Closed Beta Update V39035 (25th September 2014)

(Silvanoshi) #161

Spookify,

Indeed, playing on a server located on another continent won’t provide the best gaming experience. While we still have optimisations to carry out, many of the issues that you’re experiencing are likely to be compounded by the latency that you’re playing with.

Performance is a tricky beast, the problem generally arises from many smaller issues that add up, rather than one larger issue that can be addressed in one fell swoop. We’ll be continuing to make optimisations for some time to come, we’re certainly not settling with what we have.

If you’re going to do a framerate comparison, it really makes a lot of sense to have your setup configured in the same manner. Comparing results from two different configurations when you’re trying to ascertain how the number of players affects performance isn’t really a great way to go about it. Don’t get me wrong, I’m not discounting the fact that additional players on a server can impact performance, but it’s much more useful for all involved to get info from configs that can don’t have any differentiators in them.

Animations in general ehh? In general, animators could tweak and tune animations for the rest of their lives and still have more than they would like to adjust. For the most part, we’re pretty happy with where we are atm.

With regards to defibs, I don’t think what you’re getting at is actually an issue with the animation there. You just need to be pinpoint accurate with defibs. That’ll likely change at some stage in the future, but it’s certainly not top of the list right now. For the time being, you’d be best advised to go for centre mass with them…reviving is just slightly trickier skill to master atm ;).

Player collision, what specifically are you talking about here? Many of the collision issues are with the maps rather than players, and no map collision isn’t final. Getting the final collisions in place doesn’t make sense to do until the maps themselves have been finalised, and since most are due to receive tweaks, it’s going to be a while before we reach that stage.

(PixelTwitch) #162

I would like to do a performance test by locking everyone bar myselfs framerate to 60fps (just for a test).

I have a feeling the higher that everyones framerate is the lower that everyones framerate is.
My reason for feeling this is due to the way that the game slows down fps based on network traffic. I feel that loweing the badwidth used would actually increase the framerate. I could be wrong but I know that if you limit your network usage in the config it changes your FPS greatly. Simply halfing the max bandwidth can give you as little as 20% the normal max framerate.

This is likely the reason behind the player connection stutter also and things like AirStrikes. Simply making the airstrike effect larger but also slower would remove the airstrike stutter while keeping its effectiveness just as high…

There is a way to get the MG nests to fire at 10x the firerate and when you do this the whole server almost comes to a crawl. Likely due to the number of traces and hit results its sending to all the players on the server.

I know its not ideal and not a long term solution. But if you was to increase the server tick rate to 120fps JUST FOR A TEST. I am pretty sure we could see a higher overall framerate (with higher bandwidth usage) and a smoother experiance.

(Voshy) #163

[QUOTE=Silvanoshi;512219]Spookify,

Indeed, playing on a server located on another continent won’t provide the best gaming experience. While we still have optimisations to carry out, many of the issues that you’re experiencing are likely to be compounded by the latency that you’re playing with. [/QUOTE]

Things spooky is talking about have nothing to do with continents and/or latency. Server still lags for a second when someone joins, playing with more people affects FPS alot.

[QUOTE=Silvanoshi;512219]With regards to defibs, I don’t think what you’re getting at is actually an issue with the animation there. You just need to be pinpoint accurate with defibs. That’ll likely change at some stage in the future, but it’s certainly not top of the list right now. For the time being, you’d be best advised to go for centre mass with them…reviving is just slightly trickier skill to master atm ;).
[/QUOTE]

It’s not a trickier skill to master, its simply bad at the moment. Crouching while reviving helps a lot, but trying to revive standing up is a wheel of fortune every time.

Even if you feel it’s about reviving being a trickier skill, you should change it. You already made medics worse than they should be with other decisions, I think at least reviving should be something foolproof.

(prophett) #164

[QUOTE=tokamak;512184]
Good side objectives give attackers more options rather than merely faster paths towards the same end.[/QUOTE]

Also on the defensive side, you also need to make decisions on what to defend. A 4/2 or 3/3 split, and positioning the defense so you can easily rotate to cover. I really hope they invest time in designing better side objectives, rather than the “faster paths towards the same end” you mentioned. The value they bring to the table in terms of opportunities for more strategy/depth has been overlooked thus far :[

(tokamak) #165

Yes! Back in the days one of the most important strategies was to put your opponent on their heels by forcing them to split. Rather than pushing them slowly back through a single route you could invest in taking detours and undermine them through entirely different angles.

The current side objectives mostly serve as anchors for the progress the attackers make. They can get you really close to the objective but that’s not the same as driving a wedge into the defenders.

(spookify) #166

[QUOTE=Silvanoshi;512219]If you’re going to do a framerate comparison, it really makes a lot of sense to have your setup configured in the same manner. Comparing results from two different configurations when you’re trying to ascertain how the number of players affects performance isn’t really a great way to go about it. Don’t get me wrong, I’m not discounting the fact that additional players on a server can impact performance, but it’s much more useful for all involved to get info from configs that can don’t have any differentiators in them.
[/QUOTE]

In my post above and my general post spooky frame rate I did use the same CFG… That is testing 101 stuff…

Using “any” CFG I can run around in a server and max out my frames at 360ish whatever it was… I can unlock that and score 500 to 600 frame in some areas ALONE in a server running around…

The big problem is when people join the server you see a steep decrease on Frames based on the number of people in the server…

JUST ME = 600+ Frames
Me + 1 = 400 - 500 Frames
Me + 5 = 225 Frames
Me + Full Server = 50 to 175 Frames

I do not pay attention to my ping any more and that a whole different issue…

This is a flat out huge problem… The more people in the server the worse the Frames are… Its as simple as that…

(Running around in an empty server the game feels amazing!!! All bullets seem to hit and fire where you want them too and you can switch weapons with out a hiccup)

Also with Medic paddles = what Voshy said…

(Silvanoshi) #167

Pixel
That is indeed an area that could see some improvement for sure.

