Slashdot Mirror
Linuxmusician.com Interviews LilyPond Authors
jcn writes "Chris Cannam talks to
the authors of
one of the best-known and most ambitious music programs for Linux, the LilyPond score engraving system. Unlike other typesetting software like Finale or Sibelius, LilyPond is not a score editor, it aims to use simple textual description of the music and turn it into the highest possible quality output, automatically.
Han-Wen says: In my opinion, any file format that claims to be universal should have two properties: it should have an expressive structure, so other formats can be expressed in it, and it should be as lean as possible, so that converting from other formats amounts to removing information. I think that MusicXML fits neither. Ouch."
Some F/OSS projects just aim to get a job done, do it, and leave it up to someone else (perhaps less qualified?) to complete things, to produce a complete package that does the job well
Han-wen & Jan have done one of the latter, this is a supreme polished job that's only getting better. Kudos
adult desktops & wallpapers
Why is it that so many Unix/Linux programs (and everything else, for that matter) do not provide simple screenshots on their products websites?
If I'm going to download your program and install it (and in many cases, take time to compile it...) I want to know that it's going to look halfway decent when I'm done.
Why is this so hard for some programmers to understand?
A good example of seperating content from presentation is to use an XML-type file (at least have a structured document model) where the music data is defined. Then, have somthing like an XLS sound stylesheet to define how the data will sound like. As a developer, this would create greater posibilities what I could do with the sound that my application processes.
On a side noce GNoise is a good sound editor that I recommend to anyone doing edeting or large sounds like game-music (that is uncompressed in raw format.)
is there anything like cakewalk available for linux?
It seems to me that they're trying to redefine "score editor" and yet generally, that's what it seems to be, more or less.
While the printed output is asthetically pleasing, it strikes me as an odd technology to persue, because I wonder how many musicians today can actually read music. I'd wager the vast majority of rock musicians can't, and that roughly half of pop musicans can't. I can't, and I've written "plenty" of material and play several instruments. It's not truly a necessity anymore, with a good ear and modern equipment, ideas can quickly be stored for future embellishment or shown to others in the absence of an actual instrument. It's not even necessary for registering with the library of congress, an audio tape will suffice.
Look back up at my post, now look back down, you're on the Internet. Now look back up. I'm a signature.
I don't understand why Lilypond aims to go back to having a proprietary textual format for typesetting music. Most people, I'd imagine, would want to typeset music graphically, as it's just more intuitive that way (I mean, I'm guessing that, for example, getting two voices per staff would be easier in a GUI system than having to manage the text input).
Anyone know of a GUI frontend to Lilypond?
Talented and creative people use Macs, or Windows at a pinch. I shudder to think at the potential deluge of anime soundtrack remixes that would be unleashed by the releasing of production software to the sweaty, overweight and unshaven Linux hordes. Thank fuck for profit motives.
Um, its a music typesetting program, not a sequencer. I am sure it would be fairly easy to convert from a MIDI file to LilyPond, so use a sequencer to get the music into the computer properly, and clean it up in the sequencer, then convert to LilyPond to print it out nice and pretty.
/usr/games/fortune
While I know that this is more of a compositing program--at least from what I read so far...as I have shamefully not RTFA--I'm going to take this opportunity to bitch about the one thing that has been keeping me from making the switch to Linux for all these years:
Audio Apps
I'm no industry elitist that demands ProTools. in fact, I hate protools. The interface leaves much to be desires...granted, i'll buffer that (admittedly harsh) opinion: I'm a huge fan of CoolEditPro.....("eww, PC audio"...I can hear it already),
The underlying audio subsystems are a far cry from what windows offers. And what I experienced with in my limiting dealings with aRTS leaves much to be desired. (Think: latency) And I'm sure that has a lot to do with it....(why hasn't ASIO or an equiv been implemented yet?)
Aside from that all I ask for is a simple audio production suite where i can record something, and then playback and record something else. Simple full-duplex operation. I've been doing it in Windows for over 7 years now.....hell, I did it in DOS with my GUS 11 years ago.
Toss in a little simple single-track editing, some simple effects (Chorus/Flange, Dynamics processing, simple verb and delay, etc) and maintain development of the project and you've won yourself a full-fledged permenent windows convert.....and i'm willing to bet I'm not the only one.
Am I just out of touch? Is there already software out there that does this?
~Dan
Jim claims that the traditional music in Sri Lanka has far greater diversity than its western counterpart. Thus a simple music notation system, in his opinion at least, is far better than a complex rigorous one.
Indefinitely Detained US Citizen
LinuxMusician.com!?!?!?
I'm a penguin fan and all, but there are some things that should not be mixed. Like....
Water and oil.
