Possible to have Slur Playback?

• Jan 22, 2010 - 16:31

Countless times when writing with musescore, I've often wished that it would have playback for slurs. Would it be possible to have them effect playback? Instruments like the Violin and horns which can naturally slur very easily anyway (not sure if there's actually a name for it), sound very awkward in play back playing each individual note as they move up or down the note range which I want them to play. If there's a way to slur it, it would sound a lot more natural IMO. I have no idea if this would be possible, but I do think it would make a really cool addition to what is already a great program.


Comments

General MIDI does not have different sounds for slurred and slurred notes. Do you know of any free sound libraries that have this distinction?

In reply to by David Bolton

No, I don't. I wish I did. Would it be possible to have a device of some sort in the mixer that could alter note pitch in order to create a slur? I have no knowledge about programing, so I don't know if this would be possible or not. From a non programing perspective, that's what I see as the best practical way to create a slur.

In reply to by Michael M

Synful Orchestra is some kind of 'procedural' orchestra and not just a set of samples. It can actually interpret the melodic lines and output a 'humanoid' interpretation, that gets better if you activate the Delay for Expression check. (It then takes 1 second to pre-process the sound, but the results are beautiful).

David, In any case, there isn't a 'slur' sound in any GM instrument. But the notes under a slur should be completely legato (not a gap between them).

It's a matter of fine tuning the duration of notes to match articulation, in the order: slur (no gap, or maybe a slight overlapping?), portato (almost no gap), normal note, staccato (short), staccatissimo (shorter) etc.
Either Synful, an external library, or a simple GM synth output would match interpretation better from this extra midi information.

In reply to by elerouxx

MuseScore does not have any gap between playback of normal notes (or slurred notes). Staccato notes play as half the duration. Coincidentally, this behavior matches that of Finale and Sibelius.

Some early versions of MuseScore experimented with some gap between normal notes (85% of full duration) but the results were generally less satisfactory than full duration (100%) for normal notes.

I'm not capable to do this, but it seems to me that this may be a candidate for a plug-in, using the pitch bend capability of MIDI. I can see a command to slur from x to y over z amount of time and when complete, change the new note to its actual pitch.

Regards.

Try using the Tenuto "-" marks, in the Display-Palettes-Articulations, Ornaments. There will hardly be a pause between a Tenuto-marked note and the next note.

In reply to by xavierjazz

Xavierjazz, slurred notes for wind instruments means that they are not tongued so they do not start with a strong attack. Slurs for stringed instruments indicate bowing. It does not mean glissando or portamento.

Slurred notes do not a break between them. Earlier versions of MuseScore had breaks between notes by default. The latest prereleases and nightly builds do not. Probably misterdls is using a version of MuseScore that breaks between notes by default.

In reply to by David Bolton

Although I think what is being talked about is more of a glis than a slur. Most soundcards support the pitch bend controller. It shouldn't be all that mysterious. As for a true woodwind slur as David mentioned, there is also a contoller for that, but it is not supported by all soundcards. It is however standardized where it is supported so using the controller (it is sustain 2, not portamento) shouldn't conflict with anything else. When this controller is activated, its function is to sustain one note until the next is played and to skip the attack section of the next notes envelope. This should work pretty well for single lines and could be notated as a functional version of the slur line.
There seems to be a reluctance to add the functionality of even the most basic midi controllers (eg. sustain for piano) to Musescore that I do not understand. Can anyone explain it?

In reply to by MDMilford

My view is that notation software developers and sequencer software developers adopt opposite attitudes. Usually notators have only rudimentary MIDI support (how many of them support sending SysEx Messages?), while sequencers are satisfied with only rudimentary notation support (how many of them support multiple voices per staff?).

In the 1990's we used to mimic legato using the Piano Roll view, and manually or automatically (depending on the software involved) prolong the endings of all notes, with extremely short durations (like make a note's value 101% of its nominal value), to overlap with the notes that happen to follow. As far as I can remember, it wasn't bad at all...

Hosam Adeeb Nashed

In reply to by HosAdeeb

Hi HosAdeeb, Sibelius, for instance has a facility for directly editing the output MIDI file - and as such, sending SysEx messages. The purpose of 'notation' facilitiy of MIDI editors is rather a simplification only of seeing the notes per channel,; and not to produce a playable score. After all, if there are 2 flutes, seeing the notes on the (MIDI's) staff is simpler then reading the long, monotonous lists of "timeframe", "note on+pitch", "velocity" ..... "note off+pitch" . MIDI, by nature, does not (isn't instructed to) see the 2 flutes as 2 different entities (as it is performed by the same channel). In contrasts to any notation software that has to show the voices separately (on the same staff or on 2 separate staves) as the given sequence is to be played by 2 separate flute players at the time of the performance.

