Split Midi track to double staff
Hello,
I'm a new user of Muse Score and I would like to do the following:
I opened a *.mid file and made an excerpt of one track. This track is a piano part which is usually noted with a double staff containing a treble clef and a bass clef. However the piano part is all noted on one staff with a treble clef and I don't find any way to split this up between a treble and a bass clef. Is this possible in any way?
Thanks for helping,
Tokke
Comments
Same problem here. Also, when I tried adding a new piano instrument, copying the notes to the new instrument, and moving the notes between the staves, it messed up the note stems. I guess the workaround would be to edit the midi file to use separate tracks for each staff you want, but it would be nice to be able to add a staff to an instrument. (And if any developers are reading this, an extra point for giving the option of choosing a split point for the notes - i.e., "if the note is below middle-C, put it on the lower staff"!)
In reply to Same problem here. Also, by gotgrip_com
The ability to split the notes of a single staff on to two staves is already implemented in the nightly builds . Right click on the staff and choose "Split staff".
In reply to The ability to split the by David Bolton
Hello,
I'm a new user too and I would like to do exactly the opposite : I opened a *.mid file for a piano solo and I have 3 staves, one treble clef and two bass clefs...
Is this possible to merge the two bass clefs in one ?
Thanks in advance for your help.
In reply to merging two staves by gambic
I can't think of a way to merge two staves in current versions of MuseScore.
In reply to merging two staves by gambic
I'm sure there are sequencers (freeware) out there which can import the 3 staves and allow you to merge 2 of them and then spit out a new .mid file.
In reply to The ability to split the by David Bolton
Hi , I am new here. Nice program and am forced here to try to find a program that does what I want as a composer.
I have tried Sibelius , and that fails, tried Finale both versions and that fails, Tried Encore, and that too fails, but showed promise.
They all have major bugs and failings for any credible use.
Essentially I have a Piano player performing my work. I need to capture midi as well as WAV Real time. NONE of the notational software progs capture what you play live on the keyboards. The On screen Notation is hopelessly unrepresentative of the live performance, and it gets worse.
Eg If I export and save as a midi file, ( a Piano score ) I end up with a midi file that is bass clef only , and hence looking for a way of recreating what the software " should " have done in the first place a Piano score. Have read the MuseScore comments, and the response said that you could do that in the nightly betas. Sorry but I tried that and there is no such new feature, beta or standard feature. Any ideas what and where you can do this task ?
Also I have used a free package called RED DOT forever , which everyone raves about and is supposed to keep things simple and basically is a midi recorder and not at all like any of the heavyweights. So am I right in thinking that midi files contain " no time signature, no key signature and no bar lines ? and no temp statement ? And that they are only assigned upon importing into notational software ?
My request also would be for someone somewhere to confirm that midi files indeed do not contain anything more than pitch velocity and tempo credentials, and that in trying to create a screen visual of sheet music, that it makes more sense to recognise fundamental errors that all manufactures of notational software are making, ie that they should not dictate rigid structures to host your musicv, assuming every piece of music is to a beat, as that is pure nonsense in the real world of music.
My material are phrase specific and some phrases cannot be contained within normal bar structures, so its very simple as a suggestion.
Importing a midi file in a " composer " mode, should NOT put bar lines in at all, and should offer a suggestion of where the first bar line should be, and allow manual insertion, and bar removal - then you tell it the time signature , and then the remaining basrs are auto spaced. BUT, IF we want to move a bar for " whatever" reason the we can delete it and reposition it, and have auto remapping of the new bar lines, or freeze the rest of the bar lines whilst the work area called edit mode working has been finalised and then allow for custom bar insertions , maybe even different time signatures for a couple of bar to accomodate the phrasing and the rest of the bar line positions remain intact.
I am utterly amazed that software is being developed for musicians without any regard for the end user. Am sure each manufacturer has its musicians who contribute the features, but fundamental workings are plainly missed out of ignorance by someone. That is why I will not waste any more money on software unless I can see the credibility of the program in action. I hope MuseScore meets my requirements as I am at my wits end.