Acids and Alkali
Nucular [sic] weapons and George Bush.
Linux and Musicians!
Music is not about the tool, particularly tools that aren't themselves musical. I mean, you *could* say: "ViolinMusician" but "LinuxMusician" comes across to me like "GasEngineMusician" or "Cassette Tape Musician".
Just dumb. Sorry. (It's late, Saturday, and I've had a few drinks. So sue me, or as Apple Computer would say, sosumi!)
-Ben
I have no problem with your religion until you decide it's reason to deprive others of the truth.
Well if you want to nitpick. The "intuitive" way for music is the ear, not the eye. But I'm certain that a GUI would be easy to create.
Guys, I am a professional musician who occasionaly makes a few hundred bucks setting out of print scores to finale or sibeleus. I also use linux, and like the open source model.
The problem is that programmers arent creative in this department... those coders all work at apple.
This is never going to get off the ground, and is a hindrance to the adoption of linux by musicians, when in reality things like jack, ardour, and alsa make it an excellent platform for creative types, a la Pd, miller puckette's wonderful synthesis program.
The developers seem to be focusing on making things "right" and in a description language. Fine, but i dont see how this is going to help inspire musicians to use this arcane latex garbage to print out a set of exercises. Most of my musician friends cant even use finale well, so how can one expect the same of this program.
On the other hand, if your objective is to create a framework for music notation software, midi in, etc, etc, then you need to work with people in that community so that you can have more attention and people drawn to that project.
As it stands now, this software is like enlightenment 17... by the time it gets ready, all the interested people and developers will have gone elsewhere or vanished in disgust.
I agree, most amateur rock musicians probably dont read music.... But think about pro musicians for a second, u know guys in orchestras and such, where everyone needs a piece of music in front of them. This program looks like it would make it much easier to look professional for starters, instead of having to make excuses to the conductor or the violin section on why they can't read the music properly because of glaring errors in it.
music monkey n
1. A person who is capable of making music without actually understanding what they are doing.
Also see: code monkey
Where was it inferred that this was supposed to be easier? This is meant for musical typesetting, and, as was mentioned, is not a score editor.
// file: mice.h
#include "frickin_lasers.h"
That might be perhaps the most ignorant comment i've read here.
Linux is *very* useful to musicians. A Prof at my university, a quite well known computer science phd, is also a music prof. His research is 12 century Chinese music (i'm not making this up). He's relying quite heavily on Gentoo on Alpha for a lot of his custom software.
Musicians and Linux mix together just fine.
As a musician, and someone who publishes their own work, why would I go through the effort to use this program? Using Finale with TgTools gives me just about everything I could want in a music notation program.......
It's either on the beat or off the beat, it's that easy.
I moderate therefore I rule!
--
Because the software is free/open source! As you know, that is more important than being useful.
Planet CCRMA http://ccrma-www.stanford.edu/planetccrma/software /
The AGNULA Project
http://www.agnula.org/
Enough toys to keep you busy for a day or two.
Free or not. Typesetting or not. This program has a long ways to go......
It's either on the beat or off the beat, it's that easy.
I moderate therefore I rule!
--
I was reading this, and it basically summed up how good ideas can go horribly wrong. Basically, the authors are trying to make a tool that matches their ideal of music engraving. So, the use LaTeX markup ideas, add in a Scheme interpreter, don't really bother with MIDI import or other standards, focus on one thing to the exclusion of all else and basically come up with a tool that almost nobody will probably use.
Because most musicians just want to make readable scores quickly and effectively. They aren't looking to make works of art. Those people that want engraving, will probably pay an engraver to do so. And engravers have their own tools.
The whole thing seemed to be "we make better printouts that anybody else" seems awfully subjective and not really the main point.
A tool that likely takes 10 times as long to make a simple score for band class (not to mention the huge learning curve) is not a good computer tool for most musicians. A tool that bangs out pretty nice scores fast, that's a good use of software.
Ask yourself why the printed word still exists in the midst of audio tapes and CD's? The score is the stone tablet.
What about the slurs? They look OK to me. Maybe a bit far from the notes, but that can be adjusted by the user.
/usr/games/fortune
That's a bit like comparing a word processor with PostScript output to PDF.
Rodegarden is a proper comparison--if not perfect--that runs on Linux (under KDE), and it does, in fact, export to Lilypond...
Ok, everybody seems to knocking LilyPond so far, so I thought I'd put out my initial opinion. I've been learning LaTeX recently, and in spite of the waves of horror you feel the first time you look at it, it is actually extremely good at what it does. Revelation, I know, but the point is it ISN'T made for high schoolers writing their history reports. Same thing with LilyPond here. It doesn't look easy, but then, typesetting music isn't easy. LilyPond and LaTeX are an order of magnitude less complex, even if the coefficient is higher than, say, MS Word or Finale. I know I would die if I had to write a book in Word. :) I thought, "If someone did it for typesetting, can't it be done for music?"
