Question About Recording Vocals Over Mastered Beats

[waterise]

Hello,

My name is Branden and I rap. I have a Shure SM7B mic to a DBX 286a pre-amp, then from the pre-amp I have a little wire from the back of it straight to my PC. I use Cool Edit Pro 2.1 and I wanted to know about some things.

1) Clipping When Recording
Ok, so the way I make songs is this: I simply import the beat in multitrack view on the 1st track, then just record over it while the beat's in my headphones.
Now, when I am in the process of recording, should my vocals not be spiking over the top horizontal line in Edit view? Or is that after I mix it with compression, etc? 'Cause sometimes I record and it's spiking over just a little bit, but when I apply noise reduction & mix it, it's not really near spiking anymore. So, should it not be spiking before mixing the vocals first?

2)Normalization
When I go in Edit view to edit my vocals, I use noise reduction, Very fast attack compression (4:1 ratio I believe), Spit Clear, a FFT Filter preset, little reverb, then a hard limiter of -6.5 max I think it is. Then after all that, I normalize to -3 db. That makes the vocals louder since the compressing lowers their volume. Am I going about this the wrong way? Should I not normalize? I don't really hear distortion when I do it or anything. Or do I need to lower the beat's db level on multitrack view? 'Cause I've heard vocals should be -3db, so if I still can't hear my vocals too good over the beat after normalizing, I just lower the beat db volume.

3)Mastering for an Album
I'm working on an album & I wanted to know how to make all the tracks have around the same volume like professional albums are, so people don't have to turn their volume up & down for each song.

In the reply under this post, you can see the already mastered beat. Now, the beat is spiking without my vocals even on it, so it's gonna be spiking a lot more with my vocals added. Is that fine? How should I go about mixing it down correctly and whatnot?

Thank-you.

Mic:


Before Normalization:


My normalization setting

 

[waterise]

DBX 286a:



After Normalization:


Already mastered beat:

 

[datboysc]

This should help you out. Check out this thread: https://www.futureproducers.com/forums/production-techniques/recording-mixing-mastering/how-gain-stage-your-mic-pre-interface-302602/

It's all about gain staging my friend.

 

[waterise]

This should help you out. Check out this thread:

It's all about gain staging my friend.

What about my other 2 questions? & what's the fader mentioned in that post?

edit: Ok, so I guess the fader is just the slider/knob used to increase or decrease?
Are the knobs on my DBX 286a faders?

 

[audioguy]

Clipping While Recording


When recording vocals make sure that your input of your mic pre IS NOT IN THE RED. if you do not se red, but hear distortion, turn the it down about a db or so. Remember, you are recording in digital, you don't need to record hot levels.


Normalization


No. Do not normalize. I was told that if you have to normalize you are doing something wrong. Normalization is the last thing you want to do. Check out my blog article for setting up a mixing session. If you recording vocals over a mastered beat, turn the beat down about -6db and mix in the vocals around it. Or, mix the vocals at the level of the mastered beat.


Mastering


I would suggest that you find a good mastering engineer to master your records. Most of the time, it's best to let some else master your album if you did the mixing. There's a lot of particulars about mixing that can interfere with mastering if done by the same person.

 

[bandcoach]

Hello,

My name is Branden and I rap. I have a Shure SM7B mic to a DBX 286a pre-amp, then from the pre-amp I have a little wire from the back of it straight to my PC. I use Cool Edit Pro 2.1 and I wanted to know about some things.

1) Clipping When Recording
Ok, so the way I make songs is this: I simply import the beat in multitrack view on the 1st track, then just record over it while the beat's in my headphones.
Now, when I am in the process of recording, should my vocals not be spiking over the top horizontal line in Edit view? Or is that after I mix it with compression, etc? 'Cause sometimes I record and it's spiking over just a little bit, but when I apply noise reduction & mix it, it's not really near spiking anymore. So, should it not be spiking before mixing the vocals first?

2)Normalization
When I go in Edit view to edit my vocals, I use noise reduction, Very fast attack compression (4:1 ratio I believe), Spit Clear, a FFT Filter preset, little reverb, then a hard limiter of -6.5 max I think it is. Then after all that, I normalize to -3 db. That makes the vocals louder since the compressing lowers their volume. Am I going about this the wrong way? Should I not normalize? I don't really hear distortion when I do it or anything. Or do I need to lower the beat's db level on multitrack view? 'Cause I've heard vocals should be -3db, so if I still can't hear my vocals too good over the beat after normalizing, I just lower the beat db volume.

3)Mastering for an Album
I'm working on an album & I wanted to know how to make all the tracks have around the same volume like professional albums are, so people don't have to turn their volume up & down for each song.

In the reply under this post, you can see the already mastered beat. Now, the beat is spiking without my vocals even on it, so it's gonna be spiking a lot more with my vocals added. Is that fine? How should I go about mixing it down correctly and whatnot?