Voshy
The feeling of server responsiveness and the accuracy of 3rd person animations can most certainly be impacted by network latency. As I said in my previous post, I’m not stating that latency is the cause…but rather that latency can potentially amplify the symptoms of the problems.

Regarding reviving, the winky smiley implies that I wasn’t being entirely serious with that last statement. As I’d already mentioned, we’ll likely change how defibs work in future.

Spookify
I could be wrong, but the images that you attached are taken at different resolutions. Is that not the case?

(PixelTwitch) #168

even if spooks images where at different resolutions I can assure you I experience the exact same issue.

(spookify) #169

[QUOTE=Silvanoshi;512250]

Spookify
I could be wrong, but the images that you attached are taken at different resolutions. Is that not the case?[/QUOTE]

You are correct that those two screen shots were different resolutions… That is what I was testing at the time… I do have a screen shot with the same resolution to make it all match up but do to laziness and a mistake I garbed the wrong image from train.

Also, while testing resolution and alone in a server I did not see a drastic Frame change.

I will try to be 100% correct next time with my posts…

(spookify) #170

Pix could you grab a couple Frame Screen Shots like I did to back up what we are seeing?

Go in an empty server unlock the frames and see how high you can get… Run around and so on…
-Then hope in a full server with the same everything to compare.

Now that I am reinstalled I will try it again and sort my screen shots :smiley:

We can probably dig even further to see what an empty server vs a full server does to our machines (ie cpu, ram and GPU) usage…

I think that this is way more important then the bugs at the moment…

(Silvanoshi) #171

Sure, and as I’ve said before, there certainly are improvements to be made.
But if we’re going to be getting as granular as to discuss specific framerates as Spookify was doing, ensuring that the test environment is as consistent as possible is the way to go. I’m certainly not looking to place the performance impact solely on resolution, or indeed latency, there are other much larger factors that need to be addressed.

Would still be interested in seeing the results that you guys come up with though. Always interesting to see what happens on different hardware.

(spookify) #172

I will take a lesson from the movie Frozen and

//youtu.be/moSFlvxnbgk

[QUOTE=Silvanoshi;512258]
But if we’re going to be getting as granular as to discuss specific framerates as Spookify was doing, [/QUOTE]
FYI I am talking “hundreds” of frames…

(PixelTwitch) #173

Without a stable testing environment that makes things much harder for us to do.

What we do seem to know is that due to the fact the framerate can change greatly without anything additional being rendered… The slowdown is mainly due to CPU and NETWORK issues.

(spookify) #174

I will have to say after playing Medic again something is up with defib paddles… It could be the speed at which they come out or the animation of that speed but something feels off…

You are very correct the paddles are very accurate… Why? You literally have to be crouched right next to the body… No fly by’s or standing revives unless you are luck or lucky good…

I will try to go for center mass I guess but seems a little crazy when you have a noob or even me zapping your 3 or 4 times…
-The game feels so slow when you have to run in and crouch next to a dead body… I just dont like it…

I want to long jump in and zap the closest part of the body and bring up my main gun to defend myself…

Also bring out the gun and bringing out medic packs is crazy slow too

(trickykungfu) #175

[QUOTE=Silvanoshi;512258]Sure, and as I’ve said before, there certainly are improvements to be made.
But if we’re going to be getting as granular as to discuss specific framerates as Spookify was doing, ensuring that the test environment is as consistent as possible is the way to go. I’m certainly not looking to place the performance impact solely on resolution, or indeed latency, there are other much larger factors that need to be addressed.

Would still be interested in seeing the results that you guys come up with though. Always interesting to see what happens on different hardware.[/QUOTE]

you guys have to know why playing on a full server has soooooooo extrem impact in FPS. Cant believe there are many little things adding this up.

Just a little tip stop posting on Facebook the response in the comments is awful…

(Mustang) #176

Hint, it’s Nexon that control the Facebook account and they seem to have the impression that any advertising is good advertising.

(spookify) #177

He never said which facebook account… Sil posts and so does miss

(tokamak) #178

Having frustrated people on Facebook is acceptable. What’s less acceptable is that DB’s Facebook promotion doesn’t really correspond with all the other promotion. It’s weird seeing the game being pushed like that on there while it’s very much a mystery everywhere else.

(spookify) #179

Whats the good word SD?

I heard rumors that this next update is going to be pretty awesome…

One New (Old) Map?
Changes to Current Maps?

I am expecting something big!
Red-Dots?
Second pass crates?
Shotgun Tweaks?

Double Abilities this patch?

  • Less Clustering?

Match Making Back?

Execution?

Patch Date = What My Guess is Tuesday November 4th…

Can you give us a little hint? We really havent asked lately and this could be our Halloween sugar rush!

SD, you could even say something super cryptic like you always do… Like “There will be a patch…” HAHA! No crap, I would hope so…

We are going to see an old map re-appear soon but what else? :smiley:

News? Its been one month since this update hows life at SD?

(Silvanoshi) #180

These rumours you’ve heard, set them aside. Whomever spoke those words knew nothing of what they were saying :P…

It will indeed be a bit of a while before the next update hits, and there’ll be some nice stuff in it, but it’s not going to be an absolute behemoth. I’d say that, as is often the case, patience (without getting caught up in the hype), is key.

We’ll certainly be talking more about DB between now an then, so more info’ll be on the way for sure.