Hope someone understands all this, cheers if anyone cares
Steve
In reply to Splitting a Stave to Piano Score by steve2010
Listening to a performance and turning it into music notation is difficult for humans and even more difficult for computers for anything other than very basic pieces. A MIDI file contains information about how the music is performed: note pressed on the keyboard (pitch), time the note is held (duration), speed that the note was press (volume), etc. A MIDI file does not contain information about how to present the notes on a page. Add to this mix rubato and tempo changes and the computer will inevitably make mistakes.
There is plenty of software that is developed for musicians and the end user that do lots of incredibly useful things. Its just that the specific feature that you are looking for is not what computers are good at. You wouldn't expect an old-fashioned pen to magically write out the notation for you. This doesn't mean the pen wasn't designed for the "end user". It means it was designed primarily for a different purpose. The fact that computers can notate performances even a little bit is pretty incredible if you think about it!
In reply to Listening to a performance by David Bolton
Hi , thanks for explaining. I agree that it may be by total misunderstanding of the correct procedure for Piano midi capture. What method would you recommend? I accept that it is an impossible task we set our Notational software to guess or anticipate our next move and inevitably it will come out wrong.
So if Midi doesnt contain the musical structures that we recognise in sheet music, then there is hope for mye. I suspect that midi is quite capable of recording the 100% performance across the pair of hands playing a piano. I will try your program to capture , but am I not going to be doing same mistake again ie capturing in the notational screen ? Or do you have a non notational screen elemntary capture utility like Red Dot that allows importation into the notation module later for analysis and sheet music interpretation and indeed editing modes that allow you to correctly align bar lines and tidy the " correct " midi capture and - eventually - have the precise notational cleaned and refined notation as the result. At least if I coul eg define the first bar ( as often melody lines start mid bar and to define the precise position in the notation software simply has to be a manual preference and placement. So can I subsequently remove all bar lines or manually reposition or nudge them forward or backwards by predertmined amounts such as a crotchet or quaver to have eg genuine beats arrive on the appropriate points in a bar as opposed to the unsightly position of a beat mid bar - which is totally wrong emphasis on reading or playback. I hope you understand the inadequacies of all notational software, with respect to cater for the all musicians, and not ju=st composers the way things are at present. Somewhere somebody seems to be assuming what is right, when it is clearly wrong to me. It would be fantastic for you to understand these elementarry things and implement as it is so easy to do as the exprerts. I like your software and hope I can use it in the way a composer needs to use, eg like a blank manuscript book we all used to use
regards
Steve
In reply to Midi Music failures and inflexibility of all notational software by steve2010
Here's a review of MIDI import on several programs: http://www.skytopia.com/project/articles/notation.html
The reviewer was working with MIDI files that were created from a computer (strict tempo, tracks, etc.). Converting from a real performance (as I mentioned earlier) is even harder. The article is dated to 2004.
In reply to Midi Music failures and inflexibility of all notational software by steve2010
The "normal" workflow when transcribing real music is IMHO bottom up. Bottom up starts with determining the rhytmic structure of the music. Then fill in more and more of harmonies, melody etc. until it sounds more and more like the recording. "Top down"
starts with a midi recording and tries to reduce it to a readable score. This does not work well with current MuseScore. It does work somewhat with simple music but produces garbage with complex expressive played piano.
Traditional composing works with a piano, pencil and paper. Compared to that a computer scorewriter is much more comfortable.
In reply to Workflow by [DELETED] 3
Hi, Thanks for the alternative viewpoint. I can't fully agree, as a composer needs space to as you say, prepare the dots on paper with a pencil. The computer way of course is more comfortable and I see it as a boon send compared to the laborious way in which Beethoven, Mozart, Chopin and company all wrote down their compositions. But you know what, You start with a blank manuscript and like painting by numbers you start to fill the staves. Your idea starting with a rhythmic structure " pre-supposes " that it's a regular piece ie constant time. Let me share a few thoughts for the night.
