please help.
i'm trying to analyse some recordings with audacity and am really being hampered by the lack of labelling. when you import a file you get the long window with the sound represented in it (forgive my lack of vocab but i am still finding my feet a little in this stuff), there is a + and - scale on the far left, just right of the bit that tells you 32bit float etc. what is that scale of? i can't find a label or explanation of it anywhere.
unlabelled axis
-
archaeologist
- Posts: 5
- Joined: Tue Sep 04, 2007 11:50 am
- Operating System: Please select
-
kozikowski
- Forum Staff
- Posts: 69365
- Joined: Thu Aug 02, 2007 5:57 pm
- Operating System: macOS 10.13 High Sierra
Re: unlabelled axis
See the little teensy downward-facing black arrow to the left of the waveform. Click it and select Waveform (dB).
Then grab ahold of the bottom of the audacity windows and pull down. the taller the window is, the more accurate those dB number become.
Indeed, percent is a way of measuring video, not audio. Everybody gets that wrong. Everybody expects "50%" to be half volume. It's not. Generally half-volume is -18 dB or half three times (half or double voltage every 6).
Audio doesn't act like you're expecting. That's what makes this so darn much fun.
Also, if you're doing sound analysis, I don't know if I would be using floating point encoding. That tends to change the number scheme around depending on how loud things are. I'd be using 48 KHz, 16 bit, which is the video standard. It's a actually a bit better than CD quality and you have to downgrade the audio to make a CD. I don't remember if Audacity supports 24 KHz, but that's another way to get a "clean" recording with no magic. It depends on what you're trying to do.
Koz
Then grab ahold of the bottom of the audacity windows and pull down. the taller the window is, the more accurate those dB number become.
Indeed, percent is a way of measuring video, not audio. Everybody gets that wrong. Everybody expects "50%" to be half volume. It's not. Generally half-volume is -18 dB or half three times (half or double voltage every 6).
Audio doesn't act like you're expecting. That's what makes this so darn much fun.
Also, if you're doing sound analysis, I don't know if I would be using floating point encoding. That tends to change the number scheme around depending on how loud things are. I'd be using 48 KHz, 16 bit, which is the video standard. It's a actually a bit better than CD quality and you have to downgrade the audio to make a CD. I don't remember if Audacity supports 24 KHz, but that's another way to get a "clean" recording with no magic. It depends on what you're trying to do.
Koz
-
archaeologist
- Posts: 5
- Joined: Tue Sep 04, 2007 11:50 am
- Operating System: Please select
Re: unlabelled axis
thankyou, that was so helpful. you are a liesaver.