V. 2.0 Yamaha Grand Piano
The V.2.0 Yamaha Grand Piano sounds like a harpsichord. Is there some adjustment I'm missing to make it sound like a piano?
Do you still have an unanswered question? Please log in first to post your question.
The V.2.0 Yamaha Grand Piano sounds like a harpsichord. Is there some adjustment I'm missing to make it sound like a piano?
Do you still have an unanswered question? Please log in first to post your question.
Comments
It is indeed very bright, but no harpsichord I have ever come across sounds like that :)
I hope you are aware that MuseScore 2 can load or play multiple soundfonts, so if you don't like a particular default sound you can source one which meets with your approval and run it together with the default.
I have noted your concern though, this maybe a candidate for some tweaking of the filters :)
In reply to It is indeed very bright, but by ChurchOrganist
Well, no piano I have ever come across sounds like that. I can't believe Yamaha let them put their name on it. In V.1.3 at least you could choose between various piano "sounds" with the mixer. But in V.2.0 you're saying there's only the "Yamaha" sound and the User has to go out into the esoteric world of sound fonts and through trial and error maybe find one that has a better piano sound. I'm too dumb to tackle that. Thanks, but I think I'll just stay with V.1.3.
In reply to Well, no piano I have ever by hmscomp
Wait—now we may have a bug. You should have lots of alternatives in the Mixer:
![Piano mixer.png]( "Piano mixer.png")
Maybe there's something else wrong, too, which accounts for the sound coming out like a harpsichord. It should be fairly similar to the default sound from 1.3, except fuller.
Just because this is always the first step in figuring out a problem, and usually it fixes things, could you try resetting MuseScore 2 to factory settings? See Revert to factory settings . See what happens. If it doesn't get you back the full set of options in the Mixer (also, try creating a new piano score, to rule out the possibility that there's a problem with your file), report the issue to the "Support and bug reports" forum and we'll try to figure out what's going on.
Also—ChurchOrganist said "if you don't like a particular default sound you can source one which meets with your approval". You already have one that you like—the one that came with 1.3! You can use the sounds from 1.3 in 2.0. See SoundFont for details.
In reply to Well, no piano I have ever by hmscomp
Simply add the TimGM6mb.sf2 soundfont to the "Soundfont folder" in MuseScore 2.0 - see step #1 here:
http://musescore.org/en/node/50721
This will give you the Piano 1, Piano 2, Piano 3, Honky Tonk, E. Piano 1, and E.Piano 2 sounds.
The TimGM6mb.sf2 soundfont can be had here:
http://musescore.org/en/node/36171#list
(fifth in the list - no need to unzip)
Regards.
In reply to Well, no piano I have ever by hmscomp
You can still choose between different sounds in the mixer; this hasn't changed. And in general, the Yamaha Grand sound in 2.0, while definitely bright, should still be a very noticeable improvement over the much less realistic sound in 1.3. One other poster had an installation problem or something where for some reason, his piano really was replaced by a harpsichord or some other sound. The bug went away on a factory reset (run with "-F" option on command line. So maybe this is happening to you too? For reference, the piano sound should be as demonstrated here:
http://musescore.org/en/node/20818
Hearing the same piece back to back in three different soundfonts, to me it's pretty clear the 1.3 soundfont is by far the worst of three - sounding very electronic, twangy, and nothing like a piano. The FluidR3 version - the default for 2.0 - is a very realistic rendition of a rather bright piano. The Salamander version sounds best overall, although that soundfont *only* does piano..
In any case, as mentioned, installing a new soundfont is easy, not "esoteric" at all. Trust me, you really don't want to give up the amazing improvements in 2.0 - not to mention, cut yourself from the abiltiy to load scores created by others.
In reply to You can still choose between by Marc Sabatella
Marc, I don't want to beat a dead horse when there's absolutely no need - there are choices, after all, with which everyone can be happy now - but I continue to disagree vigorously with your characterization of the piano sounds in the respective soundfonts. I would suggest that it's all in the ear of the beholder, and thus a matter of opinion.
Instead, you continue to treat the issue as a matter of fact when you claim that the 1.3 soundfont's piano sounds are 'nothing like a piano' while 2.0's is a 'very realistic rendition of a rather bright piano'. That is just not true as a matter of fact - it's subjective - and, as with any such matter of opinion, we should all just agree to disagree about it.