Music notation was invented as a means of somehow documenting the composition for others to play. But in the real world of music and performing we have an entity that makes music more beautiful I call " light and shade ". To create that is not all softness and dynamics in a performance it's also very much faster or slower performing of the melody down to an individual note being emphasised almost to a pause in the music and re-commencing. In a real performance the conductor of an orchestra does this based NOT on bar lines and time signature but variables we come to know as expressions, which litter a musical score and actually makes a nonsense technically of a musical score. Midi,is notationally speaking overlooking the obvious need to drag music scoring into the 21st Century and is modelled on perceived past requirements and does not reflect accurately a musical performance " as heard ". That is the fault of the programmers who don't understand the needs of composers, only musicians who play the scores.and the two are different fulfilling requirements. So this was a missed opportunity to be innovative and perfect midi notational screen displays to intelligently allow for Light and shade performance by if necessary allowing custom bars of different eg BPM content - but even that would not go far enough. On the other hand a simple midi recorder eg RedDot should be able to faithfully capture a midi light and shade performance and it couldn't care less about bars and staves etc, because for the first time, electronically we capture what we play. Yet when it comes to creating a notational interpretation, it's as crude as working by candle light in the dark ages. Can you see, the problem ? Notational software cannot even create a decent midi file from its own notation let alone record in real time a performance properly when it is trying to write notes on the stave and " anticipate " what the next few bars might be, so the cache and buffering and screen writing methods used by all software is by experience flawed. Better, to capture raw midi, create a suggested screen representation " post " recording with highlighted bars and zones requiring clarification and re conforming to the accepted mathematics of note timings , EVEN allowing for intermediary custom bars and overide editing capabilities that include bar line nudging or removal or insertion elsewhere in the phrase.
Finally, as a composer, why would I wish to be constrained by a suggestion of a predetmined rhythmic structure ?. . It''s like telling Michaelangelo he must paint within a graph paper style grid. When in fact he is painting a distorted image on a curved domed ceiling, which is intended to reflect a 2D plain illusionary corrected perspective when gazing up ? Imagine looking at your reflection in a spoon and the image being correct and not bent. Now how innovative is that by a painter, who thinks outside the box. Well, music is no different, but the tools to assist us are underdeveloped and miss the features that we need to compose our individual creations.
I hope that others understand the fundamental issues her and that the good writers of this software are listening with interest to be bold enough to feel inspired to incorporate this common sense approach to modern music needs. We have fabulous DAWS but ancient midi and notational software interpretations of music that is like a blind man asking " what colour is the wind ? " Hope things improve as am so frustrated with all the stuff out there inhibiting creativity.
thanks for listening to my rant
Steve
In reply to Listening to a performance by David Bolton
Thanks, I sentiment the observation, but much of the software could be easily developed to cover that need because there is nothing out there that does which you seem to suggest.. There is nothing out there and someone kindly pointed me to a review of notational software to interpret midi, and surprise surprise , nothing does it properly. So the " clean " and tidy type editing features INCLUDING bar shifting nudging or whatever else you may wish to call it is very much needed as a staring pont. s for software incorrectly writing midi to screen is an example of poor prpgramming by some companies who do not test their products before putting them on sale. No excuse there.
In reply to Thanks, I sentiment the by steve2010
As I've already mentioned, the fact that none of them do it well is not an example of poor programming. Have you tried to write a program that could do that?
In reply to As I've already mentioned, by David Bolton
Hi David, No offence was intended and is merely frustration on my part not to be able to use notational software from a composers point of view. I would be the ideal person to give programmers an insight and ideas on how composers work and why they need simple fundamental scratchpad features before pressing an " automate " or " auto conform " button that tries to tidy up what you have done.
Was hoping I could open minds to the general failures of notational software programs that have missed the core point of their mission statement to make a worthwhile product available to us users. It is not intended to be a criticism but a simple observation where others may not have seen the wood for the trees. The reason I can see the wood properly is because I am looking in at an early planting stage before the woods become a mass of foilage and fully developed, and in planting the saplings I have a choice on control on my own spacing of the saplings eg further apart on flat ground and closer together on inclines where I want the roots to bind the soil to avoid future landslides of ground etc. Sorry about the analogy.
