Guitar finger picking pattern arpeggios with held notes

• May 9, 2021 - 03:49

A guitar player often plays arpeggios with the right hand while holding a fixed chord with the left hand. This is extremely common in guitar playing, often sloppy written down using TAB notation without any proper note lengths. However, any guitar player knows how to play these: Any plucked note is supposed to ring until the same string is plucked again (or the chord changes, but sometimes even beyond). Many songs are played this way, sometimes using fixed chords, often varying then during picking. A very famous example would be the intro to Stairway To Heaven.

Musescore doesn't allow entering these so that they are easily editable, look correct and are played correctly. I appended a small example score demonstrating the problem. It contains four instruments all playing exactly the same, only the display variant is different (standard notation, simple TAB, common TAB, and full TAB).

The easiest way of entering this in Musescore is to simply denote the pattern in the TAB score, but then everything is played in one voice which makes any note stop as soon as another one starts, which is not what is supposed to be played. This I did in measure one. It looks okay in any display style, but of course because all notes are eighths, it doesn't sound right. It sound as if after each plucked string, that string is muted again before playing the next one. Polyphony only exists if several strings are also plucked at the same time.

So I figured that I could use voices to make the notes ring longer while others are played. This has been done in measure 2, and it has been done again in the measures 3 to 5 with notes crossing the bars. These sound (about) right (except for point 3, see below).

This has three problems:

  1. Entering this isn't straight forward anymore. First you have to enter the score as before, then assign to each note on any string a different voice (e.g. notes on the G string are all voice 3), then you have to make each note longer, so that it rings until this string is plucked again (at least that's the best way I came up with). Entering such a thing once is manageable. To play around with different patterns (i.e. editing such a thing) is a PITA.

  2. It looks horrible. Look at the screenshot. Only the simple TAB is readable anymore. The common TAB can still be used, but it you want to let notes ring beyond the measure borders, you get lots of ties (I only have 2 but there can be much more of course).

  3. You only have four voices. Guitars have six strings (some even more). So this doesn't cover all strings. In my example, I share the voice 2 between the B and the D string. Of course, notes played on the B string are now aborted as soon as a note on the D string is played, etc.

Is there any better way to enter these absolutely standard things of guitar playing in a way which doesn't only look good but which eventually will also play the right things?

If not:

I would like to propose some new feature which allows guitar players to enter guitar tabs simply and which doesn't sound wrong in the end.

One way of doing this could be
* increasing the voice count to (at least) six,
* adding a function which assigns each note to a voice corresponding to the string it is played on, and
* adding a function which simply prolongs every note in a TAB to fill the time until the same string is plucked again.

Then we could use the simple TAB notation which hides all the (not helpful) uglinesses like ties and note lengths.

Any other new feature solving this problem would be welcome as well :-)

Alfe

Attachment Size
Guitar_Arpeggio_Example.mscz 17.31 KB
Guitar_Arpeggio_Example.png 85.35 KB

Comments

Many of us would like to see an option for automatic Guitar durations. This is needed whether scorists write in a single voice and write in three or four voices, which is likely the limit found in scores produced by major publishing houses.

The points you raise are discussed here:

https://musescore.org/en/node/319340

https://musescore.org/en/node/319324

https://musescore.org/en/node/310986

It has probably been discussed elsewhere as well, but I don't recall if an official request was ever made in Musescore's Issue Tracker.

In reply to by yonah_ag

I was referring specifically to the first measure of the attached file. Is that, in this case, the "Let ring" line does the job. Answer: yes, absolutely.
As for the rest, you already know my opinion :)
It's a waste of time to want and believe to simulate the resonance of a guitar string on a computer. For you, play the score yourself on your guitar, and the problem does not exist.
And for those who look at your scores on musescore.com, they don't come to judge a sound rendering that would be absolutely perfect (and it can't), they come to download a tablature and play it, that's all. The playback is only a temporary help, a starting point.
Then, they're on their own with it.

In reply to by cadiz1

For me it's by no means a waste of time: for you it is.

