16 Bit Patterns

[Jeff_Birt]

As I said before, the benefit comes about when manipulating patterns on the board. An 8bit pattern can produce jaggy results, on both the screen and carving, where the 16bit won't. You want to manipulate the data with the highest resolution possible to avoid rounding errors that result in jaggy results.

This is not speculation on my part, I used to do the programming for the Pattern Editor and am very aware of the benefits/tradeoffs of both the 8bit and 16bit patterns.

[Alan Malmstrom]

Isn't it 256 including white and black?

Yes it would be 256 colors. I was thinking shades but it would probably be 254 shades between white and black. Never Mind!

To avoid stepping I usually take a greyscale image and introduce a slight RGB noise using a filter then blur it using gausian blur. Using the noise cuts down on the amount of blur you need to apply because blur only works on areas that have varying pixel values.

[JLT]

An 8bit pattern can produce jaggy results, on both the screen and carving, where the 16bit won't. You want to manipulate the data with the highest resolution possible to avoid rounding errors that result in jaggy results.

Jeff,

A follow up question, if you don't mind... Does the X/Y resolution also come into play? Ie, if I create two images, both 100x100 pixels, in 8-bit gray scale and 16-bit gray scale, does the 16-bit image have an advantage, or does the advantage come only with a combination of higher resolution image and more granular gray scale? (The reason I ask, is that I happen to be writing a program to generate gray scale images, and am trying to understand the best means of achieving the desired smooth results in designer...)

Tx, Jon

[JLT]

FWIW. Picture's worth a thousand words... spent a few minutes writing a program to generate the same image in both 8 bit (0 to 255) and 16 bit gray scale (0 to 65535), and imported them into designer, setting Depth to 0.75 and Height to 999. Not much difference at all, although if you look closely enough, there are a few pixels here and there that make the 16 bit gray scale a whisker better, if even that. Of course, that being said, if I understand this post, the value in the 16 bit image comes with scaling and merging of images, resulting in better rounding of numbers representing the Z height... There's also the possibility that my choice of image doesn't reveal the strength of the 16 bit images, or that my 16 bit image isn't really 16 bit (the only way I can tell is by looking at the image properties...)