[Lilypond] Numerous Questions & Thoughts

Recommendations concerning notation and publishing software in a non-partisan environment.
User avatar
wess-music
Posts: 51
Joined: 05 Oct 2015, 10:40

[Lilypond] Numerous Questions & Thoughts

Post by wess-music »

@Pierre
@tisimst
@Knut
and all other colleagues who work mainly with LilyPond:

I have a very sincere question to you friends:

"Why do you prefer LilyPond since there are many other application – most of them you know for sure and even better than me, because I specialised only in Finale?"

Please, do not consider my question as a provocation.
I feel myself here surrounded with colleges and friends and not of beginners (as in Finale Forum), who desperately need some help on very base level.

Just to get the picture of myself and related interests – I still continue taking pictures on Film (mainly BW and develop manually) though I have pretty good digital camera.
In other words - I do appreciate the real things in this life that bring aesthetically approach to our souls.

Does the work with the Lilypond deliver such kind of pleasure of going through the touch of every element?
In addition — I visit sites and watched more than 20 videos and I am really very intrigued from some of the result I've seen.

And the next questions: can you compare the work on Lilypond with the same processes done in Sibelius or Finale?
Is it faster of slower? How you find practical solution looking on this limited part of the screen, where the compiled view appears?
How about your customers, when they need or request their back up files (archive) in order to fix themselves some mistakes or weaknesses, however slight, after many months or years? (since the compatibility from version to version is a serious thing).
Can you manually control the horizontal spacing of single note without touching the vertical (in dense orchestral scores usually)?
How big and sophisticated score can be done and can the user decide custom size for page format?
How useful is the program when the parts need to be extracted?
The biggest concern of mine: can you control in details (long) slurs as in Sibelius or Finale, because those, many shared sample, found in internet, are looking IMO a bit confusing if not problematic.
What I mean — the beginnings and endings of the arcs, thickness, floating angle and control points for long span and so on.
What about embedding graphics? If yes, what kind of formats accepts the program?
Dash slur/tie?

Put in a few words: what are the main limitation compared to the known software alternatives?
_______

I know, there are many questions and it could take long time to answer, but please.
Thank you.

Best regards,
Wess
Last edited by wess-music on 11 Nov 2015, 10:35, edited 2 times in total.
User avatar
OCTO
Posts: 1742
Joined: 05 Oct 2015, 06:52
Location: Sweden

Re: [Lilypond] Numerous Questions & Thoughts

Post by OCTO »

Wess-music brings quite important questions.

I think that Lilypond has a great power but "fails" in two moments:
1. Engraving (music notation, music typesetting) for many people means visual work on a plate, paper or computer "paper". While SCORE confirms the opposite (that it is possible to engrave beautiful scores without GUI), today's society is extremely visual dependent (icons, apps, constantly updated GUIs, etc...). If we see Lilypond as flow of input-compiler-output, the most of engravers want to skip the first two steps. Only output is needed.
2. Engraving is not composing, and composing is not engraving. However, many composers DO both at the same time using other tools (StaffPad as newest.) It is extremely difficult to use a textual music input for composing. You cannot identify form, phrase, chord in the textual file, as it is not possible to play piano from a SCORE's or Lilypond's textual file.

I think these are two quite difficult moments. If there is possibility to fix these two, in a very clever and appealing way, Lilypond will be more popular than now.

I believe that the power in Lilypond could be in SCORE-replacement, but in that case the time is running quick. The GUI's notation tools are developing quick and constantly implementing new things. The differences are tight.
The Lilypond developers should contact the main publishers and offer their expertise in notation. Than they should compare the final output of Lilypond and other tools showing its superiority and so awake interest for publishers. As with Linux servers, you offer it for free, but you get paid for fixing it and running.

Maybe I might have it wrong, but these are my thoughts right now.
Freelance Composer. Self-Publisher.
Finale 27.3 • Sibelius 2023.5• MuseScore 4+ • Logic Pro X+ • Ableton Live 11+ • Digital Performer 10+ /// MacOS Monterey (secondary in use systems: Fedora 35, Windows 10)
User avatar
tisimst
Posts: 416
Joined: 08 Oct 2015, 17:57
Location: UT, USA
Contact:

Re: [Lilypond] Numerous Questions & Thoughts

Post by tisimst »

wess-music wrote: @Pierre
@tisimst
@Knut
and all other colleagues who work mainly with LilyPond:

I have a very sincere question to you friends:

"Why do you prefer LilyPond since there are many other application – most of them you know for sure and even better than me, because I specialised only in Finale?"
...
Does the work with the Lilypond deliver such kind of pleasure of going through the touch of every element?
These are great questions. Let me share a couple of my thoughts about my experience with this program. I have written numerous posts in a variety of places as to why I prefer LilyPond over other programs I've used (Sibelius, MuseScore, etc.). Here are a few reasons:

1. Default Output - Almost in every case, the default appearance is so close to what I want that I hardly need to hand-tweak anything. If I do want to tweak something (like the shape of a slur, the horizontal offset of a notehead, the thickness of a line), there is usually a simple "override" to do that. In other words, LilyPond gets me to a much better starting point for editing than any other program I've used. And now that LilyPond supports alternate music fonts, the appearance is infinitely customizable.