Other TAB softwares provide a better simulation of guitar playback than MuseScore does and so their programmers clearly think that it is worthwhile. I still prefer MS because of the flexibility of its engraving.

At least 1 musicologist musescorer, (BSG), regards playback important enough to him that he wrote an articulation plugin.

Whilst "Let Ring" may help in this specific instance, the OP has provided an excellent summary of the playback vs. scoring issues that fingerstyle guitarist face. His summary would provide a good scope document for improving guitar playback if it becomes a priority.

In reply to by scorster

One of your links also led to a post mentioning the Piano Roll Editor. I hadn't tried that one out yet, did it now. Actually, it can be used to make any note ring for a specific time longer than denoted, so that kind of works around my problem as well, and this time even without the limit of only 4 voices and without having to assign voices at all.

It is a little tedious to use, that editor, but so far I think that's actually a better workaround than what I found.

In reply to by Alexander Fetke

"A little tedious" is an understatement.

I would still recommend 2-voice, one for the bass notes and 1 for the melody/picking. Sometimes I use voice 3 for picking since I want to apply a reduced volume to the whole voice. With 2 voice scoring you get the voice 1 rhythm above the score and voice 2 below.

In reply to by yonah_ag

AH, now I got what you meant with 2 voices: bass and the rest.

No, that's not my understanding of how the typical harmonies work. It might sound better at least to the untrained ear if at least the bass note is continued while the rest is still crippled. But the high e ringing in the Under The Bridge example is a clear demonstration of that this wouldn't be enough.

+1 (or if I "Let Ring" this thumbs-up then +10)

I've resorted to 2 voices for sane beaming and a plugin to apply duration extension to make selected notes ring. 3 voices are better for sound playback but poor for rhythm information.

In reply to by cadiz1

Guitar Pro has an excellent sound engine with proprietary soundfonts and very good playback. One particular strength is that it allows stems in different voices to be combined which prevents 3 voice rhythm clutter. I'll post one when I'm on my computer. It also supports an automated Let Ring which can be set for the whole score.

TablEdit has a function to double the written playback duration of notes individually, which allows a manual let ring to be applied quite quickly. (I have a plugin which also does this in MS but would be better if the ring could be 4x or 8x. Unfortunately the plugin API imposes a 2x limit).

In reply to by yonah_ag

"One particular strength is that it allows stems in different voices to be combined which prevents 3 voice rhythm clutter. I'll post one when I'm on my computer. "
Yes, please.
"It also supports an automated Let Ring which can be set for the whole score."
Some difference with MuseScore by doing: select all (Ctrl +A) -> click "Let ring" in Lines palette ?

In reply to by cadiz1

But MS "Let Ring" is too broken to be useful. It stops any note ringing at the end of a measure. It also plays thru a subsequent note on the same string within a measure.

Here's an example from Guitar Pro 6. There are 2 voices present, (1) melody and (2) bass, and I have chosen to show both of the rhythms both below the stave. (Any of the 4-voices per stave can be set to above/below or hidden).

The screen snips show that the first note is voice 2 is a minim and followed by a quaver in voice 1. However GP has combined the stems for visual simplicity whilst preserving the note durations. Neat. This becomes even more useful for 3-voice fingerpicking patterns.

Voice2.png

Voice1.png

Choosing rhythm position in GP6:

rhythm.png

In reply to by yonah_ag

I think this feature would be only a very small step in the right direction (and actually one I wouldn't bother using, I would then stay at my current workarounds).

If I pluck the bass note of an E major chord, that rings far longer than just for the duration of two eighths on a real guitar, and that also is relevant for the sound. If the computer mutes it after two eighths instead of after one eighth this isn't much better.

In reply to by Alexander Fetke

But if it allowed up to 16x written duration? Really easy via a plugin.

I have complained about the very things that you are talking about but it's not a priority for development so we have to come up with workarounds. All the things that you want can be achieved in 2-voice scoring and the built-in Piano Roll Editor, (or in my case a plugin).

In reply to by yonah_ag

Any fixed ring duration (twice, 16-fold, whatever) would not be appropriate. I could have a 128th note to start with. The duration just should not be depending on the length of the original note but on the position of a mute point.

