mono (L) and mono mix conversions to stereo (or L and R)
Dear Forum friends,
I record single speakers (stories) with an AT2020 external mic (mono) to a H4N Zoom recorder, 48 Khz/24bits waw files. I need the highest quality possible (but no need to increase hz and bits) for the archiving of my files. I must make the audio files stereo (or L and R, practically the same in this case) to be able to listen to them with both ears in closed head sets.
Q1. I handle the files like this 1) split stereo (file menu) 2) close right empty channel (file menu) 3) Duplicate (Edit) 4) Export file as L and R channels. Or 4) make stereo track (I think the result must be the same) 5 Export.
Why do the files not increase in size after I have duplicated a channel (and exported), 48Khz/24bit)?
When I applied the same procedure to L channel files made from the same recorder and an ext. Shure 058 mic (mono), the file size increased 30%. Why?
Q2. Same equipiment, settings, except I mono mix directly in the recorder. Because one channel is empty I thought that I would get two channels that are the same as if I recorded mono. It is not the case, the channels are the same, but the signal diagram shows shorter spikes. So this was a mistake. But does it have any significance? I wonder if I can gain anything if I make stereo from my mono mix recordings? I did it, with the same procedures as above, and the resulting file increased by 30 %. Any gain? Volume?
Q3. Why do Audacity open my 24 bit files as 32 float? Default (pref) is 24 bit. What happens when I return the 24bit file to 24 bit? Any loss?
Q3. Why do Audacity open my 24 bit files as 32 float? Default (pref) is 24 bit. What happens when I return the 24bit file to 24 bit? Any loss?
Audacity always works internally at 32-bit floating in order to avoid damage from effects and filters. It’s possible to overload 32-float (go up beyond 0dB) and recover by simply turning the volume down. In the fixed formats, the show would be permanently trashed. Yes, there is always a very tiny error when the output file is created. The output dither settings affect how the output file is created. The default is lesser acoustic error.
This process regularly kills scientists who depend on bit-level accuracy for experiments. Audacity is a show editor not a WAV editor.
“Mono” sound files always play to both speakers/headphones. It’s not necessary to “convert” to stereo to get left and right presentation.
Try it. Create or open a mono show. Make sure the panel to the left of the track says Mono, or use the pulldown to force it. Export to the format of your choice and I bet it plays to both left and right.
As koz says - your recording are mono, so best to work in mono throughout.
Also best to keep the “Quality” setting as 32 bit float. When exporting to 24 bit, the recording will be “perfect” down to about -120 dB (even “state of the art” 24 bit DACs produce only random noise in the last couple of bits, so better than 120 dB is not achievable).
I tried making mono of a Left channel (stereo in Audacity) file, and noticed that the file size was cut to 50 %. So there must be loss. That is why I make two channels (and or stereo after that, which is the same actually. Perhaps L and R does not play in a mono speaker, like on computers, just one, then Stereo or Mono must be better.
No more ideas about recording mono mix on recorder, compared to making L and R from L (stereo), Stereo or Mono?
For uncompressed file types such as WAV, each audio sample is saved and occupies the same amount of disk space as every other sample. Thus a 2 channel recording with audio in the left channel and silence in the right channel will be double the size of a mono file because the silent channel has exactly the same number of audio samples as the audio channel. Recording mono audio as a 2 channel (stereo) file wastes space.
So I guess the wisest thing for me to do is to set the audio recorder (H4N, to which the mono AT2020 is connected XLR to one channel) to Mono. Done, if there is no win at all with copying the left channel.
PS. 16 bit files opens as 24 bits in Audacity. My annotation software did not accept the 32 bit float, so problem solved as well. Must 24 bits. And a 32 bit file is double the size of a 16 bit. Just learned! That explains another question posted above. Amateur…
Remains the question why the waw diagram is smaller on my mono mix recordings than the mono (Left channel) recordings that result by default from the H4N recorder? That was what made me fear I was wrong to use mono mix. At the moment I believe mono mix is what I should use.
My educated guess is that “monomix” is recording the average of the left and right inputs. As the recorder only has one (mono) mic plugged in, one input channel will have the mic recording and the other channel will be almost silent (just a tiny bit of noise). Averaging the two input channels will therefore produce a mono recording that is half amplitude (half height).
Does the recorder allow you to record just one (mono) channel (the channel that has the mic plugged in, and give a full height waveform)?
Yes, with one channel recording (w one channel empty).
It might not matter after all, each way gives good results, but the monomix Rec is comfortable. If it results in noise in the empty channel, invisible in the display and unhearable, better record one channel and then duplicate it in audacity. As described previously.
We are obviously misunderstanding each other somewhere along the line. Why waste bits / disk space / time and effort duplicating it?
Does the recorder allow you to record just one (mono) channel (the channel that has the mic plugged in, and give a full height waveform)?
If yes - do that.
If the only way to get a full-height waveform is to record two channels, then do that, split to mono, and delete the “silent” track.
See here for splitting (and joining) tracks: Audacity Manual
)