2. File Format - A pure-text file format offers quite a few benefits that a proprietary binary format does not. I can read it on any computer, in the text editor of my choice. I can track the changes much easier with a source code management tool like GitHub, and thus can always revert to a previous version at any time. The format itself is much more bullet-proof (in terms of readability and future use) than a binary format and is less susceptible to problems when converting to a newer version. There are lots of editors that support many conveniences for quickly inputting/editing the data (like syntax highlighting, command completion, etc.).

3. Verbosity & Extensibility - In LilyPond syntax, you know exactly what is being attached to what (like the ends of a slur, or articulations, or dynamics) and is very human-readable most of the time. If I have transposed some music, it's obvious where that takes place. If there's a certain functionality that you want that hasn't been developed yet (especially a programmatic one), then you can use the powerful Scheme language to do that. I had to do that recently with an experimental music font so that I could rotate/mirror noteheads on-the-fly. This would have been totally impossible to do *within* any other notation program and painfully time-consuming in a graphics program.

4. Separation of Score Content & Structure - In LilyPond, I can intentionally separate the music content from the the score structure. What good does this do? It gives me so much more flexibility to use and reuse the music. I can store pretty much anything in a variable for later reuse. For example, I can enter the music for a horn part once and utilize the same content in both the conductor's full score and the individual part's score. At the same time, I can customize the layout and appearance of each of those scores completely independent of each other.

I'll stop there for now. Do I enjoy inputting music this way? I actually do because it has helped me to be more deliberate of my choices and aware of what I want to engrave and how it all works together on the page. Do I miss the instant aural feedback while entering notes? Sometimes, but I can get around that by using LilyPond's MIDI output to do audio-proofing when necessary. I find that I can input a full score just as fast if not faster than with other programs because I can input other notation elements at the same time as inputting notes without much slow-down, if any.
wess-music wrote: And the next questions: can you compare the work on Lilypond with the same processes done in Sibelius or Finale?
Is it faster of slower? How you find practical solution looking on this limited part of the screen, where the compiled view appears?
In terms of work-flow, like I said before, there's no question for me that it is just as fast if not faster to input music as text than in a graphical environment. If you've seen the tutorials by Ben Lemon, then you'll have been introduced to Frescobaldi (http://frescobaldi.org/). Each user has their own favorite program, but I have found Frescobaldi to be a total life-saver when working on LilyPond files. One HUGE benefit of this program is the point-and-click functionality (i.e., I can click on a note, articulation, slur, etc. on the engraved score and Frescobaldi automatically shoots the cursor to that location in the input file). This is such a time saver and works very well when I need to make changes.
wess-music wrote: How about your customers, when they need or request their back up files (archive) in order to fix themselves some mistakes or weaknesses, however slight, after many months or years? (since the compatibility from version to version is a serious thing).
Thankfully, there are a few things going for you (which I've already eluded to). As a human-readable text file, the customer is able to see exactly what has been done. If the syntax has changed somehow from a previous version, the developers have provided a utility for updating that automatically for the most part. Usually, the syntax doesn't change all that much from version to version, so recent versions of input files will work with little or no changes. In other words, even if you see, for example, that a file was coded for version 2.16.2, more recent versions of LilyPond (like 2.18.2) will likely compile it without problems and without you needing to change anything--and with better results!

There's also been some work to make LilyPond more friendly to the MusicXML and MEI formats, but these are still in development. Importing MusicXML is probably the most supported feature at the moment, though everyone will agree that there's still work to be done there.
wess-music wrote: Can you manually control the horizontal spacing of single note without touching the vertical (in dense orchestral scores usually)?
Absolutely yes, this can be usually done with a simple tweak. This applies to virtually everything you see on the page.
wess-music wrote: How big and sophisticated score can be done and can the user decide custom size for page format?
This is another beautiful strength of LilyPond. It has almost no limits to how large/sophisticated a score can be. Add as many staves as you want. Add as many pages as you want. Add as many voices to a staff as you want. Change it to any page format you want. LilyPond can handle it. You may want to tweak some settings to taste, but that's the choice of the user.
wess-music wrote: How useful is the program when the parts need to be extracted?
VERY useful, although LilyPond kind of works in reverse to other programs. Instead of "extracting" the music of an individual part out of the full score, the parts are created first and then added to the full score. This is a really, really nice way to handle the music, as I've mentioned already.
wess-music wrote: The biggest concern of mine: can you control in details (long) slurs as in Sibelius or Finale, because those, seen on many sample in the NET are looking IMO a bit confusing if not problematic.
What I mean — the beginnings and endings of the arcs, thickness, floating angle and control points for long span and so on.
Definitely. LilyPond does its best to optimize the shape of ties, slurs, and phrasing slurs, but there are plenty of times (as I'm sure you've seen) that you'll want to adjust their appearance. There are a handful of properties and commands for adjusting slurs to look just how you want (e.g., thickness, control points, eccentricity, etc.), even for items that continue over multiple systems.

