all the players on a WvW map. We’ll be talking in more detail about the specific changes they’re making when things get just a little more nailed down, but I can say right now that I’m very impressed with the work they’ve done already.

Since client-side is your largest hurdle, why not add “culling” as a configurable slider under Graphic Settings? Ultimately FPS is at stake, so let the players decide if they want more players visible or better performance. You should, of course, first optimize to the best you can for minimum-performance machines, but in the event you can’t remove culling completely and are being held back by minimum-performance machines, make it a Graphic Setting like any other Graphic Setting that minimum-performance machines don’t have the requirements for. It’s not like they can’t play the game still, they just can’t see everyone, but don’t let them hold back the rest of your players from enjoying the game.

Call it “Limit Visible Players” or something.

I still don’t understand why we don’t have option to turn off spell effects…i don’t need to see fireballs fly around…and it would boost up fps so much in WvW =(

i wonder if some of these ‘fixes’ will be in the Feb WvW update.

I can easily see a big change in the avatars/character models being made. If there is less data to load per character, they should load faster on less optimal computer, no? If one of the changes in feb is WvW specific armor appearances (with a few options dependent on rank, also new addition) then this could be one of the fixes we will see soon. just maybe

Burl.9567:

Nabrok.9023:

Burl.9567:

Nabrok.9023:

In the meantime, can we go back to the experimental alternative method you had a while back? It was so much better.

I disagree on this completely. I often found I couldn’t see my allies and we couldn’t coordinate anything. We never knew if each other were pushing in or pulling back. Frankly, they both are terrible. I’d rather see all my allies for more coordination and get a few of them to show up first.

You could still see the green dots on the mini map for your allies.

It’s way worse when you get the door down, charge into a keep and the only indication there’s any enemy there at all is the red circles that appear at your feet. Then as you’re sitting there dead “Holy kitten, there’s a lot of bad guys here”.

If you are balls to the walls rushing into a keep, you are doing it wrong.

Of course. Everybody knows that when the outer wall door goes down and you don’t see anybody on the other side, you just wait outside.

I am very curious why this game has such large bandwidth requirements compared to other games, when I don’t notice such a huge difference between it and other games. Really the ‘limit’ where issues start to occur seems to be just a relative handful of people compared to other games… not sure if it’s been discussed.

Also, there is a claim it works well in PvE, but that’s not the case. Temple of Grenth, for example, has issues and wipes a lot because people can’t see the ‘adds’ when the spawn. (Again, it seems to happen with a relatively small number of people).

I understand you need to defend decisions, I guess, but this like so many other design decisions seems based on the fact that you thought small fights like SPvP 8v8 and 5v5 were going to be more popular and “serious” formats and large event PvE and WvWvW were less important…. but what happens when you are proven wrong like it seems you are going to be? I think everybody in WvWvW would rather it be like most games and take update rate and framerate hits….

This might not be possible, but can’t you just make a client side setting where someone can essentially turn “culling back on” and have their game throttle info if their min spec pc can’t handle it? Rest of us can just have no culling then if our pc can handle it?

Habib: The consensus here is pretty clear, since many people are suggesting essentially the same thing: that culling should be placed in settings where the user can control it. Those who have bandwidth or rendering issues can simply leave culling on and have the exact same performance they get now. Those who have systems that can handle the increased bandwidth/rendering would have the option to turn off culling completely. If this is at all possible, please consider implementing this.

Kintari.4172:

Habib,

Appreciate the detailed response, but it seems like it doesn’t address some important questions.

1) Why is there a 2-4 second delay between destealth and rendering in WvW, even under optimal circumstances (i.e. 1v1 out in the middle of nowhere, no other players present) This shouldn’t be ‘model loading’, because I’m talking about repeated stealth/destealth cycles in the same fight, when presumably all the assets for both characters are resident already. I’m playing on a high-end machine and I get this, and so does everybody I talk to. How do the changes you outline address this behavior?

2) Why is enemy player culling vastly less well-behaved in large fights than friendly player culling?

3) Can you say with a reasonable degree of confidence that the culling / visibility system is currently free of major bugs? Because it seems to me like it isn’t. I can be more specific, but it really looks to me like the model loading gets delayed or ‘stuck’ in some cases, and sometimes gets unstuck due to user input such as changing the view direction. This kind of behavior, combined with the seemingly random selection of enemies I see popping in and out during large fights, strongly suggests to me that the system is currently bugged. Are you doing anything to make sure the system really works the way it is supposed to work?