In reply to Marc, I don't want to beat a by [DELETED] 448831
it is indeed at least partially subjective. But there are objectively measurable components as well - the amount of raw data present in the sample set. Simply put, more data *does* equal better when it comes to ability to faithfully recreate sound. A WAV file recorded at 48 Khz / 24 bits *is* objectively better than one recorded at 22 Khz / 8 bits, for example. Or, a soundfont built from a single sample of a single note thast is then scaled cannot possibly match in realism a soundfont built from, say, 5 different sample of each of the 88 notes. There will be objectively measurable artifacts from the scaling of limited data compared to a soundfont that actually contains more data.
What is subjective is that more data, while it might mean a more faithful recreation, doesn't necessarily mean better overall. In particular, a great recording of a mediocre piano might sound better to some people than a mediocre recording of a great piano. Beyond that, there is also the subejctivity in what constitutes a great piano, also difference in microphone quality, recording techniques, etc. So no, I full agree there is a subjective element, but hoepfully yuou will also agree there is a very obejctive element here too.
Anyhow I think pretty much everyone would agree a great recording of a great piano beats a mediocre recording of a great piano *or* a great recording of a mediocre piano. So if you don't like the piano that was recorded for FluidR3, or something about the micing technique or whatever, that's fine. But still, given that almost every other GM soundfont provides more raw data - objectively, measurably more data - than TimGM6mb, it's natural to assume that the vast majority of people would find many of those soundfonts preferable overall.
So again, I implore people to check it out for thermselves and try different alternatives. Even if some people subjective find the TimGM6mb piano better than FluidR3, it's still *extremely* likely that if they tried out other soundfonts, they would find most of them *far* superior to TimGM6mb. There really is on objective basis for this assumption, and it is all about quantity of data.
In reply to it is indeed at least by Marc Sabatella
Gotta love the passion in these forums.
To add my 2¢...
I know pretty much everyone would agree a great live performance on a great piano beats a great recording of a great piano; a mediocre recording of a great piano; or a great recording of a mediocre piano.
MuseScore, after all, is a notation app. (Where have I heard that before?)
:-)
Regards to all.
In reply to it is indeed at least by Marc Sabatella
I completely support everything you've written in this comment, Marc - one can't argue with facts! - and I would also like to re-post the link you referenced where at least the piano sounds in question can be compared.
http://musescore.org/en/node/20818: New audio output architecture for MuseScore 2.0
Even though that discussion thread began two years ago, I think it is very useful and valuable for the wealth of information and insight it contains. I would also offer the caveat that a soundfont contains a multitude of sounds/instruments, and piano isn't the only criterion (and possibly not the 'best' one) for judging the quality or suitability of a given soundfont - but I think most people with an interest in the topic of audio output architecture already know that. :-)
Of course, the backdrop here is that MuseScore is primarily a notation program - even as many users are simultaneously quite concerned with playback issues. Everyone who shares that concern really should investigate this area to get acquainted with options that are out there, and I applaud your effort to encourage that!
It pains me to think that anyone would turn away from MuseScore 2.0 - a truly brilliant achievement that was accomplished by the dedication of so many hours by so many people - for the sole reason of not liking one or more of the default sounds when there are other choices so readily available.
In reply to Well, no piano I have ever by hmscomp
I'm a pianist, and I use MuseScore almost exclusively to re-engrave 19th century scores for piano solo.
I too thought that the Yamaha Grand Piano sound in 2.0's default soundfont was horrible - and I still do. I expressed that opinion very early on here, but (1) there was no consensus of agreement with me, (2) the majority viewpoint was that the new soundfont was generally far superior to that of 1.3, and (3) it didn't really affect me as a practical matter because of the ability to continue to use the default soundfont from 1.3 if I so chose ... and that's exactly what I've done.
I hope you'll do the same! MuseScore 2.0 is far too good a program - and way too big an improvement over 1.3 in some very important functionalities - to forego using it because of this specific and limited soundfont issue with piano sounds, especially when you can continue using the same soundfont you've always used.
In reply to I'm a pianist, and I use by [DELETED] 448831
Hmm, if you don't like the piano sound in FluidR3, I'd still implore you to try other alternatives. MsueScore is capable of sounding *so* much better than TimGM6mb. There are plenty of *far* better soundfonts out there. The only reaosn TimGM6mb was ever used bis because it was so small, making for small downloads. But virtually everty soundfont every created is bigger and will sound correspondingly better.
In reply to Hmm, if you don't like the by Marc Sabatella
OK guys, problem solved. I took jm6stringer's advice and added the old TimGm6mb.sf2 SoundFont to the synthesizer and set it as the default. I had thought about doing that but really didn't know how (everything is "esoteric" until you get some help doing it for the first time). It sounds even better in V.2.0, as if this Acoustic Grand Piano was being played in a recital hall with really good reverberation. Sorry Marc, to my ear it beats the pants off the "bright" (harpsichordish) Yamaha Grand Piano sound.
