Staff Rastral sizes in Digital Engraving apps

[benwiggy]

As we all know, metal plate engraving used sets of different punches in different rastral sizes, numbered from 0 (or even 00) to 8 (or even 9!). I seem to recall that Sibelius uses these in its setup. I don't know whether Dorico does. Finale ... does not.

Are the continued use of these sizes beneficial, in a digital age? I've worked with graphic designers who set entire books in 11.75pt on 14.25pt leading, because with digital typesetting, there are no discrete limitations: we can use any number we can conceive.

Similarly, in music, we can makes staves to be 22.16pt high, if we so choose. However, each of the old rastral sizes frequently was used for particular purposes, and the rasters were a useful gauge of size-for-function.

Would anyone here wish to see a drop-down menu in Finale (and Dorico, if it doesn't already) that let you set a traditional rastral size?

[tisimst]

I've thought about this numerous times. I think this would be cool, but only if there were also some kind of context description of the selected size. Otherwise, it's just another number.

[Knut]

As we all know, metal plate engraving used sets of different punches in different rastral sizes, numbered from 0 (or even 00) to 8 (or even 9!). I seem to recall that Sibelius uses these in its setup. I don't know whether Dorico does. Finale ... does not.

Are the continued use of these sizes beneficial, in a digital age? I've worked with graphic designers who set entire books in 11.75pt on 14.25pt leading, because with digital typesetting, there are no discrete limitations: we can use any number we can conceive.

Similarly, in music, we can makes staves to be 22.16pt high, if we so choose. However, each of the old rastral sizes frequently was used for particular purposes, and the rasters were a useful gauge of size-for-function.

Would anyone here wish to see a drop-down menu in Finale (and Dorico, if it doesn't already) that let you set a traditional rastral size?

I don't think Sibelius uses astral sizes. It does use millimeter by default, though.

I thought about this as well. I can see some small point in including them in a scoring application purely as a historical reference, but otherwise I don't really see these rastral sizes having much practical value in the digital age. Every engraver should of course know their approximate equivalents in millimeters, inches or points and when to use the different sizes, but you don't need this system for that. There might still be those who at some point made the transition from manual to digital engraving, who would welcome an option to chose between the standard rastral sizes, but I'm guessing they get by just fine without them. For most of the younger generation I don't think they matter at all.

BTW, Dorico uses points exclusively to determine staff size, which makes a lot more sense than using millimeters or any other value.

[benwiggy]

Thanks. That's helped clarify my thoughts.

[OCTO]

Ben, that is very interesting. The old way of rastral sizes had its logical purpose. One of the "misses" of the digital engraving applications is that they allow resizing to any size equally. Also it is allowed (in Finale) to have each system resized differently and to resize small staves.
The problem is in the optical result. Smaller staves cannot be reduced proportionally because in that way they loose the weight. Therefore, a deep knowledge of rastral use would result in correct staff weight AND compensation in the reduction power. That is what the rastral sizes could be of help in my opinion: simply said, you cannot reduce to any size and if you reduce some optical compensation must be applied.

[Knut]

Ben, that is very interesting. The old way of rastral sizes had its logical purpose. One of the "misses" of the digital engraving applications is that they allow resizing to any size equally. Also it is allowed (in Finale) to have each system resized differently and to resize small staves.
The problem is in the optical result. Smaller staves cannot be reduced proportionally because in that way they loose the weight. Therefore, a deep knowledge of rastral use would result in correct staff weight AND compensation in the reduction power. That is what the rastral sizes could be of help in my opinion: simply said, you cannot reduce to any size and if you reduce some optical compensation must be applied.

I agree. And insofar as the traditional rastral sizes will reinvigorate the focus on this principle within the world of computer-based music engraving. I don't see why they shouldn't be included.

It's worth noting, however, that even though Lilypond considers what OCTO is talking about, the application uses point size rather than raster numbers as it's reference.

[tisimst]

It's worth noting, however, that even though Lilypond considers what OCTO is talking about, the application uses point size rather than raster numbers as it's reference.

This is true. It wouldn't be difficult to create a simple lookup table that grabs the proper pt size of the chosen rastral size.