If I use the PRE, I don't need more than one voice. I guess that will be my workaround for the future.

If I use two voices without the PRE tweaking, I wouldn't know how to produce three or more ringing notes which started at different times. Please enlighten me ;-)

In reply to by Alexander Fetke

Agreed. I'm not talking about a fixed duration at all but simply increasing the maximum allowed in PRE. Currently this is limited to Len = 2000 mille, (= 200%) as per image below. There are times when a shorter duration can be used, e.g. pizzicato.

Using the arpeggio symbol will start notes at different times but you can also do this is PRE with OnTime, (see image)

Mille2000.png

In reply to by cadiz1

The "let ring" feature also lets ring all the notes played consecutively on the same string, which renders it unusable in many cases. My simplistic example above doesn't show it, though. But imagine a short hammer-on and pull of for phrasing. With "let ring" all the notes create quite a dissonance bordering on cacophony.

@Alexander Fetke

"adding a function which simply prolongs every note in a TAB to fill the time until the same string is plucked again."

This would be ideal and would make working in 2 voices practical. One slight mod though: the note should also stop after its natural sustain, or if a rest/mute is applied.

In reply to by yonah_ag

Why do you think two voices would be enough? What about three strings plucked one after another? And since we have six strings, I still think that we would need six voices. (More for guitars with even more strings.)

The note should not stop after the natural sustain but the sound. (Slight difference.) This would be accomplished automatically by using samples which have a limited duration. For acoustic guitar sounds I assume this is the case anyway.

Yes, reacting on a rest or mute might be necessary. In a TAB I would notate an X on the string in the proper position then. Actually, I have never come across such a thing yet in thirty years of guitar playing, so I could live without it for now ;-)

In reply to by Jojo-Schmitz

Name it as you like (one voice or more, arpeggio in once voice, whatever). Coming from the piano originally I found the concept of what MS calls "voices" rather surprising anyway.

What I want is that several sound producers (strings in my case) of one instrument (guitar) can produce their sounds independently (so that playing one doesn't forcibly stop the other), and that's typically done using voices in MS, and I need (at least) six of them because I can pluck complex rhythmic patterns with all six strings involved, including hammer-ons and pull-offs, so it is not just an arpeggio of a chord. And that's not a theoretical case. Any decent rock song includes exactly this. Yeah, really not just some freak cases. Literally the first thing I wanted to notate using MS in TAB notation was not possible (sounded plain wrong if notated as is usual for guitar).

Other options exist of course. I can create six instruments, each being just one string. But that looks horrible in the notation then.

In reply to by Alexander Fetke

Guitar Pro also calls these voices and allows 4, but its automated ring means that 2 voices are usually enough. For some classical pieces 3 voices can be useful and I guess that very occasionally 4 could be.

The independence of rhythm is not as issue when notes ring. It would only be an issue when notes don't ring. If you have an example of a couple of measures that would still need 6 voices if there was an auto-ring then please share a picture.

Truly independent 4 voice guitar playing requires a high level of ability. I'm not sure that Koyunbaba is even in 4 voices and it's a piece for advanced players. You can follow each voice as a thread in the piece and each has independent rhythm and volume.

See https://youtu.be/szY7jmWHXJc at around 5 mins in, then 6:30, then 8:12

Even this piece can be scored in 2 voices:

https://musescore.com/user/3990006/scores/1399741

In reply to by Jojo-Schmitz

I added a real-world example (the intro to Red Hot Chili Pepper's song Under The Bridge). I notated it once in standard notation (without any ringing notes, so it sounds rather staccato) and once in TAB notation. In the TAB notation I tweaked the note lengths using the Piano Roll Editor to ring up to the intended positions. I found no proper way to add an × in the TAB notation to mark where the strings are supposed to be muted; I used a breath mark instead. But the × would be better of course as it can be placed on the specific string.

I'm not sure if you would call this kind of picking an arpeggio with phrasings or whatever (note the hammer-ons and the pull-offs in measures 3, 5, and 7). I also have no clue how many "voices" this should be called. (To me it is just one guitar and voices just a workaround to get the ringing notes.)