The only thing I can think of (related to ties/slurs) that LilyPond doesn't support out-of-the-box is a slur that has a truly flat part. I have no doubt, however, that some of the developers could implement this feature without too much trouble.
wess-music wrote: What about embedding graphics? If yes, what kind of formats accepts the program?
Yes. The main formats are: post-script (PS), encapsulated post-script (EPS). Any other graphics can usually be converted to these without much trouble. SVG graphics can be embedded, but are a more tricky.
wess-music wrote: Dash slur/tie?
Not a problem.
wess-music wrote: Put in a few words: what are the main limitation compared to the known software alternatives?
Not really any. It's mostly just getting used to a different way of inputting music.
wess-music wrote: I know, there are many questions and it could take long time to answer, but please.
I hope I've answered your questions sufficiently. To be honest, it took me a couple of tries over the years to really understand and appreciate LilyPond. I went back and forth with other programs because I struggled with the switch to text input. Once I forced myself to work through a couple of involved scores, I began to understand things much better and realize the benefits. Now, I can confidently say that I will likely never go back to a GUI-based input program. It doesn't mean that LilyPond is the perfect solution for everyone. There are definitely use-cases where I'd say using LilyPond isn't the right tool for the job. When it comes to publication-grade engraving, however, I wouldn't use any other program. With support for alternate music fonts, it is also excellent at supporting a custom house-style.

Lastly, LilyPond and the supporting programs are being actively developed and improved (and even though the development version is called "unstable", it is rarely truly unstable and I prefer to use it with the latest feature set). The developers are very responsive to users' questions and feature requests. There are so many people on the user-mailing list that want to help when someone gets stuck. Response time is usually within minutes, from my experience.

Also, the LilyPond documentation is deep and well covered, but can be daunting, so definitely start with the Learning Manual (found on http://lilypond.org/manuals.html) and then move on to the Notation Manual after that. These two will cover most syntax questions you will have. More advanced code can be found in the Snippets manual and in the LilyPond Snippets Repository (or LSR, found at http://lsr.di.unimi.it/). I've found these immensely helpful when I've needed to do something out of the ordinary.

So, take that for what its worth. I understand that it's a lot to chew on (but you asked for it ;) ) If you want to try LilyPond out without installing anything on your computer, you can always try an online interface called LilyBin (http://lilybin.com/). I've used this many times, but I mostly stick to working with Frescobaldi on my own machine.
Music Typeface Designer & Engraver - LilyPond | Sibelius | Finale | MuseScore | Dorico | SMuFL | Inkscape | FontForge
User avatar
tisimst
Posts: 416
Joined: 08 Oct 2015, 17:57
Location: UT, USA
Contact:

Re: [Lilypond] Numerous Questions & Thoughts

Post by tisimst »

OCTO wrote:Wess-music brings quite important questions.

I think that Lilypond has a great power but "fails" in two moments:
1. Engraving (music notation, music typesetting) for many people means visual work on a plate, paper or computer "paper". While SCORE confirms the opposite (that it is possible to engrave beautiful scores without GUI), today's society is extremely visual dependent (icons, apps, constantly updated GUIs, etc...). If we see Lilypond as flow of input-compiler-output, the most of engravers want to skip the first two steps. Only output is needed.
2. Engraving is not composing, and composing is not engraving. However, many composers DO both at the same time using other tools (StaffPad as newest.) It is extremely difficult to use a textual music input for composing. You cannot identify form, phrase, chord in the textual file, as it is not possible to play piano from a SCORE's or Lilypond's textual file.

I think these are two quite difficult moments. If there is possibility to fix these two, in a very clever and appealing way, Lilypond will be more popular than now.
I think you hit it right on the head, so LilyPond tends to appeal to musicians who also have programming experience and are used to a text-input environment. I am sure glad that I got over that hurdle, though (but then again, I don't do that much composing).
OCTO wrote: The Lilypond developers should contact the main publishers and offer their expertise in notation. Than they should compare the final output of Lilypond and other tools showing its superiority and so awake interest for publishers. As with Linux servers, you offer it for free, but you get paid for fixing it and running.
I wish it was that easy. These kinds of activities have been done/proposed on numerous occasions, but with little success. There are two major factors at play:

1. Work-flow - Established publishers have developed work-flows that work for them and they are hesitant to disrupt that. Some publishers just don't have any interest, regardless of the short-term and long-term benefits of using LilyPond and its text files. The funny thing about it is that both Amadeus and SCORE used to be both text-input (and much less user-friendly IMHO) and these programs are revered for their superior output quality.