Also note that this is not intended to be a replacement for Finale, but rather an entirely different way of getting the job done. They've taken to engraving what TeX took to typesetting.
The coolest thing about this project to me is that I was wondering earlier if anything existed.
jagged. Very jagged. Check out Finale or Sibelius's slurs.......not jagged.
It's either on the beat or off the beat, it's that easy.
I moderate therefore I rule!
--
Try zooming in. They only look jagged in Acrobat when it is zoomed out. Did you try printing it? I am sure they are perfectly smooth in the printout. I have printed things produced by LilyPond, and they look beautiful. Nothing is jagged.
/usr/games/fortune
In my opinion, any file format that claims to be universal should have two properties: it should have an expressive structure, so other formats can be expressed in it, and it should be as lean as possible, so that converting from other formats amounts to removing information. I think that MusicXML fits neither.
Am I missing something or are those two properties mutually contradictory? If converting means removing stuff, then the format would have to be a subset of the original, but if it's expressive enough to express other formats, then would it not also have to be a superset?
I basically read that as "It must be both more and less than what we have, and MusicXML is neither of those things"
I think that the real question should be "Does goatse have a vagina?"
A very pressing question indeed.
It all goes downhill from first post
I'm a pretty serious amateur jazz musician, and I do a fair amount of composing and arranging for jazz ensembles of about 8-16 musicians.
LilyPond is not intended for people like me. If you're less serious than I am, LilyPond is definitely not intended for you.
The most popular music notation software is Finale. Finale is buggier than Windows ME and twice as bloated, but once you learn how to use it, it gets the job done. You can enter your notes relatively quickly, tweak them a little, print, and go. While it has some very non-intuitive options, it's straightforward enough that most amateur musicians are able to sit down and click around until they get it to do what they want.
How's the output? Pretty crappy if you don't spend any time playing with it. But if you spend a little bit of time fixing the glaring errors, the result is readable by most musicians.
LilyPond, on the other hand, reads a description of the music in a text-based format, and formats it automatically - using much nicer algorithms than Finale apparently uses. It might take quite a bit longer to get your music input, but the end result will look nice - and will not require nearly as much tweaking.
LilyPond, by itself, is only of use to professional engravers, and only those who are willing to learn how to use it. If somebody ever develops a front-end to LilyPond that's actually integrated (as opposed to something like Rosegarden that can just export to LilyPond's format), then it might be more accessible to the average musician.
Don't get me wrong - I think that LilyPond is great. I just think that a lot of the complaints I'm seeing in this forum are because people don't understand what problem LilyPond is trying to solve and who will benefit.
No, LilyPond is not ready to replace all of the other music notation software out there. But it's one of the best tools for professional music engraving already, and maybe someday it can also be an appropriate tool for the casual user, too.
Because this is the Unix way. It does one useful thing, and does it well. The problem of recording (e.g. from MIDI input) and creating nice printed output can be broken down quite naturally into at least two parts. Separating out the typography part makes it simpler to implement and more reliable, and offers flexibility by not binding it tightly to particular solutions to other problems. The apparent convenience of one big monolithic software package can often be outweighed by its disadvantages.
Ok, you are right. No jags when printing. However, I'm not all that impressed. The example (Standchen), doesn't look spaced well. Measure 17, Measure 7, 8 and 9. Measure 21 where the slur collides with the triplet....
It's either on the beat or off the beat, it's that easy.
I moderate therefore I rule!
--
Yeah, the spacing is a bit weird. Measure 17 it looks like its caused by the lyrics. I don't know why 8, 9 and 10 are so squished up, I would have tried to make it do the same number of measures per line. The slurs I think can be moved. Whoever was typesetting that should have noticed the slur was a problem in 21 and moved it up a bit. 13 is also a bit weird in spacing. It looks like lyrics throw it off. Look at some stuff on Mutopia without lyrics.
/usr/games/fortune
I see a number of programs, such as Graphire Music Press, that would negate your statement.
These are eclectic composers: they blend many musical styles (ranging from medieval hoketus via french baroque to gay nigger music) into new pieces.
Wow, where can I get the CD?
Sigh, gotta love blind positive moderation of copyright violation. I'm sure there's more interesting changes in the "repost." That's two so far!
Bitchslapped. Neat.
I sort of enjoy the troll easter eggs...
I see tools like Lilypond and Rosegarden and other such tools as the audio embodyment of the Unix philosophy of "a tool good at a specific task". This works for stringing commandline tools together, and this works as well as one gets higher level. Imagine for a second collaborationware for musicians, with output everyone can be proud of.
