RECORDING LOW BACKGROUND MUSIC W/ VOICE

Special Interest Groups Recording Techniques
41

Bill,

thank you for replying with all details
I did remove the click and yes I was finally able to get a 19.2 dB in the Amplify panel
I did actually finally get an amplification, thanks

however…sorry…but again, have to ask you, I still do not know where to go, in the main Audacity project window
to choose or _" Select from 2.12 to 2.83 seconds (approximately)

  • this is the noise profile we are going to use. You select it before you do Effect > Noise removal_
  1. Where do I select these 2.12 to 2.83 numbers, what are they?
    Are you referring by any chance to those teeny little numbers that appears along the track line, on top of the line, which look like inches in a measuring tape?
    Or are you referring to some time-frame units somewhere? Where?
    I am sure these numbers play a big roll, so I am struggling to find a way how to apply them

2)About talking louder and closer, I am almost touching the mic with my mouth and I am speaking as loud as my voice allows,
I did a test this way, and yes, the volume is better, but when I play it back through a Media device , volume is still poor (and it gets especially very poor when I transfer the file to my MP3 player and listen to it through the tiny payer) but even through Windows M Player is poor

It must be the poor preamp contained in the mic, dont you think?

  1. I have the mic volume slider in Audacity on 0.7 (3 quarters of the line) is that too high? Is a little louder at least

Look forward to your answers
Thanks a lot again
Anna

42

Yes, those are the numbers in the Timeline above the track: http://manual.audacityteam.org/man/Timeline

Click and drag in the waveform, the same way as if you were going to delete a section of audio. Use the “ruler” to guide you. The section I chose was one where you are not talking. You can see the exact range you have chosen in the Selection Toolbar http://manual.audacityteam.org/man/Selection_Toolbar

As for your other points, yes, it seems there might be a poor pre-amp in the microphone.

Are you speaking as if you were talking to someone on the other side of the room? That’s what I mean by “louder”.

You can try turning up the input level slider all the way, but I doubt that will help the noise issue.

– Bill

43

@Anna: The main problem with the sample you posted is hum. There is some hiss but it looks and sounds about normal. It would be nice to track down the source of that hum. How do you have the microphone mounted? Are you holding it? Is it in a microphone stand? Is the microphone stand on a desk? Is the desk the same one that your computer is on?

@Steve and anyone else following this thread: download the sample and follow steps 1 through 3 from my post a few days ago. Then do a Plot Spectrum on it. There are definite peaks at 79, 109, 152, 188 and 212 Hz. This makes no sense to me. Thus my questions above - I’m wondering if the mic is picking up computer fan vibrations from the desk. I’d also appreciate a second opinion on the severity of the hiss versus the hum.

– Bill

44

There is certainly something producing a hum, but I probably would not get too concerned about it.
With those noise reduction steps, the signal to noise is better than 50 dB. With a better noise sample and a bit more tweaking I’d bet you could get that up to near 60 dB. Then there was your first step - Amplify by 19.2 dB. Making the initial recording a bit louder (either by speaking louder, or placing the microphone a bit closer, the signal to noise could be up to better than 70 dB. which is not bad at all for a home recording. If the voice recording was then going to be mixed with a background (music), then the “silences” between words could be gated (after noise removal) for a very “clean” sounding mix.

It would be nice to find the source of the hum, but it could be due to almost anything (apart from “mains hum”). To find if it is environmental noise, a test recording could be made at a different (quiet) location to see if the same hum is still present.

If the hum is due to vibrations coming up through the mic stand, then some sort of “suspension” for the microphone would be likely to help. I seem to remember Koz showing us a “spider” that he made from plastic pipe and some large elastic bands.

45

Bill, Steve thanks WOW, this was so much educational and really helpful, I am learning so much through your valuable lessons that I could not ever have done all this progress all by myself, even just by studying and restudying the tutorials, thanks for all effort, it really helps a lot!!

