I’m an Audacity user and noob in optimizing sounds.
I’m trying to optimize some voice recording as training and for fun.
I have multiple recording from same voice but some records are really bad.
In example, I have this file (voix_amodif.wav), a bad record, and this one (voix_cible.wav) the good record of the same person.
Is there a way to modify the first file to approach the voice quality of the second one ?
So there are others that frequent this forum that have much better ears than I and can give you much better advice. But as no one as replied yet, I thought I would take a stab at it.
Firstly, when I listen to these recordings, even though the recordings are adjusted for maximum peak, just by looking at the waveforms, you can see that there is more energy in the first recording. You can level the floor by taking about 3dB off of the “good” recording. BTW, a louder recording normally sounds “better”.
Next, I displayed the tracks in Spectrogram View using the drop-down arrow on the Track Control Panel:
Then noticing weakness in the higher frequencies, I applied a filter curve (shown later) to the second track, which makes the two tracks look (and sound) much more alike:
Here is the filter curve:
You can also use Analyze > Plot Spectrum for more detailed work, but you may also need larger or more closely aligned samples.
I hope this helps
Theses records are not good but there are old (1995).
In fact, after some research, it seems that like the “bad” record (voix_amodif) try to make an “old tape effect”, like the record was make with a walkman. And I think that I can’t “reverse” this effect.
But again, thanks for your answer (and I test your advice but it’s not result that I hoped…).
I will continue to use Audacity to learn about sounds and mix.