But I think this simple example shows clearly that using the standard notation to write down things like these just isn't appropriate and probably shouldn't even be considered for solving this issue. For instruments with several oscillators like strings ringing for arbitrary amounts of time, a notation like TAB is way more appropriate, and if internally there are several "voices" in use or just one, I frankly don't care. We just need to have a comfortable way to tell MS that a note doesn't stop right away when another (on another string) is played, that's all. The workaround using the piano roll editor works, but is tedious, imprecise, error-prone and unfit for testing a lot and trying out lots of different variants of a melody.

Attachment Size
under_the_bridge.mscz 20.22 KB
under_the_bridge_intro.png 152.64 KB

In reply to by Alexander Fetke

Your example shows that 2 voices would be enough if a better "Let Ring" option was available. It certainly doesn't show that 6 voices are needed - this would be a "sledge hammer to crack a nutshell". PRE is the current workaround and it is, as you say, really tedious. It's also limited to x2 duration.

You have used a single voice which has meant more work required in PRE.

I add "x" mutes as stave text and move them onto a string.

Hammer On and Pull Off can be indicated with texts but there is no easy playback modelling. You can reduce the volume but not much else. (With a custom soundfont you can reduce the attack).

Maybe getting "Let Ring" to work properly (and simply) would be a really useful step in the right direction.

In reply to by yonah_ag

I never wanted six voices, I only wanted a way of letting six independently started notes ring (because that's what a guitar is capable of). I agree that this probably should not be done using voices; before I started this thread (when I didn't know the PRE workaround) the only option I knew to achieve this was by using one voice per string, so I was in trouble whenever more than four strings were in use in a picking pattern.

In the PRE I can increase the length of a note indefinitely (see my example), so I don't know what you mean with "limited to x2 duration". It just involves pulling the length indicator up on the screen way beyond the UI element and the window borders. Maybe we have different MS versions if you really cannot do that on yours.

In reply to by cadiz1

With that let-ring method in measure 3 you let the phrasing hammer-on note (the high e) ring along; that's assonant and clearly not wanted as that note is just a hammer-on-pull-off phrasing on one string, so it cannot ring on a guitar.

Of course, in this example it is barely audible but a phrasing like that could have happened way earlier in the measure, then it would mess up the sound more drastically.

No, this is just the wrong approach. Not every note is supposed to hold until a global stop of the "let ring" area, just the ones whose strings aren't used otherwise (or muted explicitly) are supposed to ring. We need a way of expressing this. Otherwise we can use the workaround of the PRE. "Let ring" is a workaround for some simple cases.

In reply to by Alexander Fetke

"No, this is just the wrong approach. Not every note is supposed to hold until a global stop of the "let ring" area, just the ones whose strings aren't used otherwise (or muted explicitly) are supposed to ring. We need a way of expressing this."

Exactly. The current "Let Ring" is broken and a fix would help so much. It's basically a modified piano pedal which is too simplistic.

In reply to by Alexander Fetke

" in this example it is barely audible "
I like to hear you say it!
That's more than enough for the average person (or guitarist) and you save hours of work - which you can spend playing your instrument, instead.
If you want a perfect rendering, do it on your instrument. Otherwise, scores are meant to be read and played, first and foremost.
Just out of curiosity: what is the destination of your files? For yourself, students, musescore.com, or whatever?

In reply to by cadiz1

I just noticed that you also cut the high d right in the beginning of the measure short because your let-ring starts only just after that. And that missing high e is really clearly an audible error (how could I miss that then on first listening?). You also cannot let the let-ring start right in the beginning because of the hammer-on switch vom C to D in the lower register (you would let the dissonant C ring otherwise).

No, I don't want crutches! ;-) I want a proper solution to the problem, and currently the PRE is the best we have, but I don't like to have to do all the tedious and error-prone adjustments there.

