Greetings,
I am using Audacity Version 2.0.6 with Windows 7
I have a few tracks of old music in MP3 format. I find it very noisy and shrill.
I have tried Normalising it down by -2.5db and also by equalising it FLAT by -3.5 db.
Still it is shrill and noisy.
I wish to convert this to WAV and burn Cds after I make it little more listenable.
Would appreciate suggestions on this.
Thanks Zameen
Have you tried Audacity’s noise removal function. I have a tutorial on YouTube see https://www.youtube.com/watch?v=qW7pdQEB9eA
Stearman65
THANKS Stearman65.
I had done all that.
I saw your video ( and it is helpful ) and trying it all over again.
What I encounter , while trying the things suggested in your video is this.
After normalizing ,( and I need to do it to -7db), , if I go and do amplify ( without clipping) I am really back to the same levels as before i normalized the file.
Wonder what I am doing wrong.
Also how does one expand the track here to get a reasonable reading for NR sampling ?
Thanks
zameen
Then don’t Amplify, unless you set New Peak Amplitude to -7 dB. Do you need to use Amplify after Normalize? Does this help: Amplify and Normalize - Audacity Manual ?
Expand what track? The idea is that you sample noise only. The sample can’t be too short, but half a second of noise should be fine if it is the same noise that you are trying to remove.
Are the waveforms flat-topped? If so, the audio is damaged by being recorded too loud, probably beyond repair. Making it quieter will just make the distortion quieter, along with making the rest of the audio quieter.
Gale
Thanks, Am a little wiser now…
I tried the amplification , but to a lesser db , than what I had normalised it at. And it seems to work better .
The wave front is not flat, but spiky …
Actually , the tracks are shrill , not so much as noisy… I don’t know how else to express this.
For examples the violins come through shrieking and painfully so.
That is the thing I wish to correct.
Zameen
Have you tried reducing the high frequencies? You can do that in Bass and Treble or Equalization .
If you are still not sure, please post a short sample of the squealing violins. See How to post an audio sample .
Gale
Thanks , I am attaching a small file over here, Hope for some suggestions to help me through my endeavor .
Thanks Zameen
There is no link or attachment.
Gale
I have tried Normalising it down by -2.5db and also by equalising it FLAT by -3.5 db.
With the equalizer set to “flat”, you are NOT equalizing… You are just reducing the volume. Normalizing is also just a volume change, no different than reducing the volume at playback time… It doesn’t change the “character” of the sound.*
After normalizing ,( and I need to do it to -7db), , if I go and do amplify ( without clipping) I am really back to the same levels as before i normalized the file.
With the Amplify effect, you can choose the amount of amplification (up or down). By default, it will be set to whatever gain is needed (if any) to bring your peaks to 0dB. You’ve reduced the volume by about -6dB with Normalization and “Equalization”, so it’s not surprising that Amplify defaults to +6 or +7dB to bring the volume back-up.
I have a few tracks of old music in MP3 format. I find it very noisy and shrill.
As Gale says, bass & treble adjustments or equalization are probably your best bet. The Graphic EQ mode is generally easier to experiment with than the Draw Curves mode. This is just a wild guess, but try pulling the sliders down between about 1kHz and 4kHz. Start with about -6dB and experiment from there. That should remove some of the shrillness.
I have a few tracks of old music in MP3 format. I find it very noisy and shrill.
You may have low-quality low-bitrate MP3s. If that’s the case, you may be able to make some improvement but MP3 compression artifacts are impossible to remove. i.e. Violins encoded to low-bitrate MP3 (or cell phone compression) are never going to sound “natural”. And who knows… This may have been a poor performance on a cheap violin or a poor quality recording to begin with…
What is the file size and playing time of one of the MP3s? From this, we can calculate the bitrate (kbps = kilobits per second). There are 8 bits in a byte, so:
Bitrate in kbps = 8 x (File size in kB / Playing time in Seconds)
This calculation is approximate because the file header takes up some space and there may be embedded metadata & graphics also taking up space.
Good quality MP3s are usually around 200kbps or more. I think MP3s downloaded from Amazon are 256kbps, but they may use variable bitrate. The maximum MP3 bitrate is 320kbps. 128kbps may be acceptable for some music. (You can go lower for voice, or if you are not concerned with quality.)
\
- There are cases where the peaks can go above 0dB. But, “normal” WAV files and and your digital-to-analog converter are limited to 0dB maximum. In that case, normalizing (reducing the peaks to 0dB or below) will prevent clipping (distorted flat-topped waves).
Hi,
Thanks for the patience , I am trying to attach a small file here.
Trust it gets through.
Thanks Zamen
There are no squealing violins in that file.
Everyone has different ideas of good EQ. We can’t fix the file how you like it because we don’t know what EQ you like.
So I suggest you open Effect > Bass and Treble…, try reducing the treble, press Preview, and if that does not sound right, try another setting, such as increasing the bass instead. When it sounds right, click OK.
Gale
" for the forum.wav" Sounds OK to me as it is .
But if you want to attenuate several narrow resonant peaks with an equalizer it’s going to be difficult.
There is a free plugin for Windows , called DtBlkFx , which will do that automagically …
This is how the creator of DtBlkFx illustrates the effect of “clip” …
DtBlkFx’s “clip” applies (downward) dynamic-range-compression to the spectrogram, (not to the waveform).
To my ears the processing introduces (or exposes) distortion in the last half.
But I have no clue what Zameen wants.
Gale