Web player audio changed?
Hey.
So I recently uploaded an updated score for Queen's Don't Stop Me Now for string quartet, and the sound of the Web player has changed to the "new" sound of Musescore 2?
This really messes with the sound because I'm still using the soundfont that came with MuseScore 1.3 and in my opinion the old one sounds a LOT better, more well-defined. The new soundfont doesn't reach full volume immediately, which means that a lot of the faster notes in my score are sounding half-assed. And the so-called 'vibrato' is just ... JANKY. Yikes.
Is there any way to keep my scores using the old soundfont on the web player?
Thanks.
Comments
We are currently developing better support for soundfonts on musescore.com. For now we don't have an ETA to share, so we kindly ask you to be patience. Thanks.
In reply to We are currently developing by Thomas
No worries Thomas, as long as it plays fine on my computer. Glad you all are working on it.
Perhaps as a short term measure, you could add an option to the upload page that would allow us to select what soundfont it renders in? e.g. choose between MuseScore 1.3 and 2.0 default sound.
Just my thought.
FWIW, I have found there is a tradeoff to be made, and I don't know if there is a good way to solve this more generally. Overall, the string sounds in FluidR3 (used in 2.0) sound *infinitely* better - much more realistic - to me than the sound in TimGM6mb. And in particular, the relatively slow attack makes for a far more pleasing legato in expressive passages. But with the slower attack comes the issue you mention. It seems all soundfonts hit a different balance point - some favoring the slow attack to get a nice legato but then fast notes sound relatively weak. Others have a strong attack and thus get more clearly articulated passages but at the expense of anything resembling legato. TimGM6mb is definitely an example of the latter. Other soundfonts I have tried are way off on the other extreme - making fast passages all but inaudible.
I think FluidR3 hits a nice compromise, - I'll take the slight weakening of sound in fast passages in exchange for the far more pleasing legato. And in particular, I think your score sounds *much* better overall in FluidR3 than in TimGM6mb (I tried it both ways). But indeed, the lead line *is* noticeably weaker - the one downside.
Elsewhere, ChurchOrganist has suggested there is a solution that could be achieved internally by decoupling the "attack" from the "sustain" portion of the sound, but it's not clear to me to what extent this can actually be implemented within our framework of using FluidSynth to play sounds using SF2 files. But maybe it's more relevant to wonder about doing this using Zerberus?
Personally, I'd have assumed the answer was to have two different sets of samples for string and wind instruments - one for passages marked legato or slurred, and the other to be used in other passages.