Since I updated from 3.0.5 to 3.1.0 the waveform/signal line is thick, even when there is nor signal running into Audacity.
I already tried changing the dB range that audacity shows on top, but the waveform line is still thick.
This is how it used to be with all older Audacity versions:
This is now with Audacity 3.1.0:
Any Ideas on how to fix this?
It does not matter, if I denoise my mic to the point where no signal is produced when quiet, or if I mute it…the db Meter on Top is showing absolutely nothing (showing down to -60dB) but the waveform shows the constant line so thick as if there was a constant signal active with 3.1.0.
I also noted this, when passing from version 3.0.2 to 3.1.2.
May it have anything with the change on Audacity’s built in volume control?
(now it changes its own volume instead of system’s volume)
It happens when recording from system’s “stereo mix”. It used to be completly silent.
In the spectrogram view it’s clear to see that even no signal is producing the thick line in waveform view.
I’ll stick with the old audacity version, because I can easily identify if there is any noice on my mic signal with the visual behavior in waveform view of the older version.
OK so I dis some testing - first three images are with 3.1.2 second two are with 3.0.5
this is “silence” recorded with 3.1.2 - and yes the line is thick (thicker than 3.0.5 - see image #4 below)
I thun used Ctrl+L to silence the section from 5-7 seconds - and yes that shows as true silence
I then amplified the clip - and we can see it’s not silence at all
so I switched to 3.0.5 to repeat thee experiment - and yes we see a nice thin “silent line”
I then amplified this - and we can see the recorded sigan is not silence in 3.0.5 either
for completeness I undid the amplify, silenced from 305 secs and then redid amplify
So yes that appears to be a slight visual difference from 3.0.5 to 3.1.x - but it may be that 3.1.x is swing a truer pictur and when quiiescent recording either my sound could is making a little “noise” - or it’s picking up “room tone”.
Here is some noise with an RMS level of around -50 dB in Audacity 3.0.2
Here is the same track with the track height reduced sightly:
Here is the same audio (from 32-bit float WAV) in Audacity 3.1.2:
and finally, the same audio in Audacity 3.1.2 with the track size slightly expanded.
At around -45 dB peak / -50 dB RMS, with the default track height, the waveform is on the threshold between one pixel high and two pixels high.
Audacity does not currently have sub-pixel rendering, so audio at this level may display as 1 or 2 pixels, depending on whether the rendering considers it to be closer to one pixel or two.
If there is sufficient track height, then the vertical height becomes closer to three pixels, and displays like this:
You are right, It is because the 3.1.2 is adding the title at the top. That makes the waveform from 3.1.2 slightly smaller so therefor should be less noticeable. And what is the difference on the other examples that are on the same screen? And why do those exported clips show up fine in 3.0.5?
The way that audio samples are translated to pixels is extremely complex. It is not just a simple mapping of “sample value => pixel position”. At normal zoom levels each pixels represents many samples, and if there is room to do so will display both the peak level (dark blue) and the RMS level (light blue). The code also needs to be as fast as possible so that you are not waiting longer than necessary while Audacity redraws the waveform. If you resize the track vertically, you may observe that Audacity redraws the waveform to the new scale extremely quickly, even when there are millions of samples within the view.
If you silence a selection of audio (Ctrl + L) then you should see it as a thin (1 px) line. Is that what you see?
Noise with a peak amplitude of around -40 dB may show as 1 px or 2 px - it’s a borderline case and may do either depending on the track height, though with a very big track height it may show as 2, 3 or even 4 px high.
The “steps” in this image are simply an artifact of the pixel display, the actual audio fade is completely smooth (same audio in both screenshots):
So, I uninstalled 3.1.2 and reinstalled 3.0.2.
After recording silence from stereo mix, and amplifying (with both versions), the difference seemed to be only visual, still not sure, but seems that I mistook by worrying about this.
I hate to keep harping on this, but it is an issue that did not show up until 3.1.0 and I think, no I believe it had to do with the addition of the clip handle and here’s why.
As I stated before I did not see the issue in 3.0.5 or the older 2.42 versions. So, in version 3.1.3 I grabbed the bottom of the wave form to enlarge it vertically and when it enlarged by (I assume) one grid level the thicker wave line disappeared or returned to normal. So, I pulled it down one more level and the thicker line reappeared. This kept repeating as I enlarged the vertical view by one grid level at a time.
If you look at the left track at 0.0 you will notice the thickness of the line beside it.
Same track enlarged vertically by one gid level.
When there is a track loaded, it distorts the look of the fade-in of the track.
But when you enlarge it vertically by one grid line, the wave line looks normal.
So back to my original comment about it being related to when the clip handles were added, it changed how the default wave landed on the grid. I went back to versions 3.0.5 and 2.4.2 that did not exhibit this issue and enlarge the same track by one grid level at a time, and the silent lead-in wave enlarged the same way in both of those versions. That is what led me to believe it was the addition of the clip handles in version 3.1.0 and later.
Now with all of that being said, is there a way I could set a new default vertical height for when Audacity is initially opened so I don’t have to adjust the vertical height every time I open Audacity or drop a new track on it.