What's the differents beetween

[jrdrennes]

in "coverage adjust" in "autocumulus layer" and "coverage adjust" in "density fractal". Sorry for my english, but i'm french...

Thanks.

[Hetzen]

Not a lot to my understanding, they perform the same function, except you may not always want to use a Density Fractal before the cloud node.

[jrdrennes]

Thanks 

[cyphyr]

There is a subtle difference but I've tried three times now to explain it and deleted each post because I couldn't get it.  
You best solution is to try it out for your self.
Do a series of small renders (320x 240) and change only the coverage adjust in the cloud layer by 0.1 at a time.
Now do the same again leaving the cloud layer untouched but changing the coverage adjust in the density fractal also by 0.1.
You should see within an hour or so how the two differ.
Hetzen is of course right, you don't have to use a density fractal (cloud fractal shader), you can use normal power fractals, image maps, function networks or combinations.
Good luck and if you can find a nice concise way of describing it I'd like to see it
Cheers
Richard

[Henry Blewer]

It is difficult to describe. It's very subtle. The density fractals coverage seems to change the overall amount of clouds. The Cloud Layers coverage reduces the cloud size horizontally. I think... I began using power fractals to handle the coverage; input into the blend by in the density window.

[RArcher]

It can have drastic differences depending on how you have your clouds set up.  The best explanation I've seen is here:  http://forums.planetside.co.uk/index.php?topic=1698.msg16676#msg16676

It makes more sense when you start having multiple density shaders in your cloud forms.  Then you can either adjust the coverage of all of them at once in the cloud node, or individually in each density shader.

[Hetzen]

Just to add to the good advice so far. A cloud is defined, again to my understanding, with a black and white input (the cloud density fractal). What the density function does, is shift the bias towards black or white. Ie, you give it a negative number, it makes more black. Positive, more white. It's a bit like using levels in Photoshop.

Now how the cloud node works with -1 inputs is another story. But I guess the above is a good starting point.

[Hetzen]

To explain the -1 inputs. Some fractal noise works between -1 to +1. Like a Sine wave dips below and above the zero y axis. Perlin noise.

Ridged and billows, is the reverse of what happens below zero y. So if you look at a Sine wave fall bellow y=0, billows will make that negative y curve positive, ie there will be a reflection of what happens bellow y's zero point to make it positive. Like a pair of wholesome rounded boobs.

Ridged is the opposite. Instead of rounded boobs, you get what you would take that boob shape away, and create pointy tits.

No bad thing.

You can play with these breast shapes even more with un-clamping, but then you'll be getting into puffies and missile shapes.....

[dandelO]

Now how the cloud node works with -1 inputs is another story.

Aye, and with '+2' for complete coverage, instead of a more straightforward '1'.
-2 to +2 has always seemed a strange range to me, considering nearly every other TG parameter.

[Hetzen]

You can go well beyong +2 and -2 for coverage. I often do with unclamped noise shapes to get something on screen.

[dandelO]

Yes, I just mean as a limit to the defaults, it's a strange range. From -1 to 1 would make more sense to me, then I could be the judge of how far past that default limit it gets pushed.
I assume there's a perfectly good reason for the odd range, probably to make it easier to adjust offsets/coverage of fractals and cloud layers, because increments of as little as .025 can create quite big changes.
Just seems a weird range to go between, to me. Even if I rarely stick to the default limits.