I also cannot simply play what I write on my instrument because sometimes I compose stuff I cannot play myself, sometimes I want to play around with the speed of the stuff, different tunings etc. To produce each time a new recording is way too much work, even if I can play it. And let's not forget what computers and programs like MS are made for: To enable us to produce music on the computer. If it fails short in this — well, so be it. But then it can be improved ;-)

If your point is that scores are meant to be read and played by instrumentalists, we are just talking about different things. I'm here only talking about MS's ability to play the scores (to render them audibly or as audio files), and it already does a remarkable job in doing so. Having a great score on paper is one thing it does. If that is the foremost thing for you, fine. But for me the other aspects are also relevant (and actually even more).

As to the goal I'm aiming at: I compose. I want to memorize the things I play on instruments for later. Sometimes I have ideas and first want to hear which version I like best before starting to practice to play them. I also produce music (i.e. audio files), and I often start with a computer-generated track which then gets added on or replaced with instruments. I like to have lots of tracks to choose from, to mix. If the midi-like guitar tracks MS produces sound like shit just because the notes are all cut short, I have one source less I can use. The idea is to have something which isn't perfect but which convinces others that it is worth working on. And the people I'm interested in don't look at the sheet music and decide based on that.

In reply to by Alexander Fetke

"If your point is that scores are meant to be read and played by instrumentalists, we are just talking about different things. I'm here only talking about MS's ability to play the scores (to render them audibly or as audio files), and it already does a remarkable job in doing so. Having a great score on paper is one thing it does. If that is the foremost thing for you, fine. But for me the other aspects are also relevant (and actually even more)."

You didn't answer the question: basically, for whom and for what (in other words, what is the project or projects) do you write music on computers.

In reply to by Alexander Fetke

I was writing an EDIT during you responding me :) I.E:
EDIT: apparently (I'm even sure!), you didn't know the time when notes in MuseScore, in case of shared noteheads, let's say an eighth note in voice1 and a half note in voice2, were cut off when reading "in favor" of the shorter duration. It was really a terrible nuisance (and I worked, indirectly, to have it corrected). Now, what bothers me the most, let's say the most blatant lack, is the impossibility to make harmonics heard on guitar scores.
For me, this would be the priority of the moment.
For the rest, go to any score on musescore.com, and you will see that it is audible and correct enough in this way. And again, this is my position, guitarists (instrumentalists) will not hold this or that imperfect note duration against you, I strictly personally have never had such remarks. They want to listen once or twice, and if they like it, they print the score and play it. This is the main goal of a musician, it seems to me!

In reply to by cadiz1

That's one way of using MS. I'm using it differently, and clearly MS wouldn't need any decent samples, ways to manipulate the articulation, bends, glissandos, etc. if this was the only indented use. We could always just say: Just look at the sheet music, you know what it's meant to sound like, play it like this.

But MS also aims to produce decent audio output, even if that isn't the priority for you.

In reply to by Alexander Fetke

"Other musicians", including guitarists I suppose. And who, a priori, play their instrument :)
As for the "investors", something that is absolutely not in my thoughts and aims (it would be difficult, I do not compose!) ;)
However, I wish you all the best and hope that the next version 4 will bring improvements in the direction you hope.

In reply to by cadiz1

Alexander is only asking for a proper guitar "Let Ring". This is entirely reasonable and would improve the playback of most guitar scores - possibly attracting more customers.

There are so many decent guitar scores on .com which are half crippled by no let ring: they sound choppy.

Wouldn't piano scorers request a "Pedal" playback option if there wasn't one. Other pianists would also, a priori, play their instruments so why is there any need for it in the playback? (Rhetorical question).

In reply to by yonah_ag

"I don't want other users to feel that I am being critical of their scores."

?? You are on the forum (not of MuseScore.com).
The problem is not to criticize a score (besides, if it was argued and constructive, what would be the problem?), it's that you find a score (scores) on musescore.com that you find "choppy". From there, we can have a concrete exchange.

In reply to by yonah_ag

It is surprising how many musicians can look at a score and decide if they like it without hearing it. It seems to me that if a score says "let ring", then that is what they will do. If a piano player sees a pedal marking, they will play it regardless if playback has it or not.
Let ring doesn't work in Sibelius, either.
It seems far more important to not sacrifice the score just to get play back.
Or, have two scores. One that is notated correctly and one marked up however needed to get the desired playback. Too bad the two can't be combined.

