Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[WxTech]

After many years of never discerning any impact of the .eff file parameter Rnd on particle-based effects, today I finally stumbled upon this one aspect of its operation.

First, I've long been aware that for trail-based effects, such as contrails, tracer smokes, ship wakes and wingtip stall vortices, Rnd will apply a random variation in the intensity, or opacity with time. This adds a nice irregularity which adds to the organic, natural appearance.

But until now I could see no sign of operation on particle-based effects, such as smokes. This led me to think it was disabled for this type of effect.

Well, earlier today I thought to get back to seeing if anything could come of it. In the past I tried values like 0.3 or 0.5. On a whim I edited my static plane long-lasting smoke to have large value for Rnd of 35. Below is the result, after shooting up a bunch of static planes and one tank. The tank smoke is dead center, and all the other smoke columns are from destroyed static planes. Previously, because I have applied a high degree of uniformity of smoke behaviour by design, the static plane smoke very closely resembled the tank smoke in form. But now the difference is very obvious. The larger value of Rnd causes a notably greater variance to be applied to at least one parameter, as evidenced by the greater dispersion in particles in the vertical. Instead of a more gentle widening of the column downwind, now the particles fan out over a wider range of angle. In particular, previously the particles never touched the ground, but now they can and do.

I tried other values, decreasing in magnitude each time, such as 25, then 10, then 2, then 0.99, then 0.67. It was only once the value got below 1.0 did the dispersion decrease. And so it seems that permitted values are between 0.0 and 1.0; if > 1.0 it is treated as 1.0.

In this case of ground smokes, where the particle emission vector is oriented vertically, this randomness parameter operates ONLY in the vertical; the dispersion laterally remains as before. That is, when looking at a smoke column from either the upwind or downwind direction, the column width is no different. Instead, the particles are spread out in a wider fan in the vertical plane as seen from the side. Now, I haven't divined which one or more parameters is operated upon. It could be VertAccel, GasResist or EmitVelocity. Perhaps with further experimentation I can determine which.

I have now used this to good effect on all ground smokes, applying a further degree of irregularity and hence organic naturalness, by use of Rnd 0.66 (previously I had largely set this to 0.0). That results in less dispersion than seen here, it being roughly 2/3 of the way from the older tank smoke to the wide fan-shaped pattern in the image.