Sorry if this is a double post. I shared this from mobile, and then had someone tell me it didn’t show up in the FastLED community.
Originally shared by Jason Coon
I wired up a stereo MSGEQ7 board and added a couple of music reactive patterns.
I switched to a thinner section of 2.5" PVC pipe, which makes the spiral taller and provides more distance between the LEDs and the frosted glass for better diffusion. I also added a rolled sheet of vellum paper inside the glass for even more diffusion.
I saw the other post, not sure why it wasn’t showing up for some people (though, the new G+ UI makes it way to easy to accidentally end up in a view for just one of the subsections).
Also - as an aside, G+ sometimes likes to randomly flag posts as spam, which immediately hides them from everyone but me until I review/approve them. I try to check a couple times a day, but if you do a post to the group and it appears to not be showing up, that might be what’s going on. (though, that didn’t happen in this case)
Top! Especially the rhythmic noise at the beginning. 2 layers right? One moved along x and one along y. Nice.
edit: Checked the code. Aha, one layer - mapped with index and brightness swapped.
I’m thinking about using two noise layers instead, for more variation. After looking at it, it’s pretty easy to tell it’s the same layer, just rotated 90 degrees.
One short remark: I saw that @Jason_Coon still uses the inoise8 function. Try inoise16 - it looks way better at edges and with “striped” palettes. Needs just little adjustments: x, y and z are 32bit then, the deltas (you called them noisex and so on) have to go up, the scaling too and the results need to be bitshiftet by 8 to fit into the 8bit for the palette index space. Result: less artifacts at edges, less flicker with thin color lines. Computing time price: on a Teensy nearly nothing.