69 replies to this topic

[One of Little Harmony]

FPS is marginally worse. I play at the highest settings on this game because I get more fps than lower. I have no idea why.

Live
27-35fps full window

Test
19-25fps full window

i7-4770 3.5ghz oc 4.5ghz
Nividia 650GTX (shhh, I used to have a very small monitor)
Windows 7 Pro 64-bit 8 gig memory at 1920x1080

Also, I feel like the lasers looked much better on the test version, other than that I didn't notice too much of a difference.

I'd like to suggest capture going to 1000 resources instead of 750.

BlueSanta, on 18 July 2013 - 05:22 PM, said:

So...about 12 mans...

THIS. As soon as you loose a couple mechs you're so much more likely to lose them all.

[muskrat]

Had 1 return to MechLab game 4. (never had this on live) FPS was considerably less 15-18 average (had 1 match with a 20) normal live FPS is 30. Ping was much higher in afternoon match (122 low 155 high avg 143) consistent ping live is 91-98. I played 8 matches all on river city or forest colony (way to small for 12v12) I run a fairly low end system, AMD 4600 dual core, 4gig Ram, Diamond Raedon 5470 1gig Ram. I would probably play less if this was a live server.

Edited by muskrat, 18 July 2013 - 05:38 PM.

[Roheryn]

BlueSanta, on 18 July 2013 - 05:22 PM, said:

So...about 12 mans...

Yeah you dropped ~1000 tons against our ~800tons and had 6 headshot kills...so about 12 mans....Honestly you guys were beasts. Just shows how badly we need Weight limits PRIOR to 12 man.

Also it feels good that we at least managed to kil la few of you looks like everyone else did even worse.

[Z0MBIE Y0SHI]

I had no crashes, hit detection was night and day. So much better for whatever reason. Latency was about the same.

We did have one odd thing happen, where all 12 of us crashed at launch, and our mechs were in game for a pretty long time. Client resets didn't bring them back. Other then that, I enjoyed it much more then the other server.

[Offrezwolf]

A few things that were causes of concern:
Maps are too small, there was a lot of friendly fire happening.
My fully armored undamaged Stalker was getting one shot by a PPC/ER PPC blowing out the entire L/R torso far too often (No ammo to explode in these areas).
Damage on weapons vs armor value appear to be consistently off.

[Rippthrough]

Roheryn, on 18 July 2013 - 05:36 PM, said:

Yeah you dropped ~1000 tons against our ~800tons and had 6 headshot kills...so about 12 mans....Honestly you guys were beasts. Just shows how badly we need Weight limits PRIOR to 12 man.

Also it feels good that we at least managed to kil la few of you looks like everyone else did even worse.

You think that's bad, look at that atlas steamtrain

Although, I have to say, the 12 mans are resulting in some serious steamrolling - even on those games if you check the figures there was some (deliberate) TKing going on once the roll was almost finished. 2 in that one with Rock having a squabble

I'm really not sure what can be done about that, theoretically it should be better as one mech is less in percentage terms, but in reality, 24 mechs manouvering means the disadvantaged team just cannot get the hell out of the way, press the advantage and mech parts just start exploding everywhere.
Maybe after the strategies have settled a little it might get better.

Edit: I think it was actually your game I was having the most FPS issues with cents on the screen. Were all of you firing SRMs? Seemed like it at the time

Edited by Rippthrough, 18 July 2013 - 05:56 PM.

[impaledface]

My experience is about the same as the live version, only had one time my frame rate went from 60 to 15 for a little bit on river city, but the frame rate went back up after so its no big deal..

[Patient X]

Noticed that the frame rate take more of a dip on Forest Colony then the other maps. On a GTX 580 running at 1920x1200 medium-high settings (some sliders are at high) from 60 to 40 FPS looking around, maybe lower when there is action. Hit detection seems a bit better on spiders and jenners in the test version.

[KingGnome]

Crashed 3 times in a row at the beginning eventually putting my graphics from high to low. Didn't know about the F9 for the fps. Moderate freezing a on a few games. One resulted in like a a 20second freeze when it ended iI was on the destroyed screen. New patch have been having issues crashing about 25% of the time. The patch before that worked just fine. Nvidia drivers were updated last night with the new release.

E8400 duel core 3.6GHz Win7 Ultimate 4GB 1333 DDR3 nvidia 640 1GB

3DMark Score 13417 3DMarks SM2.0 Score 6 K HDR/SM3.0 Score 6 K CPU Score 3280 Game Score N/A GT1 - Return To Proxycon 51.290 FPS GT2 - Firefly Forest 47.107 FPS CPU1 - Red Valley 1.029 FPS CPU2 - Red Valley 1.672 FPS HDR1 - Canyon Flight 58.073 FPS HDR2 - Deep Freeze 65.683 FPS

[Roheryn]

Rippthrough, on 18 July 2013 - 05:50 PM, said:

You think that's bad, look at that atlas steamtrain

Although, I have to say, the 12 mans are resulting in some serious steamrolling - even on those games if you check the figures there was some (deliberate) TKing going on once the roll was almost finished. 2 in that one with Rock having a squabble

I'm really not sure what can be done about that, theoretically it should be better as one mech is less in percentage terms, but in reality, 24 mechs manouvering means the disadvantaged team just cannot get the hell out of the way, press the advantage and mech parts just start exploding everywhere.
Maybe after the strategies have settled a little it might get better.