In reply to by bobjp

But would piano players be requesting a playback pedal if there wasn't one? I guess so since it adds a lot to playback. Without it they would be resorting to the workarounds mentioned in this thread. It's the same for Let Ring with guitar.

In reply to by bobjp

I disagree: the two can be combined. At least in my case, in what I ask for. I only ask for TAB notation letting strings ring in playback until they are used again or explicitly muted. That doesn't clutter the score at all (explicit mutes are, you might guess it, rather rare).

In reply to by cadiz1

cadiz1, "That's more than enough for the average person (or guitarist)".

It's not enough for the average guitarist.
7/10 guitarists want this feature.

In reply to by Alexander Fetke

I have used X to mute strings from a harpsichord piece arranged for guitar, to match the original.

Limited sample durations can work but there are complications, e.g. the ringing duration should vary with strength of pluck; the same note on different strings sustains differently. I have found that some SF2 guitar presets have quite a short sustain so I have increased it to 8s in a custom font.

6 voices does make logical sense for 6 strings but it's not practical. I have tried scoring in 4 voices and it becomes unwieldy. Just imagine all the rests that MS would generate in 6 voices. 3 strings plucked one after the other could be achieved in a single voice if auto-sustain was implemented.

Once 3 voices have been reached there is really a need for rhythm beams to be combined across voices. I now score in 2 voices and apply my own "let ring" plugin.

See also https://musescore.org/en/node/319324

In reply to by yonah_ag

Some things strike while me trying to read through this thread.

It is obvious to me that of course Guitar Pro is going to have a better playback engine than MuseScore. And probably always will.

As to let ring note durations, I wonder what we are talking about. Nylon or steel strings. Light, medium, or heavy gauge. Played with a finger nail or finger tip. Flat pick. Strummed with a thumb, or brushed with the back of the fingers. What dynamic?

There might just be too many thing affecting note duration that can't be replicated. I think is is more the job of notation to tell the player what to play, and less how to play it. Musicians know how to play their instruments.

In reply to by bobjp

Things like nylon or steel, gauge, finger nail, tip, pick, thumb, etc. are all a question of the sound font. If the sound font provides extremely short sounds (pizzicato or palm-mute or similar), then the topic just doesn't arrive because MS's behavior of stopping a note when the next is played will not be audible anyway. But for any sound font with longer sounding notes we have our problem.

Here we are just talking about how to denote the proper length of the sound if the string supports it. And typically we are talking about a duration like one measure, that should be supported by any string, unless you palm-mute it or similar. Even a hammer-on on a nylon string is audible for quite some time. We just want to avoid the automated mute which MS applies as soon as another string is plucked. (Given we don't use any of the workarounds but notate just simple notes one after the other.)

In reply to by Alexander Fetke

The MS synthesizer does not stop a note immediately the next is played; this is soundfont dependent and related to the decay parameters.

However, the lack of a well implemented "Let Ring" causes problems. A single measure should be the very least that a note is allowed to ring, (if not muted or replaced by a new note on same string), and this should happen even when the note starts on the 3rd beat of a 4/4 measure, i.e. it should ring thru until the 3rd beat of the next measure, instead of being stopped at the end of the current measure.

In reply to by yonah_ag

Sure, the sound font can specify that any note is supposed to ring independently of the others, but then again this is a global setting (for a harp, for instance).

I still need a way of specifying that the low C in measure 3 of the Under The Bridge example is supposed to be stopped when the hammer-on D in the next eighth happens (because it's on the same string) while the high e at the same position in time is supposed to ring. I still don't see how to do that with a general let-ring feature unless it can be applied specifically to just some notes, not all, which start at this position.

Furthermore, not shown by this example, I would need a way to specify where a ringing string is supposed to stop (be muted) independently of the other strings which might be supposed to continue to ring. That's likewise a thing the let-ring feature only can provide if it can be applied to specific strings.

