Notation Programs' Output Comparison

Have your scores reviewed by other users. Comment on old and new published scores and on publishers.
Post Reply
jrethorst
Posts: 122
Joined: 09 Apr 2016, 18:48

Notation Programs' Output Comparison

Post by jrethorst »

Some years ago I created a PDF document using Encore, Finale, LilyPond, Overture, Score and Sibelius, to compare their output. Several people, accomplished users of these programs, contributed scores; I did the Overture sample.

Keeping in mind that reasons for choosing a program are largely of user interface and feature set, it's also worth comparing the look or "personality" of the finished score. As I told Abraham Lee, a notable music font designer who was interested in the project:

> The original point of the doc was to show what each program, using more or less its
> default settings, looks like. Although most of the programs are flexible enough that
> their output can be made to look like the other programs', some parts, e.g. music
> font and slur curves, may be less flexible. It's difficult to call it a rigid comparison
> since anyone's notion of defaults can differ. All of the entries were done by people
> pretty well versed in their respective programs, but it's interesting how the
> personality of the pages varies.

He replied:

> I couldn't agree more! It really is amazing to see how each application embeds its
> own personality into the score's appearance, beyond what is "correct" or not.
> Truly fascinating, indeed!

This eight-page PDF file, "Six Music Notation Programs", with the six versions of a one-page arrangement for solo classical guitar, is at:

https://dl.dropboxusercontent.com/u/643 ... %20v.2.pdf

and it has received some interesting attention and discussion on various websites.

It's ready for an update, first to include the free program MuseScore, and secondly to include output from current versions of the other programs. Although the present document is privileged to have LilyPond done by Han-Wen Nienhuys, one of the program's creators, and Sibelius done by Daniel Spreadbury, formerly that program's lead developer, they were not done using current versions.

The arrangement in MIDI is at:

https://dl.dropboxusercontent.com/u/643 ... ure%29.MID

I have several contributions for the update, but need scores done in Encore and Sibelius. I don't own the current versions of these programs. If anyone who does would like to do an update, I'd appreciate it. You'd receive credit in the introduction: if you do engraving as a business this could be a helpful reference for you. And it's an interesting project.

Thanks,
John
John Rethorst
User avatar
OCTO
Posts: 1738
Joined: 05 Oct 2015, 06:52
Location: Sweden

Re: Notation Programs' Output Comparison

Post by OCTO »

jrethorst wrote:This eight-page PDF file, "Six Music Notation Programs", with the six versions of a one-page arrangement for solo classical guitar,
Absolutely amazing!
And welcome to the forum.

The PDF file looks like it is scanned, but I like that, somehow lines get smoother.

I know which output I don't like, but I will not discuss it.
I have never looked closely at Encore and Overture.
SCORE doesn't look like having the original font.
In this version Finale does it very well IMO.

The only thing I would change in the (future) comparison is to add some other values, such as whole notes, flags, rests, time signatures, accidentals and different values for spacing comparison. And also more dynamics such as sfp, mp, rinfz. etc.
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)
Knut
Posts: 867
Joined: 05 Oct 2015, 18:07
Location: Oslo, Norway

Re: Notation Programs' Output Comparison

Post by Knut »

Interesting! I agree with OCTO in terms of increasing the complexity to showcase the programs more thoroughly.

In terms of output quality, I'm pretty much of the opposite opinion, however. To my eyes, Finale comes across very badly in this comparison, even the worst of them all in some respects. It's no secret that Finale's default settings are pretty bad, and I also think it's really clear from this comparison (provided it's actually objectively executed, of course) just how much manual editing is needed in Finale to obtain a decent result.

That said, I have issues with all the output in this document. And even though SCORE's output, perhaps unsurprisingly, comes across as the superior one, comparing the default output from different scoring applications like this doesn't necessarily say much about which application is the most powerful one.
User avatar
John Ruggero
Posts: 2449
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Notation Programs' Output Comparison

Post by John Ruggero »

This is certainly a worth-while endeavor, and I am amazed that no one has done it previously. I would make the following suggestions. (But please consider that I am strictly a Finale user with only slight experience with Sibelius.)