Edit: I think it was actually your game I was having the most FPS issues with cents on the screen. Were all of you firing SRMs? Seemed like it at the time

I wasn't firing anything I was face down almost immediately ...but yeah we had 3 cent A-s and some DDC's with SRM's

[elsie]

1st match of the session, it crashed back to the launcher. But after that it was pretty much the same as the live server except for the crowds and being more hectic.


elsie

[Caseck]

Had one crash, one false drop where my mech probably stood on the battlefield like a tool.

FPS was worse from 15fps to 9FPS. Damage textures were showing on my graphics LOW settings which I hold responsible for much of the FPS dip. I don't look at it, I look at the damage charts anyway.

I need a better GPU.

[elsie]

BlueSanta, on 18 July 2013 - 05:27 PM, said:

You could turn and look out the open back of the cockpit (since they have yet to close them).

I could on my Highlander, but not on my Awesome. It was interesting being able to see behind you.elsie

[Derick Cruisaire]

I wish to add a caveat to my hit detection vote. It is very likely that the poor hit detection I perceived had more to do with the 6-8 FPS I was getting than the actual hit detection.

I also would like to mention that I ran LRMs in four matches this evening. In each match I noticed that when I fired my missiles I would see some of them fly laterally, from right to left, directly in front of my cockpit. The first time it happened I thought somebody had snuck up on me, and was firing LRMs at me from inside of minimum range. After turning to my right and finding no one their, I realized that the event always coincided with me firing my missiles.

Sable, on 18 July 2013 - 03:49 PM, said:

Reviewing the feedback from the midday test i saw people talking about the SRM pods on a highlander where one was shooting in a cluster like normal and one was shooting in a stream of 1 missile at a time volleys. I don't know if this was intended but i actually really like this idea!! Maybe not for SRMs though but i'd love to see this tried out for LRMs. I tried filling all the missiles slots with LRMs but they fired in clusters like normal. I'd be really interested in seeing if PGI would just throw this option on test and see how it performed and how people felt about it.

There is a Trebuchet variant that comes with a Narc launcher. If you put an LRM launcher in it's place the LRM's will come out one at a time.

Edited by Derick Cruisaire, 18 July 2013 - 07:25 PM.

[Deathlike]

It seemed laggier for some reason for me.

IMO there should be a question for ping stability in these tests (like, does the ping vary much or frequently change from a lot to a little or something like that). That seriously is a problem ping-wise for me as I go above the 100ms at a time, and not in a good way.

[3ffigy]

SilvaDraconis, on 18 July 2013 - 05:02 PM, said:

Single tube SRMs on my ravin (LA RT) was very disruptive. didn't notice that happening on the atlas.

I had a similar issue on my Heavy Metal. One SRM-6 launcher fired as a salvo, the other fired single missiles. The model has plenty of tubes in the torso, so it shouldn't be getting this weird salvo treatment as far as I know.

Played several matches in the 1-4 queue that went fine. I did experience one match instance crash/freeze mid-way through a match on Alpine. The other members of my group were able to confirm that it wasn't isolated to my client and that they also froze, so we exited that match and requeued in different mechs.

Edited by 3ffigy, 18 July 2013 - 08:18 PM.

[Narcissistic Martyr]

fx8320@4.2GHz
HD7950 @ 1100/1575
8 GB 1660MHz RAM.

FPS: I'm losing an average of 10-15 fps from the live version and frankly I can't figure out why since I still have power to spare.

CPU: I'm seeing 6 cores on the fx8320 being used during play at around 60% utilization per core with a max temp of 50 C. This is higher utilization than in the live version which only uses 4 cores for some reason.

GPU: I'm seeing around 50% utilization of my GPU with spikes to around 80% utilization with a maximum temperature of 60C at 50% fan speed. As my GPU isn't being maxed out I returned it to stock clocks with no detrimental effect.

Issues: There is delay between when I pull the trigger and when my Gauss, AC20, and SRMs fire. I had one C2D and one disconnection halfway through the match.

Edited by Narcissistic Martyr, 18 July 2013 - 11:19 PM.

[Rippthrough]

Roheryn, on 18 July 2013 - 06:17 PM, said:

I wasn't firing anything I was face down almost immediately

That'll teach you not to run out at me from cave, I have a nervous disposition, I tend to alpha things when disturbed

The reduced missile trails on LRM's was a big performance improvement for me, plus, you could actually see instead of the trails of 90 million LRM's blocking the sky - which was sometimes worse than any FPS drop.
Something else I forgot to mention is the bug with groups getting stuck without being able to reinvite people when they crash/drop seems to be back/more prevailent on the test server.

Edited by Rippthrough, 19 July 2013 - 04:35 AM.

[Zyllos]

The difference between the live client and this test client in terms of FPS performance is significant (+/- 10 FPS).

With the hardware I have, I would not be expecting myself to be running that low of an FPS. I even dropped the resolution one step (from 1440x900 to 1280x768) and turned Post Processing to low and Shadows to medium to gain stable FPS.

I am hoping with the new DX11 coming in the future, allowing for parallel processing and Direct Compute, this will increase my performance in the 12v12 environment.

Hardware:
i7-3820
XFX Radeon 7970
16gigs of DDR3
SSD (but the 12v12 test client is running on a HDD to save writes to my SSD)

[Ardlen]

I could not get UAVs to launch in this test.