Re: rand() pattern in texture?

From: codemagic (codemagic_at_gmail.com)
Date: 10/13/04 

Date: 13 Oct 2004 04:52:13 -0700

I attempted to reply before but google groups beta messed up somewhere.
oh well.

Anyway, after more investigation i determined the following:

1. using the high order bits of the result of rand() does indeed remove
the anomaly. Thanks all!
2. I notice that the anomaly only occurs (or is obvious) with rand() %
c if c is a power of 2. Maybe this is related to the texture being a
power of 2 in size? Example: rand() % 7 fine, rand() % 8 horrible.
3. rand() % 2 looks kinda random, certainly not 1, 0, 1, 0, 1, 0... the
win32 implementation has probably been improved at some stage.
Cheers,

Tom