1. The PDFs are not comparable, which may be seen by enlarging them from 200-1200%. The PDF of the Finale is the fuzziest and the LilyPond and Sibelius are also much less clear than the others.

2. Use a piece of music that has more variety of symbols. This piece has one dynamic mark, no articulation marks or slurs. A very short piano solo would probably be best, because one could see braces, F clefs, symbol positioning decisions. Here are some possibilities:

Beethoven Bagatelle op. 119 no. 11
Chopin Prelude op 28 no. 18
Schumann Kinderscenen op. 15 no. 7 or 13
Brahms Waltz op. 39 no. 15
Grieg Lyric Piece op. 12 no. 1
etc.

3. Allow the engravers to use their personal settings so that the programs are seen at their best, not their worst. The Finale default settings are notoriously bad (first and second endings, and the position of some of the ties, for example.) and that puts this program at a disadvantage.

4. Supply the original from which the engravers worked. It is not clear why there are differences between the different versions. Why do the ending chords of the LilyPond differ from the other versions? Why is there a precautionary accidental missing from the A in many of the examples in measure 3? Why does the wording and symbols vary for the Da Capo and Coda? Were the engravers allowed to change anything they wished?

Impressions:

All the versions have strengths and weaknesses which often reflect the engraver more than the program used. Things that stand out to me:

LilyPond: the page looks uncrowded and very readable. This required the used of 7 instead of 8 staves on the page, which had nice consequences for the layout. All the other versions have too much white space on the page. The constant use of 4 measure on a line works against readability.

LilyPond, the "6-str guitar" is redundant since the instrumentation is clearly stated above. Nice 1st and second ending numbers and brackets, but second set are positioned too high. Sibelius and the other versions have better positioning of the brackets.

Finale. Aside from the excessive white space, there is nice general layout throughout, except for the copyright notice that too close to the bottom staff.
Sibelius, similarly and with less white space, but now the top of the page is crowded.

Sibelius: Best layout of the Da Capo and Coda by placing a white space between the piece and the coda. However, the small coda sign plus the word Coda looks strange to me. Runner-up: LilyPond for at least putting a space on the staff between the body of the piece and the coda, but where is the "then to the Coda"? Finale could have added the white space but would have needed a work-around.

Most consistent arpeggio signs: Sibelius.
Best G clefs: Overture, runner-up: Encore
Best ties: Finale, but the default runs them into the dot in a few cases.
Best black note heads: Finale, Score and LilyPond
Positioning of the rests: all do better than Finale, which has some badly positioned rests.
Best overall font: Finale Maestro
Warmest over-all appearance: Sibelius and LilyPond, then Finale, then Score
Last edited by John Ruggero on 11 Apr 2016, 12:12, edited 1 time in total.
M1 Mac mini (OS 12.4), Dorico, Finale 25.5, GPO 4, Affinity Publisher 2, SmartScore 64 Pro, JW Plug-ins, TG Tools, Keyboard maestro

http://www.cantilenapress.com
User avatar
OCTO
Posts: 1738
Joined: 05 Oct 2015, 06:52
Location: Sweden

Re: Notation Programs' Output Comparison

Post by OCTO »

I would agree with John R; there should be a template (InDesign, for instance), with all information around the music, so that it is the same.
Also, use exact Times (for instance) as the main font in music with exact same size would be an additional help.
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:

Notation Programs' Output Comparison

Post by tisimst »

Sorry, but I think we've missed the point of this little exercise. This isn't necessarily a "challenge" between programs to see who is "better" or the most "correct" since all programs can output the same level of "correct", but rather an exposé to show how each program presents the same musical content and a chance to give people the opportunity to enjoy the different presentations. There is no question that the ground rules are fuzzy, but that's ok. I've encouraged John to seek out competent engravers for each program in this "comparison" because, let's be honest, there are quite a few issues with the 2nd edition engravings, as has already been pointed out. I've provided an updated LilyPond sample* and I hope others will take their competency and do similar with their own favorite app.

