Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
The Laggy Drawing Systems
#1
Well, a list of laggy systems which modder should not really use much if any at all unless don't mind lag in online games as skirmish misleads thinking performance is good when it isn't!

1) Translucent

Very bad, this routine does a fading like translucent effect for anims and it works by some kinda hard coded delay between the transitions so does not sync with the lenght of the animation, if its really short, it will only manage show the anim with 25% which is pointless idea, could have just used Translucency= to set solid amount.

Online wise, this should not be used at all as even 2 animation instances bring about noticeable pauses.

2) Translucency (25,50,75)

If you must insist using Translucency, this is it, I recommend value of 25 for most minimal lag and only sparingly use value of 50 if really needed and another note about UseNormalLight=yes, you should use animations 100% light if using translucency or else visuals will be less nice in my experience and default is not to use 100% and go according to ambient light.

This can also be used in particle systems that allow SHPs to control the amount and it is suggested to use as default seems to use the slower routine if not specified and most interestingly Translucency=0 can be used to disable translucency of the particle system animation in my tests.

3) Particle systems

They are all prone to causing lag unfortunately and in cases of spark or otherwise purely engine drawn particles, lag can only be reduced by cutting the amounts of particles and often simplest way is define ParticleCap= into the system to get some control as i suspect there may not be hard coded limit in place by default.

As for particles that allow SHPs (Smoke/Gas etc) the lag comes from the default of translucency being applied and as i said above it can be disabled or put to lower level and MaxEC (Duration) should ideally be low value as longer the particle sticks around, it lags by being present and disallows same unit make another particle until existing has ended.

Trying use the Particle light is also prone to lag like its CombatLight counterpart.

4) Debris

Shamefully, it does lag fairly bad and something like 6 debris is optimal at once (total) and past that you can expect pretty annoying lag and curiously enough, the rules specified voxel debris do not seem to lag as much as the pure shp debris being called from art.

I do suspect the system is prone to stack-up lag due to loopcount=-1 which is mainly used to keep it animating but also means infinite loop, incase engine does not properly end it it would mean it is adding more and more so called active debris on top of existing while they do seem to disappear but code may not end properly, was kinda observed years ago in original game with online action as more bases got destroyed, more debris flew and slower and slower game got and eventually it crawled and did not recover from this lag unlike translucency.

5) Combat Lights

Yes, they are pretty cool but as usual, prone to lag and online wise, should not be used in rapid weapons if intend using them.

Using values past 100 cause definite momentary lag and it seems to accept about any value for CombatLightSize= while beyond some point it won't get any larger visually.

Further testing of even low values like 40 still cause brief lag spikes online :/

6) Trailering Anims

Best kept often short and using any Translucency on these moving anims is even more laggy than stationery animation and should not be done at all.

7) Radiation Fields

Potentially laggy if occupies very large areas, like map designed around it to emit lots with lightposts.

8 ) Cloak Generator

Ideally keep this low radius as pretty unoptimized routine that can add noticeable lag easily to online games.

9 ) Excess Laser shrapnels & LaserDuration

If you go really crazy with shrapnelling lasers like http://www.moddb.com/mods/robot-storm/im...b-of-beams then pretty bad lag, else works smooth in basic laser draw uses.
Another bad idea is use shrapnelling laser in an very rapidly firing original weapon and if warfare is really rapid, i advise using as short as possible laserduration as even maintaining dozens of beams at once can strain the game like long lasting particles.

10) Excess Cluster= values use in projectiles

Causes noticeable momentary lag especially in rapid burst weapons like saturation machine guns so best used sparingly.
Reply
#2
*looks conspicuously inconspicuous*

Oh what's this? A test that can be used to measure only anim slowdown over a long period of time?

TEST
[Image: alexstand1.gif]
Reply
#3
@Vinifera7, I believe you conducted this test in skirmish which is not any value determining lagging unless tried online where syncing needs to be or game keeps stalling so all maintain sync.

I find best debris test for me was put debris spawn into every warhead of the game and go play...wow the lag Smile
Reply
#4
Sorry for the huge bump. Just want to tell something.
I removed every tag in my ini files that Apollo wrote. It took some times becuse my ini files are kinda big.
But the results online was amazing Smile

The game look a bit odd without debris, but its worth it!
Reply
#5
Its probably good idea to try consider a gamemode with full visuals or do online mode with cuts as online with most effects is pretty horrible.
Reply




Users browsing this thread: 1 Guest(s)