Slurs -- Line cap styles & Line join styles

[q]

With Apple's Quartz it's possible to apply a Round Cap style so open paths (such as those used in slurs) can terminate without a blunt end.

And it's possible to apply Round Join style to prevent long spikes at acute angles. This might prove handy if you ever choose to draw slurs using a single closed path rather than a group of open paths.

The result of Round Cap style on the slur below (from a Windows PDF) is small but significant. In the first image the paths have butt caps; in the second they have round caps.



For more information see:


http://developer.apple.com/documentation/graphicsimaging/Conceptual/drawingwithquartz2d/dq_paths/dq_paths.html

Table 3-2: Line join styles

-- explains the Round Join style.

Table 3-3: Line cap styles

-- explains the Round Cap style, one of the three line caps styles for the ends of open paths; butt cap is the default.

I assume there are similar properties in the Windows drawing system.

q

[Doug Kerr]

Hi, Q,

With Apple's Quartz it's possible to apply a Round Cap style so open paths (such as those used in slurs) can terminate without a blunt end.

And it's possible to apply Round Join style to prevents long spikes at acute angles. This might prove handy if you ever choose to draw slurs using a single closed path rather than a group of open paths.

The result of Round Cap style on the slur below (from a Windows PDF) is small but significant. In the first image the paths have butt caps; the second they have round caps

Looks good to me.

I assume there are similar properties in the Windows drawing system.

I'm not sure there is such a thing as the "Windows Drawing System", although the GDI (graphics device interface) sort of has that mantle.

In my Windows-based vector illustration program (Micrografx Designer 7.1), one can in fact apply a round end to a finite-width stroke open path. One can also have a closed path with a finite-wide stroke and apply a round join to the joints (although there happens to be a bug in it).

The bug is that the round join doesn't happen at the joint that is somehow deemed "last". My workaround is that after the path is finished, I break it in the middle of a segment. That becomes the "last" joint and is where the round join fails. But since that it just a "continuation"of the path (control vectors collinear, for one thing), the shape of the join there is moot.

If I have to modify the shape of the figure, I delete the added joint, reshape the real figure, and then put the extra joint back in.

Best regards,

Doug

[q]

I'm not sure there is such a thing as the "Windows Drawing System", although the GDI (graphics device interface) sort of has that mantle.

Thanks for clarifying Doug. GDI is the system to which I was referring, but my Windows parlance is far behind my Mac lingo.

In my Windows-based vector illustration program (Micrografx Designer 7.1), one can in fact apply a round end to a finite-width stroke open path. One can also have a closed path with a finite-wide stroke and apply a round join to the joints (although there happens to be a bug in it).

I'm not familiar with the term "finite-width stroke."

The bug is that the round join doesn't happen at the joint that is somehow deemed "last". My workaround is that after the path is finished, I break it in the middle of a segment. That becomes the "last" joint and is where the round join fails. But since that it just a "continuation"of the path (control vectors collinear, for one thing), the shape of the join there is moot.

If I have to modify the shape of the figure, I delete the added joint, reshape the real figure, and then put the extra joint back in.

Don't quite follow. I've never had to tinker with an open path to get caps to work. Are you using cap styles are for open paths, and controlling angled joins with the Line Join styles? (sometimes called miters.)

q

[Doug Kerr]

Hi, q,

I'm not sure there is such a thing as the "Windows Drawing System", although the GDI (graphics device interface) sort of has that mantle.

Thanks for clarifying Doug. GDI is the system to which I was referring, but my Windows parlance is far behind my Mac lingo.

In my Windows-based vector illustration program (Micrografx Designer 7.1), one can in fact apply a round end to a finite-width stroke open path. One can also have a closed path with a finite-wide stroke and apply a round join to the joints (although there happens to be a bug in it).

I'm not familiar with the term "finite-width stroke."
This is contrasted with a path that is (in my drawing program) defined as of "hairline width" (numerically: width =0), meaning mathematically of zero width, to be actually presented as the smallest pixel width that makes it reasonably visible at the scale of viewing or printing. (More about this later.)

