ingwa@prefix.liu.se (Inge Wallin) (01/07/88)
The current disussion about circles made me think about something else
in the same genre. How do you, efficiently, draw lines, arcs, circles
and so on with a line width bigger than 1?
Although I have read many articles about fast line and circle drawing
algorithms, I have never seen anything about thicker ones. Neither
have I seen any algorithms in any computer graphics books. I find this
a bit strange since such algorithms seem to me to be very important.
An obvious example is in a laser printer.
Could somebody please explain how this is done, or give me some
pointer to where I can find the information?
ADVAthanksNCE
/Inge Wallin
mail address: ingwa@majestix.liu.senelson@sun.soe.clarkson.edu (Russ Nelson) (01/08/88)
In article <694@prefix.liu.se> ingwa@prefix.liu.se (Inge Wallin) writes: >The current disussion about circles made me think about something else >in the same genre. How do you, efficiently, draw lines, arcs, circles >and so on with a line width bigger than 1? The typical technique is to generate a stream of pixels and run them through a routine that plops (it's a technical term that means to OR or AND NOT a specific shape) something down. In Painter's Apprentice (A soon to be released paint package for the PC's EGA <shameless self promotion>), I have a VERY efficient routine to plop a 16x16 bit shape down. For example, the code is split into the following cases: clipped/non-clipped 16 bits wide/24 bit wide OR/AND NOT/other I also pre-shift 8 different copies of the shape, one for each different byte boundary. (Oh, for a barrel shifter!). I use this code even though I have a "Chris Pike" style compiling blit. For obvious reasons, the speed of this code dominates the line drawing speed. Every cycle you can save in this routine will be amply repaid. -russ
kgdykes@watmath.waterloo.edu (Ken Dykes) (01/10/88)
In article <694@prefix.liu.se> ingwa@prefix.liu.se (Inge Wallin) writes: >in the same genre. How do you, efficiently, draw lines, arcs, circles >and so on with a line width bigger than 1? > >Although I have read many articles about fast line and circle drawing >algorithms, I have never seen anything about thicker ones. Neither >have I seen any algorithms in any computer graphics books. I find this >a bit strange since such algorithms seem to me to be very important. >An obvious example is in a laser printer. > > /Inge Wallin I myself a few years ago during a night decided that i needed to do this in some page-printer software i was doing. Not having taken any graphics courses, and not having a reference library on-site where i was working, i ended up doing the obvious "concentric circles" (you now, plotting N circles simultaneously where N=thickness) this hack has bothered me ever since everytime i think about this software. in the next few months i may be getting an opportunity to revise this program so i too would appreciate info/references. (especially ones that dont assume a graduate degree in computer graphics) (ps: i wont mention that even for "simple" circle drawing i suspect that i used the most brain-damaged childs algorithm one can dream up with no sleep). [and if it turns out that i *am* using the most optimal way for what i need -- i will loose all faith in the mystery of computer science! :-] -ken -- - Ken Dykes, Software Development Group, U.of.Waterloo kgdykes@watmath.uucp kgdykes@water.bitnet kgdykes@waterloo.csnet
holloway@drivax.UUCP (Bruce Holloway) (01/12/88)
In article <694@prefix.liu.se> ingwa@prefix.liu.se (Inge Wallin) writes: >The current disussion about circles made me think about something else >in the same genre. How do you, efficiently, draw lines, arcs, circles >and so on with a line width bigger than 1? We do it by drawing a rectangle with the desired width, the same direction as the line. This makes patterned lines (-.-.-., etc.) difficult, so we don't do 'em. - Bruce -- ******************************************************************************* * Bruce Holloway - Terminal Netnews Addict uunet!amdahl!drivax!holloway * * ALBATROSS, ATARI*TROS @ Plink ALBATROSS @ Delphi * *******************************************************************************
kjepo@portofix.liu.se (Kjell Post) (01/12/88)
In article <16183@watmath.waterloo.edu>, kgdykes@watmath.waterloo.edu (Ken Dykes) writes: > In article <694@prefix.liu.se> ingwa@prefix.liu.se (Inge Wallin) writes: > >in the same genre. How do you, efficiently, draw lines, arcs, circles > >and so on with a line width bigger than 1? Why not use a thicker pencil? What I'm saying is that you can print a larger dot instead of just a pixel. This is the method used by ditroff where they draw lines, arcs and b-splines with the '.' character. By altering the fontsize you get different line widths. -- ----------------------------------------------------------------------------- "The nice thing about standards is that you have so many to choose from" Dept of Computer & Info Science ...liuida!majestix.liu.se!kjepo Univ of Linkoping, Sweden (kjepo@majestix.liu.se)
johng@ecrcvax.UUCP (John Gregor) (01/17/88)
In article <696@portofix.liu.se> kjepo@portofix.liu.se (Kjell Post) writes: >Why not use a thicker pencil? > >What I'm saying is that you can print a larger dot instead of just a pixel. >This is the method used by ditroff where they draw lines, arcs and b-splines >with the '.' character. By altering the fontsize you get different line widths. That may be fine for output to a laser printer, but it is extremely inefficient for any sort of interactive graphics. Here is an example: Assume your big dot is: ## #### Also assuming square pixels #### ## Then drawing a line (slope == 1) would look like: 11 2211 23221 23232 The numbers correspond to the number of times the 12232 pixel is written. 1122 11 Even though there is a fix for lines, this scheme is completely impractical for curves. I recommend scan converting the polygon. John Gregor johng%ecrcvax.UUCP@germany.CSNET
kjepo@prefix.liu.se (Kjell Post) (01/19/88)
In article <471@ecrcvax.UUCP> johng@ecrcvax.UUCP (John Gregor) writes: >In article <696@portofix.liu.se> kjepo@portofix.liu.se (Kjell Post) writes: >>Why not use a thicker pencil? >Even though there is a fix for lines, this scheme is completely impractical >for curves. I recommend scan converting the polygon. > > John Gregor > johng%ecrcvax.UUCP@germany.CSNET I wouldn't be too sure about that. Some of the modern graphic processors can bitblt a pattern as fast as they set pixels. -- ----------------------------------------------------------------------------- "The nice thing about standards is that you have so many to choose from" Dept of Computer & Info Science ...liuida!majestix.liu.se!kjepo Univ of Linkoping, Sweden (kjepo@majestix.liu.se)
rmf@actnyc.UUCP (Robert Fuhrer) (01/29/88)
This is my first posting, so I respectfully request no flames please if the following does not represent the optimal (or even a working) solution: Well, it seems to me that, especially in the case where the pen's "footprint" is larger than 1x1, and is a pattern, e.g., a grey pattern of alternating white & black pixels, you CAN do something: Suppose, when drawing a line and incrementing in the vertical coordinate (similar concept applies when incrementing the horizontal of course), that you draw successive 1-pixel-high horizontal bands from your pattern. When drawing conics, corners may require special treatment (such as plotting the entire pen's footprint when changing from incrementing x to incrementing y?), but I think the general approach works. Comments, anyone? Am I on the right trail? * No snappy sign-off *