And, yeah, I have that note-dependent let-ring feature in form of the PRE. But it's a crutch, unwieldy, it doesn't reflect visibly in the TAB scores, it's easy to make errors, inserting notes later will let some of your notes stay ringing too long (because they haven't been touched in the process), you also have to manually adjust the stuff which is obvious from the TAB score already, etc. A really good solution would be something which takes into account what is written on the TAB score and which understands that there are a number of strings which might ring until they are used again or explicitly muted. That doesn't sound too hard.

In reply to by yonah_ag

Try maintaining such a score (i.e. play around with it, try out variants, edit it massively) and then say again that this is less labor intensive ;-) I doubt it. Anything invisible is a PITA when it comes to changing things. The PRE is at least one single place where to look for the problem.

But if you prefer invisible tied notes, keep in mind that then again you need many voices, in the general case one per string. If you then happen to switch your staff to a standard notation it looks horrible.

The solution using only the PRE to tweak the lengths can be applied to a clean TAB score which looks like lots of eighths when switched to standard notation.

In reply to by Alexander Fetke

It is cumbersome maintaining such scores, hence PRE or plugin. I always use at least 2 voices as I like the bass rhythm below the TAB and melody rhythm above. This pretty well cuts out PRE editing for the bass and also allows for independent volume control without needing to select individual notes.

Invisible element can be toggled on/off with Ctrl-Shift-V

But all these workarounds would not be required if "let ring" worked properly.

It looks like this would be easy to program in a TAB score since the information is easy to process.

Here's a short extract from a Musescore file showing 2 PRE extended note durations. The 41000 was done by mouse dragging and the 2000 was done by entering a value in the Len field. All the necessary string and duration data are very clear to see and Let Ring would be an easy algorithm to implement.

Screenshot_20210512-232744_HTML Viewer.jpg

In reply to by yonah_ag

yonah_ag

I really appreciate your continuing exploration and reports on Musescore's "Let Ring" and our pursuit of realistic guitar playback implementation, despite all the headwinds.

MIDI's sustain message CC64 (≥64 = ON / ≤63 = OFF) is a simple logical switch, perfectly analogous to Pedal Down and Pedal Up on a real piano or a MIDI device. Thus piano sustain is easily implemented and works just as a pianist would expect.

The current "Let Ring" (aka "Laissez Vibrer") is accessible through Musescore's Line palette. It's name is a common/archaic term indicating the piano's sustain pedal; alternately it indicates that a decaying sound (like a cymbal crash) should be allowed to ring out naturally. While the "Let ring" label itself is meaningful in scores but the common definition MIDI implementations via sustain may fail to replicate playback of crash of a cymbal interrupted by a crash of the same cymbal.

As previously discussed, Musescore's "Let Ring" (Laissez Vibrer) implementation merely sends a MIDI sustain message. And that is readily problematic even for realistic playback of simple guitar finger-picking, largely because it imparts too much sustain and we will hear intervals like 2nds dissonantly overlap.

Musescore can lead the way toward improved finger-board instrument playback by providing a sophisticated approach for realistic playback for guitar and other fingered instruments. Requiring users to spend hours in the PRE is, in fact, proof of the lack of that implementation.

I'd advance the terms FILR or FILZ as acronym candidates for this proposed feature (and its score marking):

FILR — Fingered Instrument Let Ring
FILZ — Fingered Instrument Laissez Vibrer

The FILR/FILZ concept has been clearly expressed on this board multiple times ... and in this thread. I will attempt to clarify further in my Request post and I'll add an official request in MuseScore's Issue Tracker.

Imagine if there was not piano sustain ... then realistic piano playback would be achieved only by hours of tedious PRE editing.

scorster

In reply to by scorster

Thanks for this info and for the support. I am not familiar with how MIDI works so didn't know about sustain messages. Do they work at the measure level or at the note level? I'll try a PED symbol, listen to the playback and see what ends up in the MSCX file.

I'm not sure what you mean by "too much" sustain. Do you mean that the sustain is too long, or that it applies to inappropriate notes, or that it is at too high a volume, or possibly a combination of these?

