View Single Post
Old 29th May 2005, 15:10   #231  |  Link
MOmonster
Registered User
 
Join Date: May 2005
Location: Germany
Posts: 495
@tritical
Oh, I also thought about this problem, but I wasn´t able to test it before. I thought, we still get the blended field, if we compare the combing levels for both fields, but I seem to be wrong.
If MIN(0,4) is really significant higher than MIN(1,2,3), would you also test the field C and than compare the differences or would you use a combing difference parameter to decide, if the test of field C is really necessary.
Your method seems to be accurate enough to find the blend fields.
By trying to match A & B or B & C we get the the combing levels for 0 and 4 and now compare it with MIN(1,2,3) to find the blending.
Maybe you could make this method a bit more effective.
If B is really a blend of A and C, maybe we could compare the combing levels from matching A with B and B with C to find out how strong the blending factor is.
If B is for example an A:C => 30:70 blend, the combing level from matching B with C should be much higher than for matching A with B.
So, it shouldn´t be necessary to create three new frames. We should be able to use an adaptive blend factor by comparing the combings. The rest would work how you said it.
Maybe it is possible to choose the right blendfactor a bit more differenced, for example choose between 20, 35, 50, 65 and 80 percent, but it is only an idea.
If C is the blend and not B, the blending factor should be unimportant, because the combing level should be still much higher than for the blended field, so we could work the same way.

Thank you for your work, tritical.

Last edited by MOmonster; 2nd June 2005 at 15:38.
MOmonster is offline   Reply With Quote