lilypond-user
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: How to get irrational tuplets inside a regular meter like 9/8


From: Andrew Bernard
Subject: RE: How to get irrational tuplets inside a regular meter like 9/8
Date: Tue, 8 Nov 2016 21:08:54 +1100

It's a pity you have shown aggressive contempt for programmers and computer 
scientists in general on this list, and yet you seem to want to use computers. 
Any introductory course in numerical analysis will teach you that floating 
point calculations on computers are inherently inaccurate, due to the design 
and limitations of computers. Floating point is an approximation to real 
numbers, not an exact map. The people you dislike so much have devoted decades 
of study to floating point theory and chip design. The scientific community has 
been aware of these issues and topics since the 1960's. For example. you cannot 
compare two floating point numbers for equality - you can only test to see if 
they are within a specified epsilon of each other. There are papers on floating 
point accuracy that are considered to be essential reading for any programmer. 
I am able to send you the links should you care to avail yourself of that help.

Therefore, claiming it is a lilypond problem that floating point calculations 
is incorrect, and lacks insight, and is simply an empty opinion based assertion.

Andrew



-----Original Message-----
From: lilypond-user [mailto:address@hidden On Behalf Of mclaren
Sent: Tuesday, 8 November 2016 6:29 PM
To: address@hidden
Subject: How to get irrational tuplets inside a regular meter like 9/8

An alternative method involves simply adding in the required rests, rather than 
squishing or stretching the entire measure with a large tuplet. Since this 
method is easy to calculate exactly because you're just adding and subtracting 
various integer fractions, it always produces guaranteed results. With very 
large tuplets, there seem to sometimes be floating point inaccuracies in the 
calculations done inside Lilypond, so sometimes the large tuplet method doesn't 
work exactly and barlines don't precisely add up. But with this method, 
barlines always line up exactly and no extreme unction is required to add 
barlines or to fiddle with any kind of make-moment value to get the barlines 
right.





reply via email to

[Prev in Thread] Current Thread [Next in Thread]