My Pianist cannot get his head around some of my phrasing, simplybecause the beat stops and freeform takes place over a partial bar before it picks up again. His body clock cannot cope with what I am doing to play in time. My explanation is simple, he doesn't know the tune and therefore he cannot accompany something he doesn't know. Despirte his Diplomas in Music etc, he says I can't do this and I can't do that. My response is Rubbish ! a composer can do whatever he likes, and should not be constrained by some music rules invented centuries ago but never updated to make allowances. Music notational software is meant to be reasonably representative of the composers work but never is.. He is quite offended by what he sees as a musician composer who studies music at grammar school 50 years ago and despite less expert knowledge ignores the rules and what he has learnt. What rules ? Who has the right to say how Chopin intended his music to play, and are we so naive to believe that a piece of manuscript paper handed down over the centuries that relies more on expressions above the stave to convey the piece than the dots below, is an accurate representation. I think not. If you were a witness to a Chopin or any Classical Composer performance, you would have hained from the witnessing the feel and performance . But as with all things our memory of that performance becomes muddy with time and even with notes to remind us drifts into something that could be quite different.. I saw Jimi Hendrix live twice, and even IF I could play like him and had the dots , do we honestly believe it will be representative of what he played ? Light and Shade is the key and no-one tod ate has found a programming way to cope with that element of music. Handling note pitch, velocities and duration is chicken feed compared to " really " interpreting what is going on in the performance, and that is where notational software is based on old traditions and music language now past it's sell by date, so we need 21st century notational tools that does away with " expressions " which are all way tool subjective for me as a composer to guide others how to play the tune. It's as crude as Leonardo DaVinci having post it notes all over the Mona Lisa, telling us what the colour mix was, and which way to make the brush stroke when painting etc. It is an absurd unintelligent approach to what is now a $ billion dollar music industry where technology blows us away and yet we rely on an electronic midi piano roll but with no control over the speed of the roll in technical terms which as I said makes a nonsense of the time signature system. One minute the guy is winding two turns of the handle every 5 secs, and then he does 1.5 turns and later 3 turns etc no consistency, but at least that is what conductors do. They speed us up and wind us down, and that's hardly music correctness technically , when we could document precisely thye raw midi file to be 100% what we want as a composer and not rely on stupid conductors who " try " to guess what the composer intended. - which is the truth of how absurd our music world is.
Very few drummers can hold perfect time, but that doesnt make the music poor. If it;s quantised and becomes perfect time, the we lose the human feel and end up with house style robotic stuff.
These threads are very interesting to me to see how others see music, thanks
In reply to Hi David, No offence was by steve2010
Helo there,
There are a number of ways of improving the results of translating 'free tempo' midi files into notation. They all ultimately rely on adding a midi tempo map to the original file.
First the free one. Google for DIV's Midi utilities. These are a set of command line utilities for processing midi files, so not particularly user friendly. One of the utilities is called tempo-map.exe. To make a free tempo midi file sync up with the 'midi beat' what you have to do is (working with a midi sequencer) record an extra click track into the original midi file. You make this click track to follow the perceived beat of the music, (not attempting to click in strict time). You then run the midi file with the extra click track through tempo-map.exe utility specifying that the tempo map should be created from the added click track. The resulting output file plays the same nuanced phrasing as the original, but when loaded back into a sequencer or notation programme the notes are all 'on the midi beat', or much nearer than they were originally. After carrying out this procedure of course you have to load the resulting file back into a sequencer and delete the click track .
Secondly, the WIDISOFT Able midi editor costing $25 to register, available from www.widisoft.com is a midi piano roll editor with draggable barlines, much as you have been suggesting. You can do a preliminary sync up by playing the file inside Able and 'tapping along' where the barline should be. The final adjustments of the barlines can then be made graphically by dragging them into place
Thirdly, Midi Maestro4 $99.95 from www.midimaestro.com allows you to carry out a similar procedure to tempo-map.exe where you record the playalong click track , execute a menu command and presto the job is done, all inside midi maestro.
Fourthly, this is a Not Yet/ But Soon. The next major release of Notation Composer from www.notation.com is supposed to be adding a system for synchronising the transcription to free tempo metronomeless midi files. This has been discussed on the sites forum. Its supposed to be due for release later this year. There are no details about how this will work, but I strongly suspect that this wll end up as some variant of the draggable barlines method. You could ask on the forum.
I hope that these suggestions are of some help to you.
