99 replies to this topic

[Phaesphoros]

I may not be up to date, but I thought 12v12 was only meant for the big maps (Alpine, Tourmaline, Canyon). It is fun atm to get massive brawls on the smaller maps, but that's just because it's new and exciting. IMHO 12v12 is not a good idea for the smaller maps.

Performance: Not improved at all. If anything, it's worse. I got about 23 fps average during a brawl, that's unplayable for a light 'Mech. Yeah I could join the heavy & assault meta...
Also, I don't understand why you don't tackle the HUD performance impact. I still get a steady 20 FPS boost when deactivating the HUD with right-shift + F11. I do not know if that's the case for all other players using low-end PCs, but I hear similar reports from other players.

If I deactivate HUD, the FPS is high enough for me to consider it playable: 40-60 FPS.
If I don't deactivate HUD, I cannot enjoy playing a light 'Mech any more in 12v12.

You can do slight to unnecessary performance improvements of 1-2 % for several patches or try to work out why the HUD consumes about half the FPS on low-end machines, which might involve building a new HUD system from scratch.

Btw my specs are still higher than the minimum requirements. I don't think anyone could call 23 fps playable.

Edit: yes I've set to lowest settings

---

---

---

It might be the case that there's a bug with the paperdoll:

I watched twice now a 'Mech with a cored CT (= no armor left) get damaged at the CT, and during the resulting flash of the paperdoll's CT, the armor flashed, too. That is, during this 3 second flash, it looks like the CT has still some armor left.

Edited by Phaesphoros, 18 July 2013 - 10:30 AM.

[Vernon Jettlund]

Vernon Jettlund, on 18 July 2013 - 10:11 AM, said:

No luck after reboot, same thing.
W7 Pro, 64-bit, SP1. AMD FX 8120 (8-core) proc, 16 gigs of DDR3 RAM, SSD, DX11 if it matters. Graphics is AMD Radeon HD 6700 series, I forget exactly which model. Let me know if you want any other info.



Ok, trying a re-install.

Go-go Gadget Comcast Burst!

No luck, same thing. Sending DXDiag export to support

[xengk]

seymourbalzac, on 18 July 2013 - 10:20 AM, said:

So my camera decided to glitch out hardcore in my last game, I was just moving my mouse to look around to find an enemy mech and the next thing I know I'm looking out through the back of my mech and I can see the body of my pilot, it then became extremely hard to move the camera where I wanted it to go and ended up staring at my side window for the rest of the game. I was not holding the ctrl button and pressing it/holding it didn't help anything.

first glimpse of 3rd person

[Matthew Craig]

It's not that straight forward Phaesphoros we share the frustration, we are not able to easily take Scaleform updates as the Scaleform implementation that is in CryEngine is linked to the CryEngine Scaleform interface. CryEngine 3.4 was supposed to come with an updated interface layer but it didn't happen.

Later versions of Scaleform add batching which significantly helps performance but we are limited to what we can do for now. There were optimizations made to this layer recently but as stated there is only so much we can do in the short term. UI 2.0 is being done in a way that will allow us to more readily take future Scaleform updates.

We will keep looking at the HUD for more gains that can be made in the interim but it is not as simple as just making some tweaks to the HUD.

[Corbon Zackery]

Everything seem to be the same huge issue with missiles I think that's the huge reason testing was delayed 30 min.

Some issue with frames per second. I'm running a i5 gaming computer with a Flattron 60hrtz monitor, and a huge graphics card. I'm only getting 9 to 15 frames per second on high graphics, and 1400 by 900 resolution settings.

So I don't think your getting accurate public data.

[MrMasakari]

Mostly fine for me, but I'm having issues with Srm hitboxes. Was 1v1ing a raven in my raven, and my lasers were hitting fine,but my srms hit but seemed to do little to no damage from an almost full salvo (Srm6x2 aswell)? I thought this was odd especially since this buff, and I don't have this problem on normal server so. Thought I would mention it anyway.

[Faithsfall]

I'm having an issue with my Heavy Metal the 2nd missile pod is firing as if it was a narc postion.

bloody annoying to fire 6 missiles then a train of 1 missile x6

[Matthew Craig]

Regarding the camera issue that is a bug that only exists on public test and is in the known issues. Please disregard it and remember what you see on public test is just a test build nothing is guaranteed to make it's way to production.

