Using a DAW to make Musescore scores sound better

• Mar 11, 2021 - 18:00

I've been using Musescore for a while now and have been making music for orchestras and bands alike, but I've run into the problem (I think other people are having) that my scores sound like a computer. I've tried "humanizing" pieces with the TempoChanges plug-in and custom soundfonts, but to little to no avail. I want to take my scores a step further and improve the sound of my scores by moving the tracks into a digital audio workstation (DAW) but I'm not sure where to start. Any help would be greatly appreciated and I will take all opinions into consideration.

(As a side note, I have tried composing in Finale, but I prefer Musescore's style of engraving because I feel it is more efficient. I know Musescore's reluctance to support VSTs (I.e. NotePerformer, Kontakt, etc.) within the program, so I'm looking elsewhere for better sound quality - namely within a DAW or similar.)


Meanwhile, you should be able to route MuseScore's MIDI playback to most DAW software either by using direct MIDI out (if the DAW creates appropriate device drivers for this) or by installing, configuring, and using JACK.

I've done exporting Musescore MIDI to DAW, too. There are lot of details, for example, there is a procedure for MuseScore to generate Midi CC11 for "expression" waveform, for VSTs that support it.

That said, I've always been very impressed by what MuseScore can do in making your playback amazing (for a free open source notation software)!

In addition to tempo changes you mentioned, I would try few more things on MuseScore:

  • Use MuseScore built-in Expressive Soundfonts (EXP suffix) and Single Note Dynamics
  • Notate lots of dynamics (mf,pp,ff...) and Hairpins (<, >) in your score
  • Add a little reverb on synth (View->Synthewizer->Master Effects), pan your instruments slightly (F10)
  • Use lots of articulations for your specific instruments (marcato, staccato pizz), grace notes, gliss (portamento)

Here are some examples, using MuseScore & MuseScore-Soundfonts only (no DAW):

In reply to by Shaw Takeuchi

I have to agree that I'm impressed with Musescore's current features as a free program, and I've been doing many of the things you listed below, but there's only so much I can do to humanize it. While these features are pretty decent at "exploiting" the sound, I feel I can further improve my scores through external software.

Here's an example of the aforementioned recommendations I use

In reply to by Kaden Dayog v1

Firstly, this arrangement sounds very cool! That said, it's a difficult piece to discuss expression because this segment seems to be an impressionism of "musical box" or "clock" -- so it is supposed to be mechanical, not human expressive.

Unfortunately MuseScore soundfont (I think) is only 16-bits, it does lack dynamics and fidelity, not to mention expression control. With DAW you are looking for not just the instrument sound, but all sorts of mixing/mastering tools.

One gentleman I follow on mentioned he uses MuseScore to compose, export musicxml to Sibelius, then to Noteperformer for audio playback. I haven't the time to look into this method, but this "AI technology" seems very interesting to me, and less work for us! May be you can check out his work.

In reply to by Shaw Takeuchi

16-bit samples (96dB) are enough for playback and human ear.
There are no microphones capable of recording above 20-22 kHz in any studio. There is no musical instrument that goes that high . 44.1kHz /2 = 22.05kHz <== It's above the hearing limit of the human ear.

24-bit recordings (140 dB) are used in recording-studios for both distortion free and working in a higher dynamic range and headroom (so as not to worry too much about the audio peaks).
About playback: The sensitivity of 100watt high-end speakers has been measured as 86dB (for 1 watt).

Normally (in my opinion) anything beyond 16-bit 44.1khz for playback is a waste of data.

In reply to by Kaden Dayog v1

16 bit encoding does not lack dynamics: a 96 dB range is huge. Lack of fidelity is linked to the quality of samples, the variety of samples and the ability to apply them to a score. I'm generally impressed with Musescore's playback although I have a few 'improved' presets in my custom soundfont. For me personal!y, I think that using a DAW would be too time consuming unless MS4 can somehow automate the process.