Obviously you didn't bother to read past the first of these FAQs, which is a bit sad. It's exactly as if you said "Why bother using TeX when I can typeset mathematics in Word?"
Well, you can. But no journal will accept your output, because the quality just isn't up to snuff. Likewise, Finale's output is not up to the time-honored standards of music publishing. To musicians who sight-read at a professional level, the difference matters. In the long run, they can't stand it any more than Knuth could stand the piss-poor, headache-inducing math typesetting that enraged him into writing TeX. Quote:
Plus, proprietary programs lock your scores into proprietary formats, which you can't further process and share like this (scroll down the page).Timeo idiotikOS et dona ferentes
Analogous to the world of word processing, this software is more in the category of software like TeX, LaTeX, or even Postscript and PDF, to a lesser extent. This is software made for pretty printing music. It is meant to do this job, and this job alone very, very well. While one could edit it directly (it's not that difficult to work with), that would be something like using a flathead screwdriver on a screw that is clearly a Philips.
What people should do is look for a score editor that can export LilyPond documents. I'll help start you off:
I'm sure there are others out there.
Honestly, I have no idea what TeX is, so, I don't know the comparison.
I do know that a number of companies are using programs like Sibelius and Finale to do their music.
"Time-honored standards of music publishing". Does this include all the errors that continue to make it in publications as well? I swear the Rose etude book for clarinet has errors that are over 50 years old (I've seen a first edition).
Any how, I really don't see this program doing new. I'd much rather use Graphire Music Press for "engraving standards". Hell, I know a couple of people who make their score look great in Finale and get them published, and the publisher has someone HAND WRITE it out.
That kind of cool to be able to search and stuff like that. But can't that be acomplished with MusicXML?
It's either on the beat or off the beat, it's that easy.
I moderate therefore I rule!
--
Timeo idiotikOS et dona ferentes
jagged. Very jagged. Check out Finale or Sibelius's slurs.......not jagged.
You are confusing the artifacts generated by a non-perfect conversion to PDF with problems in the output of Lilypond.
I have been using it to typeset music for around six months now, and it generates pure vector output that is, quite frankly, stunning. Producing music notation that looks good is a surprisingly subtle problem, and so far I have no complaints about Lilypond.
---
If you think Sibelius is better than Lilypond, bully for you. Go use Sibelius or Finale. Really. No one will mind if you do. After all it's about playing music, not just looking at the printed page.
In the meantime there are people who care about getting it right, and who are willing to put in 10+ hours a week for years on hacking it. In addition to practicing and playing, and let's not forget their day jobs. It seems obvious to me that a labour of love will be of a higher quality than work done for hire by some schmoe who goes home after putting in his eight hours. Lilypond is well on its way to proving this.
Unlimited growth == Cancer.
Funny, it seems obvious to me that a labor of love that's good enough for people to get paid to work on is going to be far superior to a bunch of /. readers wanking away in their mothers' basements to produce a program which nobody will ever use because it's too "perfect" (i.e. unusable) to be actually useful for any real-world application.
Ok, well, that would account for like 95% of the musicians I know...
Huh? What issue of "how is a computer supposed to distinguish between a staccato quarter note and an eighth note?"??? Who the hell cares? I want something that I can enter the music into, and it looks good. A lot of programs can do that. Sibelius, Finale, Graphire Music Press, Encore.....a lot.
"You also threw in a cheap shot about lilypond's slurs -- a known deficiency, you didn't add anything. I have no doubt that slurs will soon be fixed."
Aww, sorry. Did it hurt? If there is anything people complain about, it's slurs. And, well, they ain't happening in Lillypod as far as I can see. Slurs and things colliding with other things. Finale and Sibelius have made a lot of headway in this area.
"If you think Sibelius is better than Lilypond, bully for you. Go use Sibelius or Finale. Really. No one will mind if you do. After all it's about playing music, not just looking at the printed page."
Exactly!! So, why would I waste my time using Lilypond? I could whip something out in a number of other programs, and they wouldn't have slur issues.
"In the meantime there are people who care about getting it right, and who are willing to put in 10+ hours a week for years on hacking it. In addition to practicing and playing, and let's not forget their day jobs. It seems obvious to me that a labour of love will be of a higher quality than work done for hire by some schmoe who goes home after putting in his eight hours."
Day job? Oh yeah, thats right, some people have "real jobs". I guess doing music full time doesn't count as a job. Thats great people want to do this sorta thing, but I don't see it being as good as some of the stuff already out there. Free or not, I am not wasting
It's either on the beat or off the beat, it's that easy.