I have some other issues with playing scores in V.2.0 that were made with V.1.3. Should I continue those in this dialogue or start a new topic?
In reply to OK guys, problem solved. I by hmscomp
My big question is about the options in the Mixer. Did you really not have any alternatives in the dropdown menu? My screenshot above shows what it should look like—literally dozens of options. This could be evidence of a real bug, which might also have actually caused the piano to play back with the harpsichord sound.
The "other issues with playing scores in V.2.0": are they in the category of "something sounds off" (and fixed by using a different SoundFont), or something else?
In reply to My big question is about the by Isaac Weiss
I got the same drop down menu, but all the other piano options were definitely not what I was looking for.
The other issues have to do with formatting (number of measures in a system working in V.1.3 and not in V.2.0 without an Add Less Stretch modification, which may or may not work without an ADDITIONAL modification making the Scaling even smaller or the page bigger, and the "Can Not Open-File Corrupted" warning which always seems to be caused by a measure modification that added an extra beat which was perfectly OK in V.1.3 but apparently not in V.2.0. I don't want to have to re-work all my scores if at all possible.
In reply to I got the same drop down by hmscomp
Regarding 'stretch', have you tried tweaking the setting in Style > General... > Measure > Spacing to something a bit tighter? Even a very small adjustment there can make a difference - perhaps, if necessary, in conjunction with slight changes to some of the other settings in the Measure dialog box.
I would only decrease the Scaling in Layout > Page Settings... as a last resort, as it impacts nearly every element in the score instead of allowing you to choose some selective adjustments only where you want.
In reply to Regarding 'stretch', have you by [DELETED] 448831
Why are any adjustments required at all? The scores were perfectly laid out in V.1.3, so why are they opening with these issues in V.2.0. I am aware of the Ignore option and the Show Details option sites the adjusted measures as the reason for the "corruption". BTW, what's the new term for "accolade distance"? I don't see it in Style/General/Page anymore.
In reply to Why are any adjustments by hmscomp
Accolade = obscure term no one had heard of; "grand staff" is the more familair term used in 2.0 :-)
The reason adjustments are occasionally necessary is that 1.3 had some pretty serious layout deficiencies where it was not spacing things properly - not allowing sufficient space in some places, too much space in others, and collisions that never should have happened in others. 2.0 is hugely improved when it comes to spacing - and, yes, objectively so (we simply weren't followed standard engraving practice in a lot of situations). But "better" does indeed imply "different". The improvements mean that some measures might be different sizes than they were in 1.3, which in turn might sometimes mean a different number of measures fit on a system than before.
In general, the new default should simply be *better* in all cases, but if you were tweaking stretch and using line breaks and so forth to override 1.3's choices and enforce your own, those same adjustments won't necessarily make sense any more. So usually best answer will be to remove all manual stretch adjustments and line breaks, see if there is anything you are unhappy with in the new defaults, and adjust as necessary from there. As opposed to simply trying to emulate the layout choices made in 1.3, or the adjsutments you made based on whatever flaws were present in 1.3.
In reply to Why are any adjustments by hmscomp
'Accolade distance' became 'Grand staff distance', which I believe is far more appropriate and intuitive for English speakers.
I apologize in advance if the following is something you already know and have already considered ... but, given that 1.3 and 2.0 can co-exist peaceably, there's no compelling reason at all to convert 1.3 scores to 2.0 if they are 'perfectly laid out' and, for all intents and purposes, projects that are finished or near completion. It's well known, well documented and taken for granted that they will not open in 2.0 in exactly the same layout as in 1.3, though the precise reasons therefor are best left to developers to explain.
My advice is not to convert scores unless they are in the midst of ongoing work that will benefit from the many new functionalities available in 2.0 to such a degree that this consideration outweighs the occasional fine tuning that will be inevitably necessary to restore the exact 'look' that you wish to preserve.
In reply to I got the same drop down by hmscomp
The number of measures per system thing is not necessarily a problem per se—in fact the idea is that the layout should be better in 2.0. But if you strongly prefer the way 1.3 laid out a particular score, between using line breaks (click a barline in the middle of the page and hit Enter) and decreasing stretch (select measures that you want to fit on one line and hit "{" a few times), no changes to Page Settings should be necessary.
As for the "File corrupted" warning, that shouldn't be a "Can Not Open" issue. In fact, there should be an "Ignore" option to open the score regardless. And a measure that you deliberately changed the duration of in Measure Properties is not a corruption and 2.0 shouldn't mistake it for one.
