Strings in MP3 export glitch (sound glitch)

• Sep 15, 2020 - 16:28

I exported one of my projects to MP3 and gave the song a listen. This happened more than once: the strings once and a while sound hollow. It was loud and strong than became quiet and weak. Sometimes this happens on playback too but it usually occurs in the exported file.

Is there a way to fix this for future projects?


Comments

If this problem also occurred in normal playback then this is not normal.

But for MP3 export, I can guess:
Perhaps the MP3 export bitrate is set too low.
Or, if there are excessive dynamic changes in the score, the volume may decrease because normalization is active, or there may be some clipped output if normalization is not active.

test 1; MP3 bitrate:
Edit Menu > Preferences : go to "Export" tab
test 1. (in Audio section) Change the "MP3 bitrate" to 256 // If existing value is lower than this
and Click "Apply" Button. // <= Apply button should be clickable.
And test it by exporting as MP3.

test 2; MP3 normalization:
Edit Menu > Preferences : go to "Export" tab
De-tick (or tick) the normalize checkbox. // Click it to do the opposite of its current state :)
and Click "Apply" Button. // <= Apply button should be clickable.
And test it again by exporting as MP3.

In reply to by Ziya Mete Demircan

Unfortunately, both tests didn't work.

" the volume may decrease because normalization is active"
What I meant was, in the piece, sometimes the violoncellos suddenly have a hollow sound (it's on and off), which isn't normal. Every other instrument is fine, mainly the violoncellos.

In reply to by Nahti

I recently saw a score where hairpins were essentially causing the dynamics on one staff to be ignored. It was a piano score so this may or may not be related. Here is the gist of what I saw.

There were two consectutive measures. The fist measure started at a dynamic like pp with a hairpin crescendo covering the entire measure. There was an f at the beginning of the next measure with a hairpin decrescendo covering the entire measure ending with a p at the start of the next measure. This caused the notes in one piano voice to not respond to the hairpins. When we adjusted the start of the first hairpin so it's anchor was not on beat one, the dynamics issue was eliminated. What I suggest that you try is to put rests in voice 2 (or another unused voice) with a 16th note on the beat immediately after the dynamicthen double click the hairpin, select the left drag box and press alt+right arrow to change the anchor to the 16th note you entered.

If you don't understand this (it's a bit complicated) then post your .mscz file and tell us a place or two where you can't hear the cellos during playback.

My guess, without seeing the actual score, is that you have a unison - two or more copies of the same note using the same sound at the same time (either in different voices or different staves). This creates a well-known audio artifact common to pretty much any synthesis system, it's kind of inherent in the laws of physics. Workarounds are to slightly offset one note in time or pitch, etc.

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