OK to answer to your first question if I am speaking like talking to someone on the other side of the room, ??

No, unfortunately I am not talking that loud, because as it is a narration, if I do it like shouting, it sounds really bad and distorted,
but I am talking very loud as loud as I am able to in order to sound natural and clear and continuous at same time and pretty loud!!!

Yes it does help if i turn the Mic volume level (in Audacity) all the way up, but then the noise increases a lot too, I tried doing that and the noise was too much

But I am wondering if the pop filter that I finally got, is kind decreasing the power of the mic in someway?

I have placed the filter very close though, only half inch from the mic and half inch from my mouth ?, and at least is helping to remove all the pops

About how is the Mic positioned:
Yes, it is firmly positioned on the stand, but, now i am thinking, it is in its stand, but---- the stand is placed on a phone book (in order to reach the desired height without me having to bend down and get neck ache) , and the phone book is standing on an additional layer of Styrofoam and — the whole thing is standing on a standard PC table.

So do you think that that all this might be a cause for the humming and noise?

Then about where the laptop is positioned:
yes the laptop is not very close but not that far either

Now I did put the laptop on the carpet this last time, 2 feet from me, and yes, the humming noise was a little better this time, just a tiny less, however. the fan of the laptop is very loud when it gets hot and the mic might still be picking it up in anyway
so the close distance between the mic and laptop might be a key factor then?

Today I am going to test it with the laptop quite far from the mic, and also am going to do a second test from the older PC, which was very quite
About environmental noises, I always choose days when it is pretty quiet around, almost no sound at all, actually no sounds whatsoever, so I do not think that is the problem

So i am going to do this test today, with the laptop far away, and I will try to talk even louder… and I am also going to increase the volume slider of the mic just 1 mm, thats all and will see if the noise will be less (after amplifying and then doing the noise removal)

In the mean time,

  1. would you be so kind to do maybe some research on how to build this spider for the mic you were talking about? I wonder though how can I speak through the END of the mic, if the mic is hanging down from the ceiling?

2)Then. would you mind to tell me in poor terms what the following step you mentions mean and how I would need to accomplish it?:
“If the voice recording was then going to be mixed with a background (music), then the “silences” between words could be gated (after noise removal)”

Because yes, I am going to mix it with music so i wonder what “gating the silence between words” means—and if it would further help the quality of the recording if I apply it now?
or shall I re-place this question for you, later at a later time when that steps of mixing with music comes?
thank you all now, I appreciate it
Anna
PS
ooops… i forgot to ask:
Do those parameters about Amplify (19.2 dB) and 15 - 00- 150–.15 for the noise removal that you have suggested in your previous post, APPLY to all other recordings I am going to do? Or do i have to change those each time??? because I noticed that , while the Noise removal" figures you gave me work well with any recording, on the other hand, when i talk louder, and then apply the Amplify,I cannot reach the 19.2 dB any longer, the default does not reach even 5 or 4 dB sometimes, before the OK button becomes inactive, so cannot be applied, so should I change that from time to time?
thanx

46

The parameters that are important are the ones in the Noise Removal dialog. You do not want to reach 19.2 dB each time with the Amplify effect - that was just the case with that one recording. You just want to do whatever the Amplify effect will allow. The fact that you now see 5 dB or so in the Amplify dialog means that you are recording louder. Either you are speaking louder (that is good, so now your voice will be even higher above the noise), or you have turned up the recording volume (not so good, since you have also turned up the noise).