I moderate therefore I rule!
--
I guess we all feel your pain. But even in Word 2.0 days you didn't have to use Word 2.0. There were a lot of great editors much better suited for your needs then. On Mac, Windows, Atari, you name it.
And somehow I think comparing text editing and musical score editing doesn't work ...
I think the people who will most benefit from a tool like this are performers and composers in the academic vein. Someone who's studied theory much isn't going to look at .ly source and freak -- they've already spent years learning how to describe music in an abstract form. After doing Figured bass analysis on chord progressions and learning how to cut up a piece into it's atomic parts, something like this will probably make more sense than any other solution out there.
On the other hand, if someone is just looking for a program that they can play music into from a keyboard, or punch a few notes into without having to know much about how notation is structured, then of course Lilypond isn't the program for them.
Maybe some of you are getting 'ease' confused with 'instant gratification'. The only easy thing about Finale in my mind is that you can start the new score wizard set to 'Piano' and enter in notes within seconds. I won't deny this is an attractive feature. Any point past that though, and you have to learn the program and all it's quirks(and believe me if you're uninitiated, there are a few billion of them). Once you go beyond the first steps, the balance shifts considerably. Where Finale fails is in the ease of getting right all the minor details of a complex score, wheras Lilypond is remarkably consistent and structured.
And since the input language to Lily is open, non proprietary plain Ascii, I imagine usable graphical frontends will become available for those who are vehemently opposed to having to write out scores in a description language. Much like there are tools like Dreamweaver for HTML. But I think if I showed Lily in it's raw form to my old Theory and Orchestration teacher from my undergrad years, he'd fall right in love.
Have you any idea just how difficult it is to convert live music into sheet music?
Aside from the issue of having to figure out timing (that can vary) and key signatures (that can vary), you're still stuck with the very real issue of having to use signal processing, fourier transforms etc. to try and figure out which instruments were busy playing at the same time: computers wind up with the notes *after* interference, remember!
IANAM.
That being said, having RTFA'ed, I think that I can guess at the difference.
It sounds like Finale and friends can put out usable sheet music. If that's fine for you, go for it -- I don't bother with LaTeX if I'm just jotting a quick note to a friend. However, apparently you *really* have to know these programs and put serious time in to coax really, really high-quality output from them. If you're going to put out, say, a collection of sheet music, perhaps Lillypond is what you want -- for engravers, rather than merely musicians.
I doubt that so many people have spent time on this issue without a pretty decent justification.
May we never see th
Honestly, I have no idea what TeX is, so, I don't know the comparison.
TeX and LaTeX are (roughly) the general typographical layout equivalents of Lilypond -- instead of producing musical scores, one produces text and math formulas. Each is a GUIless program that takes a set of plain text input, and produces a rendered, formatted set of output.
Because BSD and Linux lacked a decent free word processor for a long, long time, a lot of people learned LaTeX in the place of a word processor. LaTeX is really intended for extremely high-quality textual output -- the sort of thing that a typesetter would use to produce a book. As Knuth (the author of TeX) put it, he was shocked when people started requiring college students to use TeX for their papers -- he saw TeX as a system that required more work to use but was worth it when you wanted to go the extra distance and get really good output.
LaTeX (a sort of child of/extension of TeX) is commonly required as a format to submit papers in, since it produces high-quality output and can be used to allow certain formatting tasks that are very difficult with a word processor. Another benefit of LaTeX is that it uses a text format to represent math formulas, and so LaTeX is often used on USENET and other text-based forums to represent complex mathematical formulas. It's relatively quick to enter forumlas into LaTeX, so it's become dominant in the computer science and mathematical fields.
If you've ever run into CSS, you may understand some of the benefit of using a LaTeX-style system over a more conventional GUI. It lets you assign meaning to elements of your document that can let you perform very powerful operations later. For example, have you ever used stylesheet functionality, present in most word processors? This is a very limited form of this. You can "create a new style", assign text to be of that style, and later tweak the format of an entire book easily by changing properties of the style sheet.
May we never see th
I should have said that TeX/LaTeX take "text input", not "plain text input". They have markup embedded in the text.
May we never see th
The thing with most unix programs is that they generally don't have a huge visual element to them anyway (witness grep, tar etc.), whereas sheet music is inherently visual.
More to the point, this forces you to use a linear approach - you can't see what you're score looks like until you've written it, as opposed to the real-time effect that you can get in Sibelius, for example. It also limits the effectiveness of graphical front-ends - to be WYSIWYF they have to reinvent the wheel.
You missed the part about taco's cock.