I'm going to program a "Let Ring" now that I have seen durations > 2000 from PRE but it will work on MSCX files rather than MSCZ.

Your request post defines the situation very clearly.

In reply to by yonah_ag

>> I am not familiar with how MIDI works so didn't know about sustain messages. Do they work at the measure level or at the note level? I'll try a PED symbol, listen to the playback and see what ends up in the MSCX file.

> "Pedal down" turns sustain on via MIDI CC64 (value 0 to 63), on the channel to which the message is sent. The sustain continues, affecting all notes, until a CC64 (value 64 to 127) is sent, at which point the sustain ceases for all notes on that channel.

>> I'm not sure what you mean by "too much" sustain. Do you mean that the sustain applies to inappropriate notes,

> Yes. Sorry I was unclear. I meant that MIDI sustain, when applied to guitar playing, readily applies sustain to notes played in succession on a string, whereas that can never happen on guitar.

>> I'm going to program a "Let Ring" now that I have seen durations > 2000 from PRE but it will work on MSCX.

I look forward to trying your FILR Plug-in for MSCX scores! It could be of great value to guitarist seeking realistic playback! And that would likely lure people from Guitar Pro and TablEdit to Musescore, which I find much more satisfying than the aforementioned options, because the engraving, layout options and due to the overall ease of use.

>> Your request post defines the situation very clearly.

> Thanks. I appreciate your clarity and tenacity. See you later in the discussion!

scorster

In reply to by scorster

OK, I understand now.

My FILR plugin will initially be a 'plugout' as it will be written in Excel VBA. I have only written 1 MS plugin and this was really just a tweak of BSG's articulation plugin so most of the graft had already been done for me. So my p!ugin skills are too slow and limited at present. It would be better in Python than in VBA but I already have an Excel file which processes MSCX files and it would still be a 'plugout'.

In reply to by yonah_ag

Phase 1
• Completed score analysis and algorithm designed
• A working MSCX editor is now available in Excel

Notes 1
• Initially work with single voice scores.
• Initially limit let ring to span up to 2 measures maximum

Phase 2
• Code and test the algorithm
• Consider 2 voice scores

In reply to by yonah_ag

Ha ha!! Good point!

Initially I liked the FILR and FILZ acronyms because they sound like like the result they would produce:

• FILR is a "filler" tool that extends certain note play times to emulate guitar's natural sustain (better than can be accomplished with MIDI sustain) without changing the face value of the notes.

• FILZ "fills" out certain note play times to emulate guitar's natural sustain (better than can be accomplished with MIDI sustain) without changing the face value of the notes. But indeed, the FILZ acronym only works if you're thinking Fingerboard Instrument LaisseZ vibrer.

So you're right. The more likely candidate is FILV (Fingerboard Instrument Laissez Vibrer). Problem is, it's just not as fun to say.

Regarding Fingered vs. Fingerboard vs. Fretboard ... the term Fretboard initially came to mind but I wanted a term that applies to fretless stringed instruments as well. So I then settled on Fingered. But now I prefer Fingerboard as it naturally includes both fretboard and fretless instruments.

So I'll cast my vote for FILR, but with F standing for Fingerboard rather than Fingered: Fingerboard Instrument Let Ring

In reply to by scorster

Just don't make it Manual Instrument Laissez Fibrer.
That acronym's already taken.

I appreciate your initiative to take this to the next level :-)

(Btw, how do I get a proper profile image / avatar? I already set a proper picture, but that seems to be working only on .com and not here on .org. But I found no way to set it explicitly for .org.)

In reply to by scorster

Finding time will be my main problem but it will start life this weekend as FILR.xlsm, (since I'm English), but I do also like the sound of FILZ, (FILV just doesn't have the same ring to it). I need to make some test measures with various time sigs, note lengths, dots, rests, (mutes), and triplets.

Anything else to consider? Perhaps a maximum ring of 2 measure durations since this is more than most soundfonts support. At least voices don't provide any complication.

Do you still have an unanswered question? Please log in first to post your question.