– Bill

47

We’re not suggesting that you should shout (unless that is appropriate to what you are recording).
There is a natural tendency for people to “talk to the microphone”. What happens then is that the closer the microphone is to the person, the quieter they talk, and they will often end up whispering. What is being suggested is that you half ignore the microphone, and imagine that you are talking to someone beyond the microphone. A good public speaker will make themselves heard at the back of a hall with or without the aid of a microphone, and they are not shouting. It’s not about shouting, it’s about projecting. From that short sample, I don’t think that you were whispering, but the suggestion was just to ensure that you’re aware of this natural tendency. The “imagined audience” is not necessarily at the far end of the room, but think about where they would be if this was a live presentation and not a recording. A news reader on the radio probably has the microphone about 10 cm away, but it sounds like they are talking to someone that is several meters away because they are projecting their voice as if they were addressing a group of people at a small press conference.

Microphone suspension (AKA “spider” or “shockmount”)
Here’s a commercial one:

And here’s one that Koz made from some plastic pipe and elastic bands:
koz-spider.jpg
(PVC Microphone Shock Mount)

Shockmounts vary tremendously in price, from less than 10 Euro (for example the t.bone SSM 4 – Thomann UK), to over 50, but they all do the same job, which is to help isolate the microphone from vibrations transmitted up through the stand. For desktop use, a cushion can work reasonably well.

Noise Gate
A Noise gate cuts off the sound when it falls below a “threshold” level. Noise gate - Wikipedia
The thing that you need to be careful of with noise gates is that it can sound unnatural if the sound suddenly cuts to total silence.
If the voice recording is being mixed with background music, then a noise gate would ideally be set up to fade out (quite rapidly) when there is no talking, then fade back in just before the talking resumes. This requires that the gate is able to “look ahead” to see when the recording is going to rise above the noise threshold. There’s a noise gate plug-in available here: Noise Gate (read the entire thread to pick up some tips on how to use it effectively). I would suggest an attack/decay time of about 0.3 seconds (30 milliseconds), gating to silence. The noise gate should be used after noise removal, not before.

A good peak recording level is about -6 dB. Setting the level higher than that is very risky as you could easily get an occasional peak that clips. Much lower than that and the background noise becomes more significant.

48

Thanks a lot Bill
I really do so much appreciate you taking the time to explain in the phrase below how I should think and speak when I record

“There is a natural tendency for people to “talk to the microphone”. What happens then is that the closer the microphone is to the person, the quieter they talk, and they will often end up whispering. What is being suggested is that you half ignore the microphone, and imagine that you are talking to someone beyond the microphone”

Because it really really helped me to understand and makes so much difference,
So i did record speaking in that way and spoke very loud, like directing my voice to the other side of the room, and tried not to shout, what a difference!!
I cannot speak too loud though, because this recording is supposed to be for a guided meditation, and my goal is to reach a DEEP-VOICE effect not a shouting voice effect
SO i think i am getting there, thanks to your advises!!
But the background whining and noise is still there though, just that after Amplifying and then removing the noise, this time I could reach a real good amplification, and the noise is almost totally gone, even though my voice still sounds squicky, not deep, I have to find a way how to eliminate that

And so, I did this experiment that I mentioned earlier, to see if I could improve the volume, but it was very disappointing:
I did buy this M Audio Fast track Audio Interface, to try to by-pass my cheap PC sound card and hoping to reach a deeper-louder voice recording, without having to shout

But, first I was not able to install their CD as it required lots of extra RAM and i would have ended up in spending at least $150 extra just for that
So they told me that I if I downloaded only drive for the “fast track” I could use it anyway, and choose fast track in Audacity preferences

This box has phantom power and a good sound card
But the result was way terrible, way more noise and even an additional whistle now in the background, plus I was shouting, but when I played back I could barely hear anything, I do not know why they say Audio Interfaces help to have a better amplification

I did apply the amplification then the noise removal to the recording, but at that point my voice would become distorted, So it did not help at all, I have to return it, I am glad you suggested to test it first

And so here is the other question
I have downloaded into Audacity 3 free Plug ins ( Blockfish, Floorfish Spitfish) from Digitalfishphones.com – (I hope is OK) because I read online that they really help further to remove noise and enhance the depth of the voice
But I do not know how to use them, i experimented with them on one recording file, and I do not know what I did, but in one option, it seemed to really deepen and enhance the poor volume of my recording even more, just that I do not remember what and how I did that

IS there a way that if you tell me how I can attach these plugins to my post for you to see them, maybe you could tell me which dial in them to use to improve the amplification and quality of my recording?