In my earlier comment (which, by mistake, appears twice - my apologies, but I have no ways to remove one of them) I described the nature of slur and why the earlier solution of slur emulation by over-extension of notes failed to give a natural feel.

The reason why the performance of slurred notes is difficult (in MIDI or otherwise) is due to the nature of physical sound and that of the perceived sound. Looking at it from a purely physical point, we may easily think that the slur effect can be easily achieved by not finishing the first note before starting the second one. (Having such an 'envelope' - the graph of the loudness of a given sound - where the 'off' state coincides with the next 'on' state). In fact, the graphic representation of 'slur' could be described as singing notes with the same syllable, without taking a breath. And what happens when two slurred notes are the same? The same note sounds longer, that's what we call: 'tie'. (In many a language therefore there is no distinction between the terms 'slur' or 'tie' - but that's an other story).

The problem with this concept is that in the real world legato (slur) never happens this way. Think of the sound of a piano. Physically speaking, the envelope of the piano sound starts with a sharp 'attack' immediately followed by a rapid 'decay' (and, even when we lift our hand from the keyboard, we have a reverberation - never have a real silence during a piece of music) Nevertheless, we definitely distinguish 'staccato' and 'legato' within piano pieces in spite of the fact that neithe one is physically what we 'perceive'.They are usually achieved by'micro-rhythmical adjustments', ie. making certain notes 'longer' (more precisely, delay the start of the following note) , or (reversing the above) 'shorter'.

I know of a (very pricy) music notation software that uses the micro-rhythmic adjustment principle for the 'life-like' playing of the score - most probably adjusting the envelope values accordingly in the course of creating the MIDI file.

I hope to meet such facility in MuseScore befoie long.

The reason why the performance of slurred notes is difficult (in MIDI or otherwise) is due to the nature of physical sound and that of the perceived sound. Looking at it from a purely physical point, we may easily think that the slur effect can be easily achieved by not finishing the first note before starting the second one. (Having such an 'envelope' - the graph of the loudness of a given sound - where the 'off' state< coincides with the next 'on' state). In fact, the graphic representation of 'slur' could be described as singing notes with the same syllable, without taking a breath. And what happens when two slurred notes are the same? The same note sounds longer, tat's what we call: 'tie'. (In many a language therefore there is no distinction between the terms 'slur' or 'tie' - but that's an other story).

The problem with this concept is that in the real world legato (slur) never happens this way. Think of the sound of a piano. Physically speaking, the envelope of the piano sound starts with a sharp 'attack' immediately followed by a rapid 'decay' (and, even when we lift our hand from the keyboard, we have a reverb - never have a real silence during a piece of music) Nevertheless, we definitely distinguish 'staccato' and 'legato' within piano pieces in spite of the fact that neithe one is physically what we 'perceive'.They are usually achieved by 'micro-rhythmical adjustments', ie. making certain notes 'longer' (more precisely, delay the start of the following note) , or (reversing the above) 'shorter'.

I know of a (very pricy) music notation software that uses the micro-rhythmic adjustment principle for the 'life-like' playing of the score - most probably adjusting the envelope values accordingly in the course of creating the MIDI file.

I hope to meet such facility in MuseScore befoie long.

I play cello, and I use to think a lot on legato...

For me slurring two sound means that the second sound, should not have any kind of attack or accent. And the first one should last until the very begining of the second. That means that any variation of the sound in the first note, like a crescendo, a diminuendo or a change in the colour of the sound, should be continued in the second.

I think that the idea of letting the end of the first note sound while the second begins won't give a good legato, since in this moment, both notes will sound and dynamic will increase creating at least a very soft accent.

Who knows... damn legato! :)

In reply to by alkayata

The fact is that markings are interpreted differently for each instrument. Piano and organ try to approximate legato by slightly overlapping the notes and often giving emphasis to the first note of the slur. From what I understand string instruments interpret slurs as bowing markings rather than legato. In a related example, trumpet players interpret a sustanuto marking as a sustained note with very light tongue (softer attack than usual). Piano players interpret a sustanuto marking as a light accent (stronger attack than usual).

In reply to by xavierjazz

Hi Xavierjazz, may I suggest an experiment to demonstrate, how every piano sound decays and every new sound has a separate attack (accent)?

Try to play a tune and hold down the sostenuto pedal (either on an acoustic or electronic piano). Can you, over the chaotic blurr, hear the melody you play?