[OCTO]

One of the possible functions that the rastral sizes could have in the digital engraving is determination of the laser printer fixed outputs, conformity with printers. You can have power of 300, 600, 1200, and so on dpi print outs. And this is very important with lines. In Finale you can have ANY line weight that is sent to the printer, and than the printer must either add or remove points to print it out. And that is bad.
Rastral sizes that would support printers' outputs would be something to consider in the future.

[benwiggy]

In Finale you can have ANY line weight that is sent to the printer, and than the printer must either add or remove points to print it out. And that is bad.
Rastral sizes that would support printers' outputs would be something to consider in the future.

I'm not sure I follow this.
Firstly, using layout settings that are device dependent is a retrograde step. That would mean different documents for different laser printers.
Secondly, how is this "bad". Even a 300dpi printer can print a line that is 0.25pt thin. Most printers these days are at least 600dpi - so 0.125pt increments -- or 1200dpi, with dots that are 1/16th of a point.
If I set a 0.5pt line, it's going to be 0.5pt on all three printers. No more, no less. If I deliberately choose a value between the dot pitch of a 600dpi printer, then OK, it's going to round things up or down.
But are you telling me that that 1/16th of point is noticeable? I was always told that dots at 1200dpi, at a hand-held distance were too small for the eye to resolve.

[OCTO]

I'm not sure I follow this.

I am sorry if my english is not on the level of what I want to explain. And what I explain, I hope I am not mistaken.

This is just a pure calculation and improvisation:
Looking at long, thin hairpins we can easily see "breaks" even in 1200dpi printers. It is really visible.
But here I think about the straight lines such as staff lines, bars, stems etc. As many printers have 600 dpi, let us take this as basic for the following calculation.
1. If Finale could cut one inch into 600 dots, so Finale would place a line-width only inside of that dots. It doesn't mean that Finale uses bitmaps, but it means that Finale could execute one line exactly at 1/600 dpi or more of whole numbers: or 2/600 or 10/600. 600/600 gives us one fat line of one inch. 300d/600dpi gives us 1/2 inch of width.
2. It gives us that printer of 1200 dpi needs two laser jet beams to cover one line of 600 dpi, because the printer of 1200 dpi does just 1/2 size of 600dpi-printer beams. It is perfectly performed. Even higher resolution printers can do the same: 4800 dpi laser beam does 1/8 of 600dpi-printers, and so on.
3. On the contrary: if you have a line width that is 1/601 dpi, or 1/601,37 dpi, so even higher resolution printers need to compensate it and re-calculate into their laserjet beam's size.

Even a 300dpi printer can print a line that is 0.25pt thin. Most printers these days are at least 600dpi - so 0.125pt increments -- or 1200dpi, with dots that are 1/16th of a point. If I set a 0.5pt line, it's going to be 0.5pt on all three printers.

Yes, only if the line is placed exactly at the "raster"-size of 1/300dpi. If your line starts at 1,33/300 dpi, than none of your three printers will execute it exactly.
4. That would mean that "rastering" of one inch into 600 dots inside of Finale for straight lines would be a perfect suited for any printer.
5. We could use that as a basic for our rastral sizes, for instance 600/2, 600/3, 600/4, 600/5, 600/6, 600/8, 600/10, 600/12, 600/15, 600/20...
6. Than: to make a smaller staff size, we could not use just 100% reduction of "something" arbitrary but of something exactly. Let us say that this following lines are perfectly visible by eye (more theoretic numbers here!):
- If we have 600/30dpi staff-size-lines it gives us 20dots/width-line. The next reduction will not be executed in % but the next whole number of dots: 600/25=24dots. It means that we could have following sizes:
600/1 =600 dots (one inch size line width)
600/2 =300 dots (one 1/2 inch size line width)
600/3 =200 ...
600/4 =150
600/5 =120
600/6 =100
600/8 =75
600/10 =60
600/12 =50
600/15 =40
600/20 =30
600/24 =25
600/25 =24
etc

I hope it is more, hope not less, understandable...

This is just an idea, nothing I firmly stand for, but rather open for comments, so please help me if I am mistaken.