About the Schockmounts, thank you so much too I will see if I can find them online and the price

And I am so glad, i also did get the Noise Gate plug in now, thank you sooo…,much for that and for explaining what it is for and what to learn
I will need to read through its tutorials and study them a lot first, as the only thing I know about it, is what you told me : “an attack/decay time of about 0.3 seconds (30 milliseconds”

When you say:
"“A good peak recording level is about -6 dB. Setting the level higher than that is very risky as you could easily get an occasional peak that clips. Much lower than that and the background noise becomes more significant”
Are you referring exactly to the recording slider of the mic volume that is in Audacity? And why is negative 6 (-6) rather than just 6?
Or are u referring to something else? I just want to make sure I understand where exactly i should see that -6 value—thank you!!

And last question:
If I record in Mono (as my mic is supposed to be Mono) --what "Default Sample Format" should I choose in the “quality” option in the Preferences? Is it 16-bit?
Or is it 32-bit float?
It seems to me that the 32-bit sounds better and less noisy, or is it only my impression?

Thanks a lot again
Anna

49

The -6 dB refers to the peak level as indicated on the recording VU meters. The meters read from -60 dB on the left to 0 dB (maximum) on the right.
See: http://manual.audacityteam.org/man/Meter_Toolbar
Note in the first illustration on that page, the blue “maximum peak” indicators are at approximately -6 dB.

This box has phantom power and a good sound card
But the result was way terrible, way more noise and even an additional whistle now in the background, plus I was shouting, but when I played back I could barely hear anything, I do not know why they say Audio Interfaces help to have a better amplification

Did you plug your microphone into the big connector on the left front with 3 pins? That’s called an XLR connector, and you would need an XLR cable to go from your microphone to the FastTrack. The Samson microphone you are using has both USB and XLR connections, right? You should only use one connection at a time. If you have the Samson microphone plugged into the FastTrack, the USB cable should not be connected to the microphone.

If I record in Mono (as my mic is supposed to be Mono) --what “Default Sample Format” should I choose in the “quality” option in the Preferences? Is it 16-bit?
Or is it 32-bit float?
It seems to me that the 32-bit sounds better and less noisy, or is it only my impression?

It is probably your imagination, although 32-bit float is theoretically higher quality. The limiting factor in your recordings is the noise generated by the microphone, or microphone pre-amp, or the background noise in the room.

– Bill

50

I posted the below post few days ago
I wanted to apologize to all, because I did not realize that
first, the question had not really much to do with Audacity itself, and so it was maybe not a proper subject in this forum, sorry if I created inconveniences and/or did not respect the general rules?
secondly, I did not realize that there were actually the PDF files manuals for each Digitalfishphones plug in, even though they are mostly directed to a highly professional public and not easy to understand

I thought that maybe, as it is allowed to download free plug ins form online sources into Audacity, that maybe my subject/question might fit in this forum
I am sorry if it did not
Anna

51

Anna:
There was nothing wrong with your post. I think it’s just that no-one here knows anything about those plug-ins.

Perhaps if you started a new thread asking if anyone knows about the Digitalfishphones plug-ins you might have more success. The “Audio Processing” forum would be a good place to ask that question.

– Bill

52

Thanks a lot for letting me know,
I am glad I did not cause any problem
I did post it where you suggested

I tried to remove that post in this forum (my previous to this post) to free some space, but I cannot find the remove button, so you can remove that post if you want, thank you

I had another question that I cannot find in the tutorials
1) When I am on my way to mix the voice with music, and I choose IMPORT (to open the music file in my voice recording) do I choose : IMPORT>AUDIO? or IMPORT>RAW DATA?
I have never known which is best

2) After I have recorded my voice file in MONO and then I open (IMPORT) the music file into my project, the music file opens up as a STEREO file (two tracks), do I have to do to go to TRACKS> and apply the STEREO TO MONO feature to make it a mono? or leave it as a stereo?