EDIT: Oh by the way. If you have loads of money to spend, I believe that this can be doe with Cubase as well. I'd be VERY surprised if it can't.
In reply to Split Midi track to double staff by BarrieB
Hi there, am very pleased with the invaluable info above. I am going to try all the suggested methods and programs, and see how that works with my stuff. But one question. I think I know the answer, but wont know for sure unless someone tells the truth.
Tomorrow I have my piano player coming to record, so should I record data as RAW midi with RED DOT just to be sure I do catch the performance and then worry about the notational difficulties later ?
I have cubase , but dont know how to use it. It also has notation feature
In reply to Hi there, am very pleased by steve2010
Only you can decide. I'm not familiar with RED DOT. Personally I would always go for capturing the raw peformance and worry about the notation later using one of the above syncing methods. Trying to play sufficiently metronomic to directly get satisfactory notation without some post processing of the midi file is nigh on futile for more than a few bars, except for the rare musical genius's. The Able midi editor is probably the easiest method of all with its draggable barlines. I think the only problem that you might have in using it for your purposes is that with Able you are supposed to configure it for the number of beats in the bar, then you 'Tap in' the bar positions (not every beat), how well it would work with mutiple time signatures I'm not sure. With midi maestro and DIV's tempo-map.exe you give it a click on the click track for every single beat, so they should work regardless of time signature.
If you have a recent version of cubase I would recommend having a good look through the manual and help files. I'm pretty sure you will find a way of doing it in there if you look hard enough.
In reply to Only you can decide. I'm not by BarrieB
I've created pieces recorded live into MIDI If I need the notation for someone else to perform, there's clearly much editing work required. Both Sibelius and Finale can handle about any notational challenge these days, but it's again hard work. Not much more work though then doing it the old way, by hand using pen and ink.
There are audio to MIDI converters. Anything they produce will also require editing. There's no really easy way. There's always labor involved.
I applaud the work that's being done by MuseScore developers. It naturally has to be rooted in traditional periodically based twelve-tone equal temperament systems, but that it accommodates time changes, key changes, even different key signatures on different staves, graphic input into frames (well, I haven't been able to get that to work yet), ,.. and it's free!
Now if the present generation of notation software packages can't do what one wants, maybe examination of something like "Scores: An Anthology of New Music" edited by Roger Johnson can provide some ideas for alternatives.
In reply to midi notation by Norman Lowrey
The required type of tempo mapping in Cubase VST was 'I believe' done using HITPOINTS so that is what to look for in the help files . Apparently the feature was removed in cubase SX. What the current status of this capability is with cubase I don't know
In reply to As I've already mentioned, by David Bolton
This is the center of the discussion or, at least, it *should be* the center of discussion. Programming a computer means taking a complex problem, splitting it into a bunch of elementary minuscule problems, then finding a way to put together those pieces and bits via mathematic procedures in order to solve The Big One.
Whenever one starts to talk about "instinct", "feeling" and such he makes apparent that he has no idea of the process and its potentiality. Now, I'm a really poor self-taught programmer who can put together nothing but elementary low-end programs, but I know as much as neededto grasp what a huge amount of difficult work the developers of MuseScore had to face. And the challenge gets increasingly challenging as the size and complexity of the software grows with new features and nug bug fixes.
Operating a software requires some level of awareness of what's behind hitting keys on a PC keyboard and dragging a mouse to and fro. Whenever one reaches such an awareness, he can appreciate the wonders of software like MuseScore, effectively use it and ask for *realistically achievable* features.
Computers-are-not-allmighty.
Software-developers-aren't-either.
That's it.
In reply to Programming is a hard task by Aldo
Thank you for this post.
Regards,
I have found the thread and comments very useful, so thank you.
I think the method described of RAW capture of MIDI, and then modificarion of the file by tempo-map.exe. " To make a free tempo midi file sync up with the 'midi beat' what you have to do is (working with a midi sequencer) record an extra click track into the original midi file. You make this click track to follow the perceived beat of the music, (not attempting to click in strict time). You then run the midi file with the extra click track through tempo-map.exe utility specifying that the tempo map should be created from the added click track. "
Problem is I cant quite understand how to do this, and as with most things with me I get HELP/Manual dizziness. , and learn faster by physically being shown how to do something. Wish there was a You Tube tutorial :)
Steve,
You have expressed the exact same problems I have written about in other places, and wrote them again yesterday when somebody pointed me to MuseScore.