Thanks for looking at it, and here’s to hoping it gets fixed.

I’m not sure I totally understand how culling is applied in GW2, but this is what I think happens:

I’m pretty sure the culling algorithm simply excludes other characters from your view that the game thinks aren’t important >>for you to view<< at that moment. That would include another character that is too far away or sufficiently behind you that you can’t normally see it. If that character suddenly pops into range or into view, the game tries to render it but there is a significant lag in doing so. So …

1. I suspect that the game does not always properly differentiate between a character that is truly out of view and one that went stealth. If so, it doesn’t always matter what else is going on or whether there are many players or few in the area. The lag comes from the server and your game client (the portion resident on your computer) trying to agree on what you need to see.

2. My guess would be that enemy players pop into your view from further away than do friendly players. The culling algorithm probably has a priority system based upon the range from you for deciding which players to cull and which not. Players closer to you but still culled may have some of their data queued waiting to be sent to you while players further away may not. An enemy further away with a teleport ability may pop into range sooner than a closer enemy that has to run to you, thereby screwing up the culled data queue. That kind of thing probably happens less often for friendlies in a mass battle than for enemies.

3. I think the behavior of culling when quickly rotating the camera should be obvious … players that the game didn’t think you needed to see suddenly have to be rendered and the information sent to you. If there are a lot of characters in the area and you swing the camera back and forth quickly, data packets in transit suddenly are no longer needed and your game client needs to request a different view. It’s not difficult to imagine temporary confusion occurring between your computer and the server while it tries to decide what you really want to see … kind of like if I order a meal at a restaurant but then keep changing my order before anything gets cooked.

So I don’t think what you are observing are necessarily “bugs” with culling … they are instead a natural result of culling that would most likely go away if culling does.

elderan.2638:

Habib: The consensus here is pretty clear, since many people are suggesting essentially the same thing: that culling should be placed in settings where the user can control it. Those who have bandwidth or rendering issues can simply leave culling on and have the exact same performance they get now. Those who have systems that can handle the increased bandwidth/rendering would have the option to turn off culling completely. If this is at all possible, please consider implementing this.

No thanks, my system is way above min specs, but giving me and anyone else the option to turn it off and thereby creating a two tier PvP system with those that can see other players and those that can’t, is simply a bad idea. (nor for that matter would it necessarily solve the issues)

elderan.2638:

Habib: The consensus here is pretty clear, since many people are suggesting essentially the same thing: that culling should be placed in settings where the user can control it. Those who have bandwidth or rendering issues can simply leave culling on and have the exact same performance they get now. Those who have systems that can handle the increased bandwidth/rendering would have the option to turn off culling completely. If this is at all possible, please consider implementing this.

I don’t that that is practical, or even desirable. If anybody who didn’t want to use culling could just turn it off completely, ANet would have no control at all over how much bandwidth was being required at any one time for their servers. Habib said that ANet was willing to buy additional bandwidth (which they OBVIOUSLY really need), but he didn’t said that ANet was willing to buy infinite amounts of it. If ANet was not able in some manner to dynamically control the volume of input/output data being required from their systems (which is exactly what culling is there for), the game could literally lock up as if ANet was the target of a denial-of-service-attack.

Sylosi.6503:

elderan.2638:

Habib: The consensus here is pretty clear, since many people are suggesting essentially the same thing: that culling should be placed in settings where the user can control it. Those who have bandwidth or rendering issues can simply leave culling on and have the exact same performance they get now. Those who have systems that can handle the increased bandwidth/rendering would have the option to turn off culling completely. If this is at all possible, please consider implementing this.

No thanks, my system is way above min specs, but giving me and anyone else the option to turn it off and thereby creating a two tier PvP system with those that can see other players and those that can’t, is simply a bad idea.

^ this too ^

Sylosi.6503:

elderan.2638:

Habib: The consensus here is pretty clear, since many people are suggesting essentially the same thing: that culling should be placed in settings where the user can control it. Those who have bandwidth or rendering issues can simply leave culling on and have the exact same performance they get now. Those who have systems that can handle the increased bandwidth/rendering would have the option to turn off culling completely. If this is at all possible, please consider implementing this.