Second experiment: try to play a 2-note sequence (eg. D-C crothchets) - obviously, without using the pedal. First time play them as close to the same dynamics as possible. They will sound as 2 separate notes. Second time play the first note (D) mezzoforte, then the second (C) mezzopiano. I bet, the second time you will 'feel' as if the two notes were slurred.

In reply to by xavierjazz

In fact, you had mentioned the sustain pedal first - and that is to what I responded. Nevertheless, did you try the two experiments? I just wanted to state that holding down the key cannot directly sustain the sound for matching the volume of the next note - due to the fast decay of the strings. (And, as it happens, the sustain pedal on acoustic pianos produce more sounds than the previously depressed keys - due to resonance etc.)

In reply to by David Bolton

I would say that there is some kind of utopic idea of legato, and every try to do their best with it's own instrument to approach to it. In fact, that's what makes possible to join the sound when many different instruments play together. What would happen then with Messiaen's End of Time Quartet, or Martinu's trio for flute, piano and cello... All that instruments have really diferent articulation.

Although every instrument does different things to sound "like legato", it seems that legato is opposed to an accent: wind players make a soft attack with tongue, string player will try to make smooth bow changings, and piano players make use of resonance and dynamic scaling to create the sense of a continuous line...

Maybe the solution of resonance in a synthesizer makes sense to emulate legato on a piano, but it sounds weird when its applied to an string instrument. Maybe, to synthesize different instruments, we should use different methodes of making legato... It sounds like a lot of work, but from may point of view, it makes sense.

Say it is worth twice: damn legato!

In reply to by alkayata

I would very much like various ways off achieving a smooth effect in Musescore. I do understand that it is first and foremost a typesetting program, but reasonably accurate playback is still important. It would be easy to get bogged down in details in this area, but generalizations could, and I think should be made. I think the instrument list can be divided into a small number of groups and legato handled the same for every instrument in that group. One group would be the wind, brass, and string instruments in which notes are lengthened to 100% of their value and the attack of every note after the first is omitted. Another group would be attacking instruments like organ, piano, harp, vibraphone, etc. where notes imperceptibly overlap, maybe 101% of time value, but the attack is sounded on every note. And finaly the instruments for which legato is meaningless, like xylophone. This would be a reasonably accurate and still achievable model for legato, without being overly concerned with the details. I could say for example that if a bassoon plays a legato octave jump from a lower note to a note between b flat, and d above the bass cleff staff, that the attack of the second note will sound fuzzy as the bassoonist flicks the appropriate key to make the jump. I think this is the type of nit-picky details that should be ignored.

In reply to by MDMilford

do we really need to reinvent the wheel? MuseScore could and should take advantage of open software and communities dedicated to improvement. Non-free notation programs have already been able to solve legato, and we would not be breaking copyright laws by analysing such MIDI files that have the best, most life-like legato. Alternatively, we could obtain MIDI conversion files from audio source.

In the past, I have seen and used a very simple MIDI assembler/disassembler for some work in order to create MIDI files 'on the spot' for educational purposes. It was not that hard. The demos of Yamaha XG 50 - back in the dark ages - achieved very good piano legato sounds and the method could be easily followed (attack and decay velocity changes slightly adjusted to the pitch too) . To emulate such a logic should only need experimenting a bit, especially taking into account that excellent soundcards are available with thrown-in MIDI studio software.

There would still be an issue to decide: although it would be more elegant to obtain excellent playback sounds generated purely by software - but with a relatively small software overhead near-perfect results could be achieved by using specialised soundcards that are not that expensive nowadays.

I can see where it could be a problem for various instruments to play legato correctly. Since I'm now taking Piano, I've found that they have almost an entirely different of idea of legato than a woodwind player does. For instance, when I play legato on the Flute, I simply "slide" from note, only tonguing the first note of the sequence. This beautifully slurs the notes together, and is something that I very much wish could be replicated (or at least partly replicated) in this software. I understand the difficulties, but I do think it would be nice to have legato play back for instruments that commonly use it (such as woodwinds/strings/brass).

Do you mean glissando? Think of a slide trombone in a circus band or a slide whistle... It sounds to me you are looking to smoothly smear from one note to another without jumping to each individual pitch between them. If that is what you mean, it is a "glissando"

In reply to by joe.shupienis

Glissando is a quite different effect. I think that if it ever is implemented, it would be through the synthesizer.
To solve the legato issue, I've found that a good workaround is to overlap about a 2 % the end of the note with the following one (actually the percentage depends on note duration). As most patches have some sort of non-instantaneous decay, this provides a smooth crossfade that is akin to a legato. To do that, select the measure, right-click and select piano roll editor, then select the note and set duration to 1020 (or something like that).

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