The bug is that the round join doesn't happen at the joint that is somehow deemed "last". My workaround is that after the path is finished, I break it in the middle of a segment. That becomes the "last" joint and is where the round join fails. But since that it just a "continuation"of the path (control vectors collinear, for one thing), the shape of the join there is moot.

If I have to modify the shape of the figure, I delete the added joint, reshape the real figure, and then put the extra joint back in.

Don't quite follow. I've never had to tinker with an open path to get caps to work. Are you using cap styles are for open paths, and controlling angled joins with the Line Join styles? (sometimes called miters.)

Firstly, this applies to a closed path join, not an open path. (I may have not made that clear.)

In any case, don't get tangled up by this. This is a bug in my drawing program, not any conceptual notion or a property of the curve defining language. (And I find just now there is in fact a way to avoid it, other than the "kludge" I described to you before.)

As a matter of interest:

I have the choice of three kinds of caps for a curve: flat, square (extends beyond the anchor point by half the width of the stroke) and round.

I have the choice of three kinds of joins for curves (whether the overall curve is closed or open): round, bevel, or miter.

If I am working with zero-width paths, I can still set both the cap and join property, and they remain properties of the curve or figure, but have no significance while the stroke width remains at zero. But if I later set a finite stroke width, those properties come into effect.

Sorry to have set you off on a wild goose chase.

Best regards,

Doug

[q]

As a matter of interest:

I have the choice of three kinds of caps for a curve: flat, square (extends beyond the anchor point by half the width of the stroke) and round.

Yes, and the Round Cap extends beyond the anchor point by a radius equal to half the stroke-width.

One can also have a closed path with a finite-wide stroke and apply a round join to the joints (although there happens to be a bug in it).

I'm not familiar with the term "finite-width stroke."

This is contrasted with a path that is (in my drawing program) defined as of "hairline width" (numerically: width =0), meaning mathematically of zero width, to be actually presented as the smallest pixel width that makes it reasonably visible at the scale of viewing or printing. (More about this later.)

... thus a line in the Euclidean sense. Something we can represent and conceptualize, but cannot see, for it has no breadth.

I have the choice of three kinds of joins for curves (whether the overall curve is closed or open): round, bevel, or miter.

If I am working with zero-width paths, I can still set both the cap and join property, and they remain properties of the curve or figure, but have no significance while the stroke width remains at zero. But if I later set a finite stroke width, those properties come into effect.

This is exactly correct.

In Adobe Illustrator parlance, when you add a weight (i.e. width) property to a line, it is called stroked. A line without a wieght property is called unstroked — but zero-width makes perfect sense for the latter. Finite-width is a term I'm less comfortable with, because I immediately begin to ponder the meaning of infinite width. Too bad Illustrator didn't use the term width instead of weight.

Also true in Illustrator, only stoked paths are affected by the non-default join and end styles.

q

[q]

 
P.S. Have fun with compound paths and their directions.

Compound paths were a core concept in PostScript that made fonts possible. While there was no transparency in PostScript 1 (or Type 1 fonts) a closed path could cut a hole though another, when grouped as compound paths. This allowed for charactors like O. (One just needs to make sure the path direction (winding?) oppose each other. They should by default.)

http://livedocs.adobe.com/en_US/Illustrator/13.0/help.html?content=WS714a382cdf7d304e7e07d0100196cbc5f-644e.html
http://livedocs.adobe.com/en_US/Illustrator/13.0/help.html?content=WS714a382cdf7d304e7e07d0100196cbc5f-6450.html

q

By the way, Adobe's entire documentation is available via LiveDocs.

[Nor]

I'd love John@developpement make ROUND caps rather then butt caps, that way would fit standard music engraving and handwritten style. And I'd love to see hat happen to the triplet arced bracket especially for handwritten style and TIES too!

I'm ready to make a SAMPLE inside a font for easy use, then John will have to tell Encore how to draw the shape in different angles and sizes.