END OF AN ERA, FRACTALFORUMS.COM IS CONTINUED ON FRACTALFORUMS.ORG

I've been rendering saving to jpg, which I would have thought at max quality would be pretty clean of artifacts, just baking in what's there. I noticed pretty severe banding in the background gradient, when I attached to this post, I cranked the jpg compression up, seeing the banding worsen, so kind of answering my own question here, but can someone say if the bit depth of the gradient used in a scene with fog as seen here is 8-bit? If saving to PNG is the fix, I can do that. Thanks.

That makes good sense, thanks for explaining the issue. Though my two computer runs 64-bit, does M3d not make use of that? As an alternative to bumping up the contrast in the background gradient, might this be my motivation to start saving out z-buffer and learn to composite? I'm seeing folks drop scene elements, e.g. sky and water into M3d scenes, helps sell the scale and worldliness in the imagery, unsure how one matches angle and lighting. You obviously can't fill my knowledge gap here, but do you know of tutorials to take basics to the next level.

I have another question, while we're at it. In response to your post elsewhere on morphing two formulas, I'm very interested in that. I must have key framed two parameters from one formula, thinking I was experimenting with two formulas, because what morphed was shockingly beautiful (I went back to add 750 between frames to expand the nutso fractilisciousness). The idea of stepping out of the way to tease deeply hidden secrets from the nature of fractals appeals to me. You outlined the means to allow the interpolation, but I'm unclear how to implement your chunk of code. If I have formula 1 at key frame 1 and f2 at k2, where do I how do I make use of your code providing for the interpolation?

Vielen Dank!

Thanks. In order to dither the gradient, I first have to select it, would the z-buffer provide that selection? For animation, that wouldn't happen in Photoshop, well possibly as a batch process, but wondering how that might look in After Effects compositing ap.

z-buffer is immediately suited for depth masks rather than for isolating the object from the background but you should be able to easily generate a useable mask from it for the latter by tweaking it with the basic color correction tools in photoshop, levels, curves, brightness/contrast or even thresholding.

You could also directly generate a binary mask from M3D by using adequate color settings. I have attached a color setting file to that effect (might or might not work for you but it'll give you an idea anyway)

Sometimes I am able to satisfactorily remove banding using the free diffuse photoshop plugin by Richard Rosenman, it'll need proper masking/selection too and several passes might be needed, a too high radius will generate visible noise. I prefer using this over adding noise.

http://www.richardrosenman.com/software/downloads/

Btw I am also attaching a preset rendering a simple 1080p greyscale background gradient which clearly shows banding, it sure would be nice if there was a dithering option in M3D.

using the chroma keying is just fine, standard rules as for real videos apply the same, i used it several times!

attached you have an animation file, load it from the keyframe window in mb3d
it is a mandelbulb8 morph to a menger3

the only steps i did there:

1. go to formulas tab
2. go to formlua index tab 2
3. assign a formula there (i used menger3 )
4. choose in the top "interpolate" as hybrid method, then a "weight" value appears in the formula index tab
5. setting weight of formula 2 to the value of 0 (just because we are in formula tab2)
6. setting weight of formula 1 to the value of 1 (this should be default
7. open animation window
8. press add keyframe from current
9. setting weight of formula 1 to the value of 0 (now reverse, formula index tab 0 should be still open)
10. setting weight of formula 2 to the value of 1 (now reverse interpolation values, making a "crossfade")
11. in the animation window press "add keyframe"
12. now check settings of animation window, make sure "linear interpolation" is selected, if not start/end values might overshoot
13. navigate with the keyframe stepper to frame 0 and enter the desired amount of inbetween frames(the example as 1000)
14. press "Preview" in the keyframe window

So, bit-depth of the processor unrelated to bit-depth of what's processed. Thanks.

Thanks, Kleinhuis, for putting the dots that close, inverting the relationship between weights within the interpolation transform, nice. I'm now curious to understand what's behind my accidentally successful approach whereby I simply key framed one set of parameters formula 1, pasted in parameters formula 2 into key frame 2, resulting in bizarre behavior. I tried this on numerous other formulas with the result predicted by Jesse, sudden change over single frame, others blank. That is, with one exception, so 2 out of 15 or so attempts resulted in nothing. So, two out of about 15 (including that first blind luck attempt)... wow. The transition wasn't smooth, like throwing a monkey wrench in, Frankenstein's M3d nervous breakdown, outgoing formula appeared obliterated, randomized colors every frame, structure more gradual, but convulsing, shattering, disintegrating eventually into dust, then blank frames, then incoming formula pops into being. I'd be curious what you mathematicians make of such odd behavior that Jesse says shouldn't be, most of the time isn't, but sometimes works.

Right, I understand about the breaks, I was asking about these odd cases whereby I don't get the break, at least not through a long sequence of frames before the break. For instance, I set 750 keyframes between formula 1 and formula 2, the first 500 or so in these exception to the rule case will return these beautifully junked up transitions before the collapse into noise. Perhaps, I should provide an example, will work at that later today. I hope my question is clear.