Title: Juliabrot raytracing and distance estimation Post by: fractalrebel on November 22, 2009, 05:51:08 AM The juliabrot was named by Mark Peterson, one of the original developers of the FractInt software. FractInt generated juliabrots have none of tenacles often seen with newer versions of software used for making juliabrots. The tenacles seem to be an artifact of the distance estimate algorithm. In fact, all distance estimate algorithms that I have tried generate tenacles. The book "Hypercomplex Iterations" also mentions that juliabrots are problematic for distance estimation. By stepping directly to the surface with no distance estimation ( a SLOW process ) a juliabrot without tenacles, like the old FractInt juliabrot is obtained. I have posted two images here, one using distance estimation and the other using direct stepping to the surface. If anyone has a distance estimation method that doesn't create the tenacles I would really like to see how its done.
Title: Re: Juliabrot raytracing and distance estimation Post by: David Makin on November 22, 2009, 01:32:26 PM Hi Ron, try using DE but with the threshold set to zero and "solid" to be either DE==zero or max iterations is hit and set maxiter appropriately. That should get rid of the tenacles. |