When in doubt, parenthesize. At the very least it will let some poor schmuck bounce on the % key in vi. (Larry Wall)
heh, i tryed such converters, oh, maybe 2-3 years ago. the conversion was quite poor. maybe it's time to give it another go.
excellent output quality, lilypond has a couple of advantages that
haven't been mentioned in the discussion so far:
Yes, it was a fair bit of work to set it all up (I even use m4 which may not be everyones cup of tea) But after that, producing a new piece of sheet music is really much faster and easier than with the traditional notation packages, and the result is a lot better.
Here's the introduction:
(Emphasis mine.)ABC is an extremely popular format for collecting and exchanging tunes. There are Large Tune Repositories and Tune Search Engines using ABC.
In the course of every project, it will become necessary to shoot the scientists and begin production.
Congrats to lily's developers for all their hard work.
I just stumbled across this online music composition generator.I wonder Jan and Han-Wen are aware? Looks interesting for quick and dirty snippets, perhaps great for a beginner's music comp class. It also appears that GUIDO has a more "natural" TeX-like command set, things like \slur, \staccato. But judging by the examples, I think lily is a bit more versatile, in the end.
To-do List: Receive telemarketing call during a tornado warning. Check.
Huh? What issue of "how is a computer supposed to distinguish between a staccato quarter note and an eighth note?"??? Who the hell cares? I want something that I can enter the music into, and it looks good.
But not good enough for professional publishing. That's the point that you remain steadfast in your determination to not get. If you'd bothered to actually RTFA, this point would be made clear through descriptions of these issues in conversations with professional publishers.
But I know, reading makes your brain hurt. Sorry.
But why go to so much effort? There are plenty of great programs out there that offer input and have engraving qualities. Finale, Sibelius and Graphire Music Press.
Engraving quality from Finale? HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA.
Sorry. That's just some damned funny stuff.
It is not even that. The problem is with Acrobat Reader displaying stuff on the screen. If you zoom in, you can see that the jaggies are reduced. If you print it, you get perfectly smooth slurs with no jaggies whatsoever. The PDF is fine, but for some reason Acrobat Reader makes it look horrible.
/usr/games/fortune
I agree with Han-Wen's criticisms of MusicXML, (some of which he voiced previously in a response to the short article I submitted in January). I readily admit that the blurb had some errors in it; and especially after witnessing the prevailing confusion over the issues involved, I wish I had written a full-length article on the state of free music score publishing and interchange.
MusicXML fails in many ways, but neither Lilypond's native format nor the various binary formats fits the bill, either. My intention in submitting the article was to make people aware that there is currently no open, editable, universal, web-renderable music notation format. Please bear in mind that MIDI is not a music notation format, and is inadequate for the purposes described above. LilyPond is a great program and a high caliber open-source development project which I admire and endorse--this is a lot more than I can say for MusicXML (regardless of the apples and oranges comparison). But I don't think it will thrive until it has a GUI and expands into the markets ruled by Sibelius, Finale, and (to a lesser extent) Encore. In other words, I think that to become a major player, LilyPond must eventually must, in addition to being the superb typesetting program that it is, it must also reach those who want an intuitive score editor.
I'm very please that open source music typesetting and publishing are topics of ongoing discussion (and controversy). Finally, I should mention that I'm affiliated with neither Recordaire nor LilyPond in any way.
You could have provided a link to the pdf, which is what the parent poster was wondering about.
// Jens M Andreasen
mvh
send + more == money?
1) Post a link to something "adult" in a first-ish post thread on a mainpage slashdot article.
2) ???
3) Death (or profit)
The reason it's not easier is because it's meant to do 1 thing well. That is to make publishable music. It is not at Cakewalk to do so.
Besides, there are things that I can do with Lilypond that I can't do with Encore.
Also, having a vast library of publishable quality music (namely Classical music) available would let small time orchestras get their music without having to pay rent for it.
If you read more about Lilypond, the question that is being answered is how do we define music? Do we define it by time/pitch/duration/velocity or is it more than that? It happens to be more than that, though those four variables will give a good approximation to what it should look like.
lilypond looks nice for many things, and i think it's a step in the right direction. the problem is, there's always a rift between what the musicians want to notate and what the software is able to do.
:)
can lilypond notate beams across barlines? can you hide rests? can you make invisible barlines? all this stuff is important to me, since that's the kind of music i write. sibelius does them wonderfully, and i've heard rumors that sibelius' base engine is written in ASM and could be easily ported to linux from OS X.
on the other hand, i have a big problem in that i wrote a lot of stuff in Finale, and then I switched to sibelius, and even the file convertor doesn't work right a lot of the time. if lilypond can offer a good long-term storage format that is easy to read by both humans and computers, it could have a big niche in digital preservation, and be a common point between notation programs.
anyone want to write a finale->lilypond convertor?