Last time i did leave the music file as a STEREO, (even if my voice was MONO) then I did save the whole project as a WAV, Audacity turned the whole file into a stereo WAV, which was not so bad at all, did not sound bad…But I am not sure I am doing the right thing

3) All my background music files in my hard drive are MP3 format, for now
When I open or (IMPORT)-- the music file (which is a MP3 format) , in order to mix with the voice,
and then I mix the project , is it OK to save the whole project as a WAV, even though the original music file was in MP3 format?
Or will create distortions?, It sounds good to me, but I might be wrong

Thanks a lot
Anna

53

  1. Use File > Import > Audio for all standard audio file formats such as WAV or MP3.

  2. Leave the imported track as stereo. When you Export your finished show your voice will be “in the middle” and the music will be in stereo.

  3. Saving as WAV is always safe. Saving as MP3 will further degrade the quality of the music, but will be necessary if you plan to post your finished show on the internet.

Also have a look at: http://manual.audacityteam.org/man/Tutorial_-_Mixing_a_Narration_With_Background_Music

– Bill

54

Thank you so much
It sounds great!

I have been able to, after recording very close to the mic and speaking loudly, then amplifying, then remove the noise and also apply some slight boost, to have my voice recording reach a normal , decent level, I might be wrong, but it sounds OK

But the problem I am having is this:
I have tested to mix it with music, I moved the gain of the music track all the way to the left, to -36 dB, it does not go further
But the music is still too loud compared to my voice, it is good if it is a song, but is supposed to be a guided meditation, so music has to be low, the music still covers up the sound of my voice

If I try to move the gain of the voice track to the right (in order to increase the voice track gain) then is when the loud background noise really appears now, so i cannot increase the GAIN slider of the voice track

What could I do, what effect could I apply only to the music track, to do the opposite of BOOSTING the volume, to reduce the gain to make it less loud even before I mix it with the voice?
Is there something I could do?
I did not find any information about how to do this specific step, in the narration with music tutorials
Thank you sooo …much
Anna

55

Use the “Amplify” effect, but enter a negative value.
Try amplifying the music by -20 dB (minus twenty), then use the gain slider to tweak the music volume.

I don’t know Bills’ opinion, but I’d probably also use a small amount of “dynamic compression” on the voice track. http://manual.audacityteam.org/man/Compressor

56

I appreciate this advise so much Steve!!!
This is exactly what I need to do, it sounds like the missing step, thank you !!! I am going to try that

I also noticed that , when I listen to the mixed voice with music file, through the PC speakers, the music sounds proportioned to the voice (much less loud and ok)
but when I listen to the mix through stereo headphones, then the music becomes more dominant and too loud over the voice
I would like to know if you think that is normal?

Thanks sooo…much
Anna

57

Hi Steve, Bill and all
Time ago I was asking you about noise removal and the steps

Now I am doing some editing again, and there are breaths, between spoken words. and also some normal zzzzz noise between words that can be easily removed

is it better to cut out the breaths and clearing of my voice between spoken parts FIRST? then applying noise removal as a 2nd step?
Or is it better to do the noise removal first, then remove all breaths ?

Or it does not matter which,?
i think I was told by some friends in the past that is best to remove breaths first then do noise removal, but I cannot remember exactly and also, I would like to ask your professional expert opinion
Thank you
Anna

58

It probably won’t make much difference, though when you select your “noise only” sample for the noise profile, it should not include any breathing - you want only the “zzzz” for your noise sample. If in doubt, test it both ways - depending on the actual audio it may be a bit better one way than the other.