[Bloody Moon]

I think the old, small maps desperately need an update. Right now there are mechs everywhere, far too many for enjoyable games.

[Mawai]

Some comments on the test client:

1) The download speed is very low. I've managed 650Mb out of 2485.99Mb is about 4 hours. Normal download rates for me are much higher so I am guessing there is a bottleneck at your end as folks try to download the test client.

2) The installer has a couple of text errors that need to be repaired. When the installer comes up the splash page showing the "Typical" installation conditions shows the install directory as C:\Games\Pirnaha Games\MechWarrior Online ... which is the live client and not the test server.

3) When I clicked on custom it showed the correct MWO test directory but then the installer crashed when I hit install from the custom selection. This may or may not have been related to the fact that the live client was active when I started the install process. I closed the live client as requested.

4) The test client insisted on closing the live client before installing. Considering that these are two separate programs with different installation locations, it isn't clear why the live client needs to be closed to install the test client unless the installer uses the same executable name and can't tell an instance of the live client from the test client.

[Ralgas]

running very high graphics saw an fps jump from ~40-45 to solid 60 over test 1 and a ~40 ping drop (Australia location). hit detection better but still off, especially on the larger maps (alpine).

[Ens]

turns out that 12v12 plays kinda different. it´s more tactical and depends more on lance teamwork

performance is great with 60 fps straight


but the small maps feel even smaller with 24 mechs on it
there´s metal everywhere

[Matthew Craig]

Mawai regarding the download speed, we distribute our patches via a content delivery network, if you are the first user pulling down the patch in your area it is slower than usual with the live patch it distributes over a much longer period for the test installer there is a much shorter window for this distribution. We'll look and see if there is more we can do to help seed the download earlier to ensure the download speeds aren't impacted. Thanks for the feedback.

[wickwire]

Matthew Craig, on 18 July 2013 - 10:38 AM, said:

Regarding the camera issue that is a bug that only exists on public test and is in the known issues. Please disregard it and remember what you see on public test is just a test build nothing is guaranteed to make it's way to production.

Speaking of which, all Atlas (at least unpainted/trial ones) seem to have a "royal" paint scheme in the test version



(i wish this WASN'T a bug )

[Protomartyr]

FPS was lower than last week on high graphics settings.
FPS was 10 FPS higher than last week on medium and low graphics settings.
Game was unplayable at 13 fps which is what I get with maxed out graphics options.
Game was perfectly fine at medium and low settings.

Hit detection was terrible. It was like playing table top, everything hit randomly. Even on stationary targets.

[Evax]

The SRM tube changes are pretty bogus..
I Hope those are bugs..not PGIs crappy way of nurfing SRM use.

[Punk Oblivion]

Everything seemed to work fine for me. From what I could tell I WAS getting a decent hit detection % on small mechs, maybe not 100%, but 80-90 for sure.

12V12 is so unpredictable, I love it. It really makes you tighten down your tactics, no more running across open spaces in front of the other team, even at 800 meters...

[pushptz]

can`t on postAA at low, med and high. Changes are minimal for core2Duo system, avr fps 20-25

[Fetladral]

most of the matches i was in this morning the other team usually had about double the number of assault mechs. only maps where it was fun was alpine and tourmaline because the rest of the maps are to small for that many mechs.

[Phaesphoros]

Matthew Craig, on 18 July 2013 - 10:34 AM, said:

It's not that straight forward Phaesphoros we share the frustration, we are not able to easily take Scaleform updates as the Scaleform implementation that is in CryEngine is linked to the CryEngine Scaleform interface. CryEngine 3.4 was supposed to come with an updated interface layer but it didn't happen.

Later versions of Scaleform add batching which significantly helps performance but we are limited to what we can do for now. There were optimizations made to this layer recently but as stated there is only so much we can do in the short term. UI 2.0 is being done in a way that will allow us to more readily take future Scaleform updates.

We will keep looking at the HUD for more gains that can be made in the interim but it is not as simple as just making some tweaks to the HUD.

Thank you for the explanation. I know that it's not simple to tweak/update, otherwise you'd obviously have done that already. What I was talking about is to ditch Scaleform completely and try to do the HUD another way (yes, that's much more work, but you'd get rid of those issues and could do that instead of optimizations). I had to look into the CryEngine SDK, but chances are high there's a way to impose an externally composed overlay to the picture created by CryEngine.