Yeah i noticed the same thing a long time ago OP, and reporting it in detail to ZOS.
Its all about sync and timing within the client. By the time its sent the sound to your soundcard its already on the next 10 down the road, they pile up and get out of sync with whats actually happening, the client tries to compensate by forcing video rendering to slow down causing the stutter effect, and eventually things get so out of line it just freezes until it finds a new start point or crashes completely.
Funny enough one of my tests was actually disabling sound on my machine instead of just in the client. Same exact effect only the client crashed and threw a sound error about 20 seconds in to a big battle.
Coincidence? I dont think so.
Yeah i noticed the same thing a long time ago OP, and reporting it in detail to ZOS.
Its all about sync and timing within the client. By the time its sent the sound to your soundcard its already on the next 10 down the road, they pile up and get out of sync with whats actually happening, the client tries to compensate by forcing video rendering to slow down causing the stutter effect, and eventually things get so out of line it just freezes until it finds a new start point or crashes completely.
Funny enough one of my tests was actually disabling sound on my machine instead of just in the client. Same exact effect only the client crashed and threw a sound error about 20 seconds in to a big battle.
Coincidence? I dont think so.
Well it seems I'm not the only one to see this issue then. Maybe we can hope they can fix it now.
Very interesting that globally disabling sound would cause such a hard crash. Seems like your able to easily reproduce this, wonder why it hasn't been fixed?
Has anyone attempted to run the program/process at a higher priority?
Yeah i noticed the same thing a long time ago OP, and reporting it in detail to ZOS.
Its all about sync and timing within the client. By the time its sent the sound to your soundcard its already on the next 10 down the road, they pile up and get out of sync with whats actually happening, the client tries to compensate by forcing video rendering to slow down causing the stutter effect, and eventually things get so out of line it just freezes until it finds a new start point or crashes completely.
Funny enough one of my tests was actually disabling sound on my machine instead of just in the client. Same exact effect only the client crashed and threw a sound error about 20 seconds in to a big battle.
Coincidence? I dont think so.
ZOS_GinaBruno wrote: »Yeah i noticed the same thing a long time ago OP, and reporting it in detail to ZOS.
Its all about sync and timing within the client. By the time its sent the sound to your soundcard its already on the next 10 down the road, they pile up and get out of sync with whats actually happening, the client tries to compensate by forcing video rendering to slow down causing the stutter effect, and eventually things get so out of line it just freezes until it finds a new start point or crashes completely.
Funny enough one of my tests was actually disabling sound on my machine instead of just in the client. Same exact effect only the client crashed and threw a sound error about 20 seconds in to a big battle.
Coincidence? I dont think so.
Could you list out the exact steps you took to disable your sounds (both in-game and on your machine) so we can try and reproduce this? That would help us a lot. Thanks!
ZOS_GinaBruno wrote: »Thanks so much, @Rylana! We'll pass all this along now.
saintmurray wrote: »Wow that's some really good info maybe this is contributing to all our problems? Could it be a backlog of sounds in an area where there are lots of people making sounds creating a backlog on the megaservers processor? I'm no computer expert but it seems we have all been blaming aoe spamming (the light generated/lighting patch), when rather maybe its a sounds that aoes make...
I think its exactly that but not just sounds. Overall clearing of previous actions seems to be a global thing.
It is as if the game just keeps on chugging forward and never clears the back end. Sounds that were still queued to play, visuals that never got cleared from memory.
Essentially the game just caches actions, they get stored in whatever medium that is in use (some to volatile memory, some to pagefile.) People with SSDs even report similar issues, so from what I can tell, its the client not cleaning up after itself.
Now granted, the older your machine, the more out of date the hardware, the worse the experience, but no one I have spoken to thus far has ever reported a seamless game, even with monster rigs that were built last year.
I think its exactly that but not just sounds. Overall clearing of previous actions seems to be a global thing.
It is as if the game just keeps on chugging forward and never clears the back end. Sounds that were still queued to play, visuals that never got cleared from memory.
I bet that's the problem, and I think I can understand why.
PvE, you're constantly in and out of instances. Quests pull you back and forth, from dungeons, to the topside, from one quest phase to another. Each time, the game essentially reloads, thus your character gets pulled away from one instance/phase to another. The surrounding world, your skills, all of it gets a quick refresh.
Can't do that in Cyrodiil, It'd be exploited. (Remember /stuck in beta?) Your toon would be plucked away into thin air before everyone's eyes.
I bet it isn't a matter of necessarily stopping the backlog, it's just the measures they have in place that normally do it are suppressed in Cyrodiil as it'd upset the mechanics of PvP.
This likely ties into the transitus system inside Cyrodiil, and plays a part into the removal of forward camps. They were a mobile version of this transitus system, and often when I either res or port to a transitus shrine, I hear a jumble of sounds (usually like a clunk) as if all my backlogged sounds are pushed through at once. As you pointed out, I bet it isn't limited to sound, but all effects. If my particle range were infinite, I bet sparkles and beams of light would process at Bleakers while I rez at Alessia.
I suspect the attempt is to use the transitus system to in part do a 'refresh' of the environment to your client. It seems pretty successful as usually any performance issues start to clear up. I can see where Forward Camps utilizing this feature created unnecessary overhead on the server end to ensure those constant rezzes were refreshing.
If we rewind all the way back to beta some of us may recall that for a while there, porting in Cyrodiil was just flat out impossible. You'd get stuck in a long loading screen. PvE side things were dandy. Cyrodiil is one huge singular instance (dungeons aside) whereas all the PvE places are segmented.
If the in-game house keeping is done in part through porting, it would explain why lots of us suffer these glitches, since Cyrodiil is massive, we see significantly more action in it, and port less on a 'suppressed' transport system. (So as not to allow player exploitation)
I feel sorry for the poor soul who has to figure this issue out.
(edit, some words, grammar, punctuation, coffee)
@Rylana During your testing of toggling sound on and off in the bigger battles, at any point does the sound ever seem to "catch up" to the action and fire off a dozen or so sound FX at once?
This is what I keep getting;
I'm in Cyrodiil and I engage in a fairly large battle, say... a resource capture with a few enemy players guarding it.
All of the sudden the sound just quits in the middle of battle.
The game continues, laggy as usual, sometimes for 10 or 15 minutes after the sound quits.
Then comes the BIG lag spike where the game freezes completely for around 3-5 seconds, and BOOM the sound effects for every ability I've cast since the sound quit, fire off simultaneously!
I didn't have this issue until the last update.
snipped to save space