No thanks, my system is way above min specs, but giving me and anyone else the option to turn it off and thereby creating a two tier PvP system with those that can see other players and those that can’t, is simply a bad idea. (nor for that matter would it necessarily solve the issues)

Agreed. Creating a game where the “haves” have a clear advantage over the “have nots” is bad for business and bad for the game. People will just tune out and quit playing if their system wasn’t “up-to-snuff”.

Sylosi.6503:

elderan.2638:

Habib: The consensus here is pretty clear, since many people are suggesting essentially the same thing: that culling should be placed in settings where the user can control it. Those who have bandwidth or rendering issues can simply leave culling on and have the exact same performance they get now. Those who have systems that can handle the increased bandwidth/rendering would have the option to turn off culling completely. If this is at all possible, please consider implementing this.

No thanks, my system is way above min specs, but giving me and anyone else the option to turn it off and thereby creating a two tier PvP system with those that can see other players and those that can’t, is simply a bad idea. (nor for that matter would it necessarily solve the issues)

PC gaming has always been like this: There are people who buy or build top-end systems who have higher framerates, better rendering, larger FOV, better sound (and sound localization), more accurate mice, faster texture load times, more bandwidth and lower latency, faster memory and disk access, etc. etc. etc.

And then there are people who buy mediocre or low-end systems, or just never bother upgrading their years-old machine. And they have low framerates, stuttering, high latency, dropped packets, slow rendering, slow disk and memory access, crappier sound, poor mouse tracking, and so on.

It’s true in all games.

And it wouldn’t be a “two-tier” situation. It’d be a spectrum: some would be able to render everything all the time, some would be able to render most things most of the time, some would render some things some of the time, and some would have trouble rendering even a handful of things all the time.

I’d MUCH rather be able to take advantage of the machine I intentionally spent thousands of dollars building by hand, from scratch for the sole purpose of getting the best possible performance out of games, than be artificially prevented from getting even playable performance out of a game (a game recently voted Best of 2012 by several outlets) because the developers refuse to remove that artificial hindrance until even the barest minimum-spec machine is capable of rendering everything all the time.

Just turn off culling. If you didn’t pay the price to have a decent machine and/or connection, you pay the price in-game due to poor performance. This has been a simple fact of PC gaming for as long as there have been PC games that were capable of taxing the systems they ran on.

If you want everyone to have an identical gaming experience, then everyone needs to have identical hardware. That’s one reason why consoles exist. If you want a console experience, lobby ANet to create an MMO for consoles.

Burl.9567:

Agreed. Creating a game where the “haves” have a clear advantage over the “have nots” is bad for business and bad for the game. People will just tune out and quit playing if their system wasn’t “up-to-snuff”.

Explain, please, how that’s preferable to having the entire player base suffer from invisible enemies, making both the “haves” AND “have nots” tune out and quit playing?

mcl.9240:

Just turn off culling. If you didn’t pay the price to have a decent machine and/or connection, you pay the price in-game due to poor performance. This has been a simple fact of PC gaming for as long as there have been PC games that were capable of taxing the systems they ran on.

If you want everyone to have an identical gaming experience, then everyone needs to have identical hardware. That’s one reason why consoles exist. If you want a console experience, lobby ANet to create an MMO for consoles.

So you want ANet to essentially exclude a large portion of the gaming population from participating in WvW (face it most of the people that play are casual) so you can look at pretty graphics? Bad business decision. I’m not saying it doesn’t need to get fixed, but this definitely isn’t the long term solution either. A slider for the short term would be OK, but long term this would destroy WvW.

mcl.9240:

Burl.9567:

Agreed. Creating a game where the “haves” have a clear advantage over the “have nots” is bad for business and bad for the game. People will just tune out and quit playing if their system wasn’t “up-to-snuff”.

Explain, please, how that’s preferable to having the entire player base suffer from invisible enemies, making both the “haves” AND “have nots” tune out and quit playing?

I’m not sure why I need to explain human nature here. When one side has an advantage that the other can’t achieve, they will quit. The culling isn’t bad to the point of unplayable either. Quit making it out into this huge issue. I have it happen maybe twice a week where I walk into a zerg (and sometimes through it). I can cough up the silver for now…