Re: Barline scans

[Han-Wen Nienhuys]

Marnen Laibow-Koser escreveu:
>> I'll keep it in mind when I'm fiddling with fonts
>> and blackness the next time.  It might be a good idea to tune the bar
>> size down a bit (say: to 1.75), but I'll have to have a good look at
>> the output first.
> 
> Is that 175% of the staff lines?  If so, I suggest that it's still too
> thick.  Of the editions I've so far scanned, the Henle edition has the
> blackest appearance, but even its barlines are only 136% of the staff
> line thickness.

Hello,

I had another look at a lilypond print-out locally, and I think that
with the settings as is, it looks good, not in the least ugly.

*However*

I already discussed some of this with Jan, and we found that printer
artefacts might also be a factor: Jan's printer doesn't produce even
output, lines near the right edge being thinner.

In my case (HP Laserjet 2100 @ 1200 dpi), it seems that horizontal
lines are printed approximately 10% thicker than vertical lines,
regardless of the method of generation. (*) This causes my staff lines
to be proportionally a little thicker.  This suggests that the
stafflines should either be 10% thicker or barlines 10% thinner,
relative to an 'ideal' device.  Unfortunately, I don't know how much
devices deviate from ideal. Changing settings may make things worse
other people.  Even more so, lots of manufacturers use the same parts,
and it's possible that most of them have the same deviation, making
adjustments turn out worse for almost everyone.

I would welcome some data points, so I am attaching a PostScript
document which prints lines at current bar size (0.19 staff space)
horizontally and vertically. You can print this out, and scan it back
in, to determine what kind of bias the printer introduces.  To be sure
that your scanner does not introduce distortion, it's best to scan it
twice (once rotated 90 degrees), and compare.  There are two files,
one uses PostScript lines and one uses rectangeles.

I was going to ask you to run a test like this, so we could determine
what kind of distortion your printer was introducing, until I saw your
scan of a LilyPond print at

  http://doko.ebon-askavi.homedns.org:8080/1200/Lilypond.png
  
All the vertical lines appear extremely jagged, suggesting that you
used an inkjet printer.  If I measure between the extremes of the
jaggies, I see 26 to 29px, which is well over the 22px we expect the
lines to be. For comparison, I put up a fragment of the 2400dpi scan
of the raster for my printer at 

  http://lilypond.org/~hanwen/raster-scan-0092.jpg

Can you print results on a laser printer before criticizing them?

Thanks,


(*) The barlines are printed as Postscript rectangles, while the
stafflines are lines.

-- 

Han-Wen Nienhuys - address@hidden - http://www.xs4all.nl/~hanwen

LilyPond Software Design
 -- Code for Music Notation
http://www.lilypond-design.com

raster-box.ps
Description: PostScript document

raster-lines.ps
Description: PostScript document