Personally, I love seeing how each program handles music. Is there a score that showcases more things for comparison? Sure, but let's not get hung up on that. If we want to put together another, more extravagant "comparison" document using a different score or set of scores, let's do it! And let's use the latest and greatest apps by experienced individuals so the results look their best. I think it's quite fun!
Music Typeface Designer & Engraver - LilyPond | Sibelius | Finale | MuseScore | Dorico | SMuFL | Inkscape | FontForge
User avatar
John Ruggero
Posts: 2449
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Notation Programs' Output Comparison

Post by John Ruggero »

I totally agree with tismst except about "missing the point." It is human nature to express preferences when comparing things and that will never change.

Maybe there should be 2 or 3 examples of the same piece and program by different engravers to show the scope of each program.
M1 Mac mini (OS 12.4), Dorico, Finale 25.5, GPO 4, Affinity Publisher 2, SmartScore 64 Pro, JW Plug-ins, TG Tools, Keyboard maestro

http://www.cantilenapress.com
User avatar
tisimst
Posts: 416
Joined: 08 Oct 2015, 17:57
Location: UT, USA
Contact:

Re: Notation Programs' Output Comparison

Post by tisimst »

Thanks, John. I could have been more clear. We are definitely encouraged to express our opinions of the different engravings. What I meant to say was that the comparison was not designed to be a rigorous challenge of, for example, how much time it took to create them, amount of tweaks needed to get to the final version, app capabilities, etc. like other comparisons have been. That's all.

No, I loved reading your assessment, John.
Music Typeface Designer & Engraver - LilyPond | Sibelius | Finale | MuseScore | Dorico | SMuFL | Inkscape | FontForge
User avatar
John Ruggero
Posts: 2449
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Notation Programs' Output Comparison

Post by John Ruggero »

As suggested by tisimst, I think that it would be a great project for Notat.io to take on. A few ideas occurred to me to start a discussion of such a project:

1.The chosen piece might:

a. be a piano solo to show two clefs, braces etc.
b. contain a great variety of symbols and situations
c. fill a single page
d. be well-known, if possible, and, of course, in the public domain
e. be a piece in which the composer's MS, first editions and other primary and secondary materials are readily available. These might be presented on the Notat.io site in conjunction with the entries.

2. All entries would be engraved by means of any computer program or programs.
3. All entries could be available for observation and comment at the Notat.io website.
4. There could be no time constraints on submissions; the process could proceed as long as the Notat.io site exists.
5. One entry per engraver, but entries might be edited at any time.
6. The editorial philosophy would be left to the engraver and range from authentic "urtext" editions to highly personal ones.

I think that such a project might create a lot of interest and bring many new engravers to this site.
Last edited by John Ruggero on 11 Apr 2016, 22:52, edited 1 time in total.
M1 Mac mini (OS 12.4), Dorico, Finale 25.5, GPO 4, Affinity Publisher 2, SmartScore 64 Pro, JW Plug-ins, TG Tools, Keyboard maestro

http://www.cantilenapress.com
User avatar
John Ruggero
Posts: 2449
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Notation Programs' Output Comparison

Post by John Ruggero »

Thanks, tisimst. I read your post after my last one. I understand your point now and think that we are definitely on the same page concerning such a project, which would be best handled in a non-competitive environment, just like this website.
Last edited by John Ruggero on 11 Apr 2016, 22:52, edited 1 time in total.
M1 Mac mini (OS 12.4), Dorico, Finale 25.5, GPO 4, Affinity Publisher 2, SmartScore 64 Pro, JW Plug-ins, TG Tools, Keyboard maestro

http://www.cantilenapress.com
Post Reply