TweakScale

TweakScale

1M Downloads

Engine Effects are being scaled the wrong way!

Lisias opened this issue · 14 comments

commented

As the title says!

Well… Search and Destroy! (the bug)

Reporter: kocmohabt , here.

Evidences:

NERV

screenshot37

Vector

screenshot38

Reliant

screenshot39

commented

Moving this one to the 2.4.6.x series, following the ROAD MAP.

commented

Decided to tackle down this one sooner.

commented

This issue is blocked by #276 .

So it's being moved to the same milestone.

commented

Just for the records: #276 is closed, so this issue is unblocked.

commented

Also, for the sake of curiosity (as it's more than evident that some scaling is happening nowadays):

pellinor0#11

commented

More interesting stuff here

The length of the plume is not being scaled at all on this specific case.
2019-03-02_02-48-18 png

commented

This can be another symptom of the problem!

commented

Is there any plan/timeframe that this bug will be worked on?

commented

Made some more tests on 1.7.3, checking also the air breathing engines too. Behaviour consistent, but it's not exactly the same as the rocket ones.

The tests were made

  • Panther: screenshot5
  • Whiplash: screenshot7
  • Spark: screenshot8

It's not something TweakScale is doing wrong. It's simpler, it's something TweakScale is not doing. My hunch is that by understanding how exhaust works on KSP, a one or two liner on the Exponents will fix this. :)

The behaviour for the air breathing engines are different from the ones for rocket engines, so it should be something on the engine module.

commented

That was unexpected. I gave this a run on KSP 1.3.1 and TweakScale 2.3.7. Same results!

Technically, this is not a bug. It's a latent new feature waiting to be developed for years! :)

  • Whiplash:
    screenshot7
  • Spark: screenshot8
commented

One fellow Kerbonaut contributed with a new test bed with a lot of engines. Thanks, Krazy1!

For KSP 1.10: TS engine test.craft.zip

commented

The Engines exhaust is implemented on KSP using Particles (and it's not exactly the best implementation I ever saw on the matter), and the thing is generated procedurally by the Engine Module.

So, I don't thing I will be able to easily tackle down this issue without some hacking on the engine modules - what potentially can create more more problems than the ones it would solve.

However, there are people tackling this differently, using meshes. This is something that TweakScale does good, scaling meshes (it could not does it badly, it's just a scale number, the scaling is handled by the engine :P ).

Check how things goes on https://forum.kerbalspaceprogram.com/index.php?/topic/196309-wip-waterfall-a-framework-for-continuous-mesh-driven-engine-effects-sept-3/ , this is promising.

commented

For the record, I think the force vector of some engines may also being inverted (and the exhaust definitively are scaling inverted).

From AccidentalDissasembly post:

I do have some small feedback about the beta - plumes are not scaled correctly (down or up), and sometimes it seems like plumes are scaled in the opposite direction of the part - for instance, in this screenshot, I think the white fluffly plume is much larger than when the part is normal scale. It's from the Snazzy pointed side tank (separation motors), which is scaled down to 0.3m in the screenshot, normally 1.25m.

On the other hand, the specific parts i tried (mastodon, kodiak thing, the snazzy tank) did not move around when reverting... so there's that

eG7101y

HOWEVER... This is not affecting every engine, jet engines appears to look good - besides perhaps the length of the exhaust should also be extended (or shrink)?

screenshot118
screenshot120

commented

In time, SmokeScreen works fine with TweakScale, as it have embedded support for it - see this line on the SmokeScreen repository (on a side note: what the hell, line 666??? =D =D )