|
DataMuseum.dkPresents historical artifacts from the history of: Rational R1000/400 Tapes |
This is an automatic "excavation" of a thematic subset of
See our Wiki for more about Rational R1000/400 Tapes Excavated with: AutoArchaeologist - Free & Open Source Software. |
top - downloadIndex: ┃ M T ┃
Length: 1246 (0x4de)
Types: TextFile
Names: »MAIL1«
└─⟦a7d1ea751⟧ Bits:30000550 8mm tape, Rational 1000, !users!projects 94_04_11
└─ ⟦129cab021⟧ »DATA«
└─⟦this⟧
└─⟦2f6cfab89⟧ Bits:30000547 8mm tape, Rational 1000, !projects 94-01-04
└─ ⟦d65440be7⟧ »DATA«
└─⟦this⟧
From: "" <@igor:uunet!ai.mit.edu!bkph@ub-gate.UB.com (Berthold K.P. H Date: Wed, 5 Feb 92 07:09:04 EST Date-Delivered: Wed, 5 Feb 92 14:07:16 +0100 To: pam@etoile.Rational.COM Subject: Looking for halftoning algorithm Oh, I see! Well, basically you want to build a thresholding patch that, when repeated, covers the plane without gaps. A point is white if the image brightness is higher than the value at the corresponding place in the thresholding patch , and black otherwise. The values in the patch are computed based on the screen function passed to `setscreen'. But not directly, that is, the actual values returned by the screen function are not used, only their order is important. Some of the tricky aspects involve figuring out, for a given user specified frequency and angle, what nearby frequency and angle to ACTUALLY use which leads to a patch that is not too large. On low end machines, the patch is usually of such a size that there are twice as many patches as halftone dots in a large image area. On high end machines, the patch covers several cycle in order to allow closer approximation of user specified frequency and angle (and also increase number of available grey levels).