This is how the logic works:
1) I'm not a professional musician.
2) I can't read music.
3) I can play instruments to some extent.
Therefore:
4) Proffesional musicians no longer need to be able to read music.
Dumb...you're not in their class, yet you claim authority. Like it or not, most musicians did study music, did have to pass reading exams, did find it a valuable tool for cummunication.
And this continues to be the case.
One interesting example is Danny Elfman, formerly leader of the group Oingo Boingo now the creator of many movies scores. Up to a few years ago, much of his working drafts were scored on paper and then entered into the computer later.
Those scores are critical for the orchestras...it's how they work. They get the scores, sit down, and play it. No one goes around the room humming their part to them.
Han-Wen says: In my opinion, any file format that claims to be universal should have two properties: it should have an expressive structure, so other formats can be expressed in it, and it should be as lean as possible, so that converting from other formats amounts to removing information.
I assume this guy didnt design GIF or PNG then (might have designed JPEG)
Perhaps, but I don't see people who write music, IE: Musicians, using this program at all. There are a number of OTHER programs out there. For "engraving quality" the one that pops to mind is Graphire Music Press. It produces excellent output, and is graphical, and is easy to use.
It's either on the beat or off the beat, it's that easy.
I moderate therefore I rule!
--
Take a look at www.tomplay.com.
What i want is something where i can copy in a sheet of music or a few bars and hear what it would sound like. if you really want someting to teach music students with this would be it because you coul experiment and verifiy ideas or intent.
Does Lilypond, or any other program, for that matter, do a good job of drum music? Don't just drumset, but marching percussion, too? With diddles, shots, pings, rolls, 32nd notes, whatever? I have never seen a program that handles all of these things, and it would be great if there were one.
A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
I'm also writing up my PhD, which also has a lot of equations in it. I knew about LaTeX from my undergrad days in mathematics and was happily writing my PhD in LaTeX until I discovered LyX.
According to the website LyX is a WYSIWYM document processor - What You See Is What You Mean. Basically, it's a Word-like frontend for LaTeX documents. Editing normal text is just like any word processor: ctrl-b for toggling bold, etc. But hit ctrl-m and you go into "math-mode" where standard LaTeX markup is understood and converted as you type into integral signs, sigmas, fractions, etc. Bibliographies, multifile documents, references and tables are all very well handled. It also has hooks for inlining XFig and other graphics including Grace plots.
At the end of the day (or paper), you still get beautiful LaTeX output but without having to worry nearly as much about compile errors. Why? Because, under the hood and behind the GUI you're editing LaTeX.
"Where is the wisdom we have lost in knowledge, and where is the knowledge we have lost in information?"-T.S.Eliot
It doesn't matter that some musicians don't read music. Some musicians can't play keyboard. Some can't sing. Some software developers don't know C!
What matters is that a good number can. I do all sorts: I've written music for a choir, played bass in the theatre, sung solos and chorus parts in in rock shows and classical concerts, I play keyboard in a band, and lots more. And almost all of it all takes printed music. Yes, sometimes I can get away with bar charts and chord symbols; sometimes I learn from recordings and lyrics; sometimes I enter music with a piano-roll editor. But I still use music notation more than everything else put together, and so do most of the people I make music with. It's the most accurate, portable and compatible way of communicating music, and there's still a huge demand for it. Yes, some musicians can't read music. Neither can some drummers [fx: ducks]. But Hovis aren't going out of business because some people don't eat bread, and neither are score-writing packages of various kinds.
I'm just getting started with LilyPod (I've completed one piece, and working on a couple of others), and I can tell you that it's much the best output I've seen, well worth the extra effort. Many packages will get you half-way there fairly easily, but half-way there does not look good. It's like comparing old, crude monospaced dot-matrix text with professional print -- it conveys roughly the same information, but it's much harder to follow and makes you work for it. Some packages can even get 90% of the way there (e.g. Cubase Score, Sibelius), but the last 20% of that takes forever, tweaking and adjusting and cursing, and the result is fragile. Only LilyPond can get the final 10% of the way there -- in a simple, robust way, producing output that looks like a piece of music rather than a bunch of music symbols on a page.
LilyPond isn't perfect. It's hard to learn even for an experiencer programmer like me; and it has trouble with complex vocal scores (e.g. where vocal lines combine or split, or switch between sharing staves and/or lyrics and each having their own; can anyone help with these cases?). But its output is far more, well, musical than anything else I've seen.
Use whatever notation or package is most useful to you. You don't have to use standard music notation, or LilyPond, if you don't want to. But there are plenty of us who do; isn't that what matters?
Ceterum censeo subscriptionem esse delendam.
Posting of the text of a slashdotted article anonomously is NOT trolling.