In reply to OK guys, problem solved. I by hmscomp
Separate issues deserve a separate thread.
Now that you have figured out how to install a soundfont, I would again encourage to try some more option. TimGM6mb is, as I mentioned, pretty much the smallest - and hence the least amount of data / detail to the sounds - GM soundfont there is. That is why it was chosen for 1.3 - to reduce download bandwidth. But I really think if you try even a couple other soundfonts, you will realize it is possible to do much, much, much better. GeneralUSerGS is one a lot of people like, and I personally think TimbreOfHeaven is perhaps the best overall. For piano specifically, there is Salamander, but it is *huge* and might not work so well on older computers. Plus it is installed a little differently since it is a different format. Anyhow, I think you find an hour or two experiemtning well worth your time lpng term!
BTW, a coule of other subjective elements worth considering:
- Speaker quality. Typical computer speakers have no low end to speak of. The FluidR# has more high end to it, no doubt. But it *also* has a lot more low end, and through better speakers or decent headphones, these balance better. The overall effect should bright but rich - more frequency content at *both* ends of the spectrum than TimGM6mb. Whereas through cheap computer speakers, you won't hear the lows, only the highs - leading you to to think it is brighter than it in fact is. On the other hand, get speakers that are better still, and now maybe the highs start to shine through more, and any flaws become more apparent. On the other other hand, the flaws in TimGM6mb become more apparent too. As I said, even if you don't like the brighter sound of FluidR3 piano, I still have to believe if you do the comparison, you will find virtually every other soundfont out there better than TimGM6mb, simply because there is no enough information in that soundfont to come close to approaching the relaism offered by other soundfonts.
- Musical context. If you are mostly dealing with solo piano, and mostly classical at that, then a certain darkness is traditional and expected. However, that same sound might not cut through well at all in a rock band or jazz big band. So that is a reason why for many purposes, the brighter sound of FluidR3 are a better fit for many people.
My impression is that a physical Yamaha grand piano has a bright ringing sound, compared to some other pianos. I found it surprising and somewhat unpleasant the first time I played one.
I have just tweaked Frank Wen's original Yamaha Grand to provide a mellower version for those who find the existing grand piano sound excessively bright.
It is currently in beta and obtainable as a separate soundfont from the soundfonts forum......
https://musescore.org/en/node/55096
In reply to I have just tweaked Frank by ChurchOrganist
Two questions:
1. Is the soundfont in beta just the tweaked piano sound or is it the entire FluidR3 with the mellower Yamaha piano?
2. What's the estimated time for it to be in beta before it is perhaps incorporated into the built-in FluidR3 soundfont of V.2.0?
Thank you for your time and effort in addressing this issue.
In reply to Two questions: 1. Is the by hmscomp
1. It is just the tweaked piano sound
2. As long is it takes to be sure there are no underlying issues with it :)
In reply to I have just tweaked Frank by ChurchOrganist
THANK YOU SO MUCH... I feel this is the solution to a lot of the original posters (and mine's) problem. This is so close to an actual piano sound, it's scary; and a good exercise for me in discovering how to install soundfonts. :)
In reply to THANK YOU SO MUCH... I feel by andrewjleggett
Actually, as of MuseScore 2.0.1, there's no need to install it separately; it's included in the default SoundFont as an alternative. Just open the Mixer, go to the dropdown menu and start typing "mell" to locate it.
In reply to Actually, as of MuseScore by Isaac Weiss
Wait, what? I am definitely not finding it in my 2.0.1. :-(
EDIT: Never mind - it's there. And ... I prefer Piano 1 in TimGM6mb.sf2! FWIW, I have very nice Advent external computer speakers. All my sounds assessments have been based on them, as I never use headphones or earbuds. I am sure I am an outlier, but I'm thankful to have options. :-)
In reply to Actually, as of MuseScore by Isaac Weiss
When I type "mell" in the mixer dropdown menu, I get an instrument that starts with an "m", then an instrument that starts with an "e", etc. If I type "mell" in the instrument search box, I get a "mellaphone". Is the default soundfont in V.2.0.1 still FluidR3Mono_GM.sf3? If so, what am I doing wrong?
In reply to When I type "mell" in the by hmscomp
The "Mellow Yamaha Grand" is a sound from the Mixer, not a unique instrument. It sounds like you're typing too slowly in the Mixer dropdown. Try setting up your fingers first and then typing "mell" quickly.
In reply to The "Mellow Yamaha Grand" is by Isaac Weiss
Typing quickly worked. Thank you.