Also, it is amazing how programmers cannot seem to follow what you are suggesting and instead take offense and bring up issues that demonstrate they were not listening to you in the first place.
If you're still getting Emails sent to you from this thread, come back and let me know if the suggestions you received from BarrieB worked. I did kind of understand the suggestion about a click track being added was somehow different than the Tempo parameter that is in a MIDI file, but the resulting use of that utility has me stymied. How does that new improv Tempo become part of the MIDI file, such that Notation software can use that information? If the Tempo Map software that BarrieB mentioned is something altogether different than the Tempo parameter in a MIDI file, I hope somebody will write in here and let me know.
Good writing Steve!
In reply to Steve,You have expressed the by Toddskins
The methods I described DO work. When a sequencer records a midi file, it does so with a concept of a fixed metronomic tempo, typically 120 bpm, unless you tell it to do something different. If you were to take a simple case where the sequencer recorded at an assumed 120 bpm, and for the sake of argument and simplicity you played a piece exactly rythmically at 125 bpm, then just about every single played beat would be "off the midi beat" and importing to translate to notation would be futile. The click track methods described above basically allow you to say "forget where the midi file says the beats are, THIS CLICK is where the beat is". With appropriate algorithms it is possible to churn the midi file data and spit out a new file which plays the same as the original, but with the musical beats lining up with the midi beats. This works even with complex tempo changes in the original, accelerando, rallentando, rubato what ever. This can be accomplished because midi files do not have to have a fixed beat speed throughout. The midi file standard allows for the "instantaneous tempo" to be defined to a resolution of 1 uS per quarter note in a "tempo map", so in the above trivial example where the recording was made at 500,000 uS per quarter note, you would essentially need to modify the timebase of the midi file to 480,000 uS (=500,000*120/125) per quarter note, if you did this by itself, then the midi file would just play faster, so the note durations (defined in terms of midi ticks) have to be manipulated also to make it all come out right. By defining a click track which contains an event at each "musical beat" it is possible to determine what correction factors are required at all points in the file.
In the postings above it was mentioned that this feature was on the "coming soon" list for a programme called Notation Composer. The feature does now exist in that programme. They call this mechanism "Rebar". If you are seriously into importing midi files to try and get readable notation, I would highly recommend that you look at Notation Composer. Its not free, shame. Its declared raison d'etre is to create notation from midi files, which it does do pretty well. It allows you to create notations that will satisfy 90% of the people 90% of the time. It doesn't do "advanced stuff" like cross staff beaming and the like and really fine detail positioning, but it does do ONE HECK of a lot rather well. Get the demo and read the documentation. I have no connection with them, just a satisfied user.
In reply to Steve,You have expressed the by Toddskins
I don't think anyone took offense at your comments. I just think you are mistaken about a number of things that I tried to point out. i think the basic problem here is that there are quite a number of *completely different problems* being lumped together here. I counted four entirely different topics in that other thread, and this discussion of free tempo makes five. Different poeple are focusing on different topics at different times, and then responding to other posts as if they were talking about a different topic than they often were. This subject of dealing with free tempo actually never came up in the other thead until now.
In reply to I don't think anyone took by Marc Sabatella
Reading about your problem,makes me realise that what you need to do is fork out a fairly hefty sum of money on a highend MIDI sequencer.
Personally I use Sonar3 which fulfils my requirements, (I think they're onto Version 7 now) but you could also look at Logic and Cubase. They all have the added bonus of extensive audio recording facilities too.
They all have the ability to create variable tempo maps, and also produce a basic quantised score display without destroying the original MIDI input. You also have direct control over any MIDI controllers that you care to name. The only problem you may run into is setting up the tempo map in the first place, as the confines of MIDI recording mean that you would need to define the click track before you started to record the performance.
Oh, and I agree about notation being totally inadequate - but then it always was. How to interpret it has always been and probably always will be passed down from teacher to pupil.
HTH
Michael