If done under your username, it is Karma Whoring, so you get mod moints.
In this case, it may be redundant, as it is seen elsewhere above. If it is posted LATER, it is redundant.
Sometimes things get posted anon and are not seen, moderators should always browse at -1.
Actually, polyphony is one of those areas where LilyPond scores (ha!) rather highly. Straight parallel-part writing is easy enough; you can either enter the two parts separately, or combine them as chords or interspersed parts. But it excels where notes overlap and combine in ad-hoc, shifting ways, such as complex piano music. LilyPond makes it easy to describe this sort of thing, to arbitrary complexity, with chords/clusters, nested parts, small- as well as large-scale parallelism, and the ability to create/use/destroy extra parts as you go along. Even in my (so far only) experiment with arranging a piano part, I found I'd naturally written stuff that even a complex GUI package just wouldn't be able to do.
Your general point is sort of valid, though. I believe there are several GUIs which can output LilyPond format, some dedicated and some more general, and it also comes with convertors from MIDI, ABC, and other popular formats. So you could easily write and/or enter music some other way, and then use LilyPond for the final engraving (which seems to be its aim). OTOH, its own text format is straightforward enough for what it does, and I personally look forward to using it some more. Whatever works for you, I guess.
Ceterum censeo subscriptionem esse delendam.
My Acrobat Reader defaults to turn antialiasing for line art off. If you turn it on the slurs are antialiased. Maybe this is why you think they look so bad. Personally, I still think they don't look that great even when antialiased.
Uh, I've played some of Brubeck's music. Um, yes I would say that he does indeed read music. Yes, lots of jazz folks start out without music, as DUH! jazz music relies on the ear and that's a good way to start. Get your story straight before making ridiculous claims man.
I think Han-Wen's real criticism of MusicXML is revealed when he says:
In other words, here is another developer who wants to lock you into his own program and format. Data can flow into Lilypond, but it should never leave. MusicXML's whole purpose is to exchange symbolic musical data between applications. So people who have a proprietary data mindset will naturally find reasons to criticize it.
People can work on an open source project and still want to lock you into their proprietary format. Plenty of open source projects use MusicXML even if LilyPond doesn't.
I don't understand how "MusicXML fails in many ways", but that is probably a question best discussed on the MusicXML discussion list rather than here.
There are plenty of hooks for doing MusicXML output from LilyPond, and there are no legal encumbrances whatsoever barring an implementation In fact, I'll gladly accept any well-coded patches for MusicXML-output. Or you can hire me, and I'll code the support for you.
However, there is no incentive for me to code this feature for free. That's why I'm not doing it.
Han-Wen Nienhuys -- LilyPond
"...the main asset of having MusicXML-output is that users can migrate away from LilyPond more easily, and that doesn't give me warm fuzzies."
Yes, that's a damning (and disappointing) quote, especially coming from an OSS developer.
Claiming that MusicXML fails on many levels was a concession to its critics, but not a hollow one. However, I'll respect your wishes and refrain from expounding on this point here. I reject the criticisms of MusicXML that condemn it simply because it's based on XML. However, there are some valid criticisms that stem from this issue, one or two of which Han-Wen articulated. Recordare itself made sacrifices in developing the format (verbalized on the website), which may or may not ultimately impair its acceptance.
To be honest, I was very excited when I first learned about MusicXML (when Sharpeye started supporting it), but my enthusiasm waned when I learned how much MusicXML's interoperability depends on propietary plugins (Dolet) and third-party converters in most cases. Hopefully this is temporary inconvenience that would be chiefly allieviated by official support in the next releases of Finale and Sibelius. OSS purists could target Recordare with comments similar to that which you made about Han-Wen: that there can exist a dichotomy of open source code and proprietary constrictions that betray a conflict of interests.
I personally think that the 1.0 DTD status should have been accompanied--if not preceded--by web browser plugins. This may be one of the most significant things that Recordare can do to bolster MusicXML's popularity and profile.
Some people who claim that they exchange scores online ask what MusicXML can do for them, since conversion apparently isn't a problem for them. I feel that being able to render music seemlessly in the browser is a big deal, and I think it would make a lot of people sit up and take notice of the format.
P.S., It would be nice if you made the mailing list archives publically accessible sans login.
Oh, you can expound here - I just thought you'd get a better discussion on the MusicXML list. I'm curious to see what you think the valid criticisms are. The interview criticisms that I saw were either 1) generic anti-XML, 2) ignorant about MusicXML, or 3) valid criticisms for using MusicXML as a native format for some applications, but not for its intended use as an interchange format. And MusicXML support has been part of Finale for Windows since Finale 2003.
And thanks very much for your earlier Slashdot story!