For a really interesting article on 16/44.1 vs 24/48, and higher resolution systems have a read of:…

You can even download the files and judge for yourself.

In reply to by Shaw Takeuchi

OK, 'musical detail' is a different thing and related more to the digital sampling rate.

16-bit / 44.1 khz is the standard used to make audio CDs so, at least in theory, Musescore playback could reach this level of musical detail if the instrument samples are at 44.1 khz. This rate is common in soundfont samples and gives an upper frequency limit of 22.05 khz to the sound produced, which is above the audible range for us humans. Some soundfonts use 22.05 khz sampling, (for smaller files), and here the upper playback limit would be 10.025 khz. This may be detectable in the sound as a lack of fine detail – but I'm not sure.

Kaden -

Just to throw you a curve-ball... I found this video recently, which will blow your mind! The audio sounds "human" because it is (using midi controller w/ after-touch and expression):

There are middle-ground techniques like "sweetening" -- where you use your exported MIDI or DAW-produced audio to play the background parts, and you hire a real human soloists (violin, piano, etc.) to record.

In reply to by Shaw Takeuchi

Well, everyone's idea of a good sound is different. For an orchestral library that cost $1000, I would hope that the owner would have an idea of what Baroque music is all about. No evidence here.

There is a false god of a human or realistic sound. It is remarkable thing that we can sit at our computers and create music. But there is only one thing that sounds like real players.

In reply to by bobjp

I've been thinking this might be a good forum "group". I looked around but I was not able to find any groups on that focus on:

  • How to make Musecore playback sound as well ('human') as possible
  • Techniques/workflow for porting Musescore MIDI to DAWs/VSTs (I've ran into several challenges)
  • Discussion on soundfonts

Anyone know of a such group on

In reply to by Shaw Takeuchi

To be clear, I spend a lot of time tweaking MuseScore to sound as good as possible. There are indeed many little tricks that can be employed.

Personally, I think there is a lot of confusion around what good sound is. There are those that try to stagger starting times of some notes to try and mimic real players. After all, real people don't play together exactly on the beat all the time. So if the timing is off a tad now and then (or even slightly off pitch) then the recording will sound more Human.

I try to make MS playback as Musical as possible. That is the job of any conductor, band, orchestra, musician, or singer. The goal is beautiful, expressive, powerful, flowing, staggeringly poignant music. Simple, yes? Humans have limitations.

Then there is the problem of what fonts work well with MuseScore.

In reply to by Shaw Takeuchi

Nope but that's why I compose and arrange a ton of music for my strings and percussion groups at school. I also have Cubase to access the VST's. I've experimented with the master effects (zita and sc4) and improved the sound but would like more info on those effects on Musescore to make the sounds even better. The trumpet sucks and you can't do a suspended cymbal roll. So, since I am a trumpet player and percussionist, I'm thinking about playing and recording those parts myself so I can get a good sound.

Here's the link to my young band song I composed and have up on JW Pepper. It sounds a little to "synthie" to me. I'm going to run it through Cubase and see what I can do. My strings pieces sound decent and so do the percussion pieces.

In reply to by Christian Paulson

Christian - LOL yes I use SC4 for "hearing protection" during the full orchestra climax (I'm very serious about this!) The parameters are generic (common in industry) -- if you know how to use the compressor for Cubase your knowledge will port well to configuring SC4.

Your band song sounds pretty good, actually! Perhaps consider adding a little reverb, and there is a Musescore trick I use that will make your drum rolls (trems or 'repeating notes') sounds a little better. I wrote series of quick tips which you may want to check out:

I shifted my focus to transcribing on MuseScore for my DAW (instead of properly engraved PDF printed scores for musician performers). Bottom line is, for audio end products I need more than "better sound font" -- things like powerful mixing bus, EQ, compression, reverbs/chorus/delays, limiters, automation, just to name a few.

My goal is to make my programmed music sound "more human than human"!

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