I was just re-evaluating some of my defaults and decided to compare spacing ratios. I had done a similar thing in Finale years ago, but this is really easy to do in Dorico by just adding Note Spacing changes. Anyway, if anyone is interested, here’s a chart comparing some commonly used spacing ratios. Obviously ratios of 1 and 2 are terrible, but there’s quite a bit of aesthetic leeway in between. I don’t currently have it installed, but IIRC Sib doesn’t use a ratio by default and the spacing values don’t really constitute a specific ratio (is it the quarter that is way off?), so 1.3 is an approximation. I think there’s a good spacing plug-in for Sib too, but again I don’t currently have it installed. Just curious if others are using ratios other than the Dorico default.
Ted Ross writes 'Each engraver may use a slightly different proportion; however, all will show an obvious difference between the spacing for notes of different values."
He goes on to give a set of spacing distances for usual note durations (whole note to eighth), which, if my maths is to be believed, vary in proportion between 1.5 and 1.35.
As with the spacing distance itself, different values may suit different musics. Something in 3/8 with lots of small notes requires a different spacing from something in 6/4, say.
Dorico uses Root 2 by default, which is a good fit for most things. 1.3 is too tight, and larger values are too slack, IMO.
Finale uses Fibonacci by default, with a reference width of 3.5 spaces. It also allows you to set individual spacing values for each duration (interpolating for other values).
Yeah, I’ve vacillated between square root of 2, 1.5, and golden ratio over time. I can usually even approximate the age of the file within a 5-yr span or so by looking at the spacing. The original Finale Fibonacci allotment library spacing from the late 90s was actually my file, but then Randy Jones coded a scaling factor so you don’t need to use allotment libraries (unless you want deliberately disproportionate spacing) so that library was sort of defunct as soon as it came out. AFAIK it still ships with Finale though. Looking through old docs, I had some correspondence with Tom Brodhead in 1999 about this too, since his SCORE spacing subroutines were so good.
Anyway, just investigating what the current trends were and if anyone is using something other than the default and why.
At what point does having (forcing?) the same number of notes per system affect the chart above? Would a longer example of each show more noticeable differences even than those visible here?
Only the top system in that example is over 100% and no one is going to use a scaling factor of 2 anyway. As the other systems range from 91.7 (for 1.26) to 97.0% (for 1.73) I’m not sure it would make that much of a difference in the scaling as the bars aren’t really forced into the system, other than the top. Forcing them into systems filled over 100% obviously will skew the results as Dorico will be compressing the bars and not actually be using the ratios for all the spacing.
i would think longer examples with a wider variety of rhythms would be useful for actually determining a definite preference though.
Here’s the opening phrase of “Big Dipper” by Thad Jones, using 1.62, 1.5, and 1.41. Does a greater variety of note durations and accidentals affect your preference? All systems are 86.8-87.5% full so scaling shouldn’t be affected by forced system breaks.
At first I thought I preferred 1.41. But as I look at my scores, it seems I’m usually closer to Fibonacci, which is a good deal less relaxed. Food for thought.
I’ve been 1.5 as a default the past few years. Just trying to re-evaluate if I still prefer it. My initial response is still yes, but I was curious what others preferred. In the original example with limited rhythms and no accidentals, square root of 2 looks pretty good, but I don’t prefer it in the more complicated example. The first bar with the dotted half, eighth, and eighth rest seems particularly unbalanced to my eye at 1.41 for some reason.
Fred, your second example shows clearly that there is a difference between the examples. I just suspect that in some systems Dorico will stretch the notes out (albeit proportionately, which may be why it doesn’t matter) or compress them slightly to fit, just as a hand engraver would. I do appreciate the examples you have put together; they are still informative and therefore useful. Thank you.
For me, I like 1.4.1 the best in the Big Dipper examples. It gives more spacing to the busiest bar, and less to the whole note. But there’s a clear distinction between notes of different duration, like the last two note of bars 2 and 3.
(Really, it’s only bars 4 and 6 where there’s any discernible difference.)
And don’t forget that in any score where there are different note values in other staves, that’s going to reduce the ‘impact’ of a particular ratio. To say nothing of lyrics and other distractions.