2. Industry Momentum - With so many engravers/copyists/publishers/composers/etc. using Sibelius and Finale (and LilyPond's weak support for MusicXML conversion), it's just easier to continue using those programs because it's what their clients use.
Music Typeface Designer & Engraver - LilyPond | Sibelius | Finale | MuseScore | Dorico | SMuFL | Inkscape | FontForge
User avatar
OCTO
Posts: 1742
Joined: 05 Oct 2015, 06:52
Location: Sweden

Re: [Lilypond] Numerous Questions & Thoughts

Post by OCTO »

tisimst, thanks for so nice explanation.

As said earlier, I love {open source, GPL, free software} philosophy. I am a composer of instrumental music who needs well designed scores...
Being so, I don't have time to learn Lilypond by myself since I am afraid of loosing control over deadlines and so on, because I have developed a network of engravers/friends who work in Finale.

Sometimes when having deadlines, I need my scores to be done in time, and correctly. There are plenty of people that can do it in Finale.
Switching to LilyPond could harm my artistic efforts, if I loose support by people who could help me engraving.

Maybe I should test it in vitro, just because it would be worth to do it.
tisimst wrote:I wish it was that easy. These kinds of activities have been done/proposed on numerous occasions, but with little success.
I have in front of me some new scores of B&H, Peters and Schott. It seems that they don't care any more about the quality.
I believe it is because:
  • composers do a lot of job by providing already done scores (badly engraved)
  • they don't care about the quality since it costs a lot of money, and they will do it properly once a work is established on the market (after many years/decades).
Last edited by OCTO on 11 Nov 2015, 13:47, edited 4 times in total.
Freelance Composer. Self-Publisher.
Finale 27.3 • Sibelius 2023.5• MuseScore 4+ • Logic Pro X+ • Ableton Live 11+ • Digital Performer 10+ /// MacOS Monterey (secondary in use systems: Fedora 35, Windows 10)
User avatar
wess-music
Posts: 51
Joined: 05 Oct 2015, 10:40

Re: [Lilypond] Numerous Questions & Thoughts

Post by wess-music »

@tisimst

I must say only — great thanks, thank you very much indeed, because such inspired answer I've newer expected.

Best regards,
Wess
User avatar
Schneider
Posts: 111
Joined: 09 Oct 2015, 06:50
Location: Paris

Re: [Lilypond] Numerous Questions & Thoughts

Post by Schneider »

With no surprise, I agree 100% with what Tisimst wrote. ;)
I also confirm that, from discussions I had last year with publishers at the http://www.musicora.com/en, most (all?) of them uses Finale and Sibelius and have no interest to use LilyPond.
And, yes, LilyPond's weakest point is definitely the MusicXML conversion.
User avatar
OCTO
Posts: 1742
Joined: 05 Oct 2015, 06:52
Location: Sweden

Re: [Lilypond] Numerous Questions & Thoughts

Post by OCTO »

hmmm LilyPond users.... Why not to start a new publishing business, only based on the beauty of engraving and on LilyPond? Für Elise - so well done that it beats all previous editions? (I am serious.)
Freelance Composer. Self-Publisher.
Finale 27.3 • Sibelius 2023.5• MuseScore 4+ • Logic Pro X+ • Ableton Live 11+ • Digital Performer 10+ /// MacOS Monterey (secondary in use systems: Fedora 35, Windows 10)
User avatar
tisimst
Posts: 416
Joined: 08 Oct 2015, 17:57
Location: UT, USA
Contact:

Re: [Lilypond] Numerous Questions & Thoughts

Post by tisimst »

I love this idea! Any recommendations of people who have been in the publishing industry who'd be willing to share some thoughts about getting things off the ground or making the best use of digitized works (i.e., a PDF download)?
Music Typeface Designer & Engraver - LilyPond | Sibelius | Finale | MuseScore | Dorico | SMuFL | Inkscape | FontForge
User avatar
OCTO
Posts: 1742
Joined: 05 Oct 2015, 06:52
Location: Sweden

Re: [Lilypond] Numerous Questions & Thoughts

Post by OCTO »

I have my own publishing company, you can drop me questions if you like.
PDF is OK, but I recommend/prefer going forward with analog paper sheets.
Freelance Composer. Self-Publisher.
Finale 27.3 • Sibelius 2023.5• MuseScore 4+ • Logic Pro X+ • Ableton Live 11+ • Digital Performer 10+ /// MacOS Monterey (secondary in use systems: Fedora 35, Windows 10)
Post Reply