Thank-you.

Mic:

1. Always record so that your levels never go over 0dbFS (your top line) - at this point you are recording essentially a square wave component which will sound as distortion when played back. If you attempt to compensate after recording all you do is the reduce the overall level without removing the distortion. As a rule of thumb I aim for -6dbfs as my average.
2. • You need to read about compressors as well - a compressor properly used will not affect overall vocal level, it is used to narrow the levels between the compression threshold and the top levels you have in your signal: if you are experiencing severe drops in level as a result, then you need to consider
     • raising the threshold setting so as to narrow the overall signal band you are compressing
     • narrowing the compression ratio so that you are gently lowering the levels in your compression signal band 1.5:1 is more sensible than 4:1 or bigger unless your signal is way too hot to begin with, in which case go back and record it again so that the levels are more easy to work with.
   • I'd also leave the limiter out of the equation here as it introduces distortion too if driven hard.
   • Effects should be send not insert unless you cannot avoid it (noise reduction is an example of an insert effect): the resulting sound of an effect is on the final version of the audio (on the master bus only) not at any interim part of the process (unlike the old days when you had to effect as you recorded, as you dubbed and as you did the mix down, due to the high cost and therefore low availability of having multiple fx units in hardware)
3. Mastering is a final stage process, not an interim process - Mastering is about getting every track to have the same subtle yet identifiable sonic signature and same overall compatible levels.
   If, as your images demonstrate, you have instrumentals that are mastered and already spiking outside of 0dbfs then you have real problems with your process before you even add your vocals.
4. The rest of this addresses mix issues
   • If you are buying your instrumentals then request that they be mixed to -6dbfs or -3dbfs at most: this is to provide you with scope inside of your daw to actually mix your vocals and the instrumental instead of killing your vocals to get them heard or, worst, compressing or re-normalising the instrumental track so that you can mix your vocals to it.
   • If you are making your own instrumental tracks then treat the vocals as part of the production process - you will be amazed at how much more flexibility you gain when you can alter track levels via automation as you mix everything together
   • If you have no room to wriggle on how you get your backings - i.e.they are delivered as mp3's or similar; then at least convert the mp3 track to a wav file - it won't eliminate the defects and artefacts found in the mp3 but it will eliminate the opportunities for amplifying these defects as you proceed with the mix process

 

[waterise]

Thanks for the responses!
So pretty much don't use a limiter, record around -6db, and maybe avoid hard limiter?

Should I use the volume adjustment level thing on Cool Edit in multitrack view to increase the recorded -6db raw vocals to 0 then? Or how should I edit now if I don't use my limiter?

The "volume adjustment thing" :


-6 on the beat. Is it ok to +3 vocals? The vocals are at -6 without the +3 on the volume bar.

 

[bandcoach]

What is you obsession with getting it to 0? -6dbfs is what most reputable mastering engineers will ask for in a final mix sent to them for mastering.

A quick piece of advice:

mixing

is

the art of making other things quieter so that you can hear what you are interested in

not

the art of making everything louder so that you can hear no one thing well
​

From this premise, we never try to make anything hit 0, no matter what the context suggests that 0 might mean.

0dbfs is as high as something can go when we talk audio, unless your daw is special and uses some other meaning for it.

Pleas read the stickies on
gain-staging
compressors and limiters
Mixing/mastering info
Definition of Mastering
Understanding EQ

In the sequence suggested

 

[waterise]

I've read the "gain-staging" thread before & downloaded the VU meter, but I don't know how to use it with Cool Edit. Is the "gain" the input knob on my DBX 286a?

When I have it mixed at -6db, it's really quiet compared to the beat. Some professional artists can whisper and talk regularly into their mic and still have it sound perfect & crystal clear.
Example: ww w. youtube .c om /watch?v=V91ByDQpIdQ

At the beginning, Tech N9ne is just talking, but it is perfectly loud and clear to understand him. If I were to do that, it would sound totally drowned out by the beat if my vocals stayed at -6db. That's why I thought you had to try to get to 0, my fault. What can I do to get the louder, but clear sound that professional artists do in their mics?<br>

Another example of not raising your voice for the surrounding loud sound:
ww w. youtube .c om /watch?v=TtA4swl7gI4

 

[bandcoach]

I've read the "gain-staging" thread before & downloaded the VU meter, but I don't know how to use it with Cool Edit. Is the "gain" the input knob on my DBX 286a?

Gain is at several points during the audio chain.

The input knob on the dbx286a is the first point in your gain staging. the output knob is the second stage of gain staging.

the gain knob/slider/input area on the input channel in cool edit is the next point in your gain staging.

The master bus is the final point in your gain staging.

Don't worry about a downloaded meter - if cool edit has channel meters and bus meters, then you are fine to use them.

When I have it mixed at -6db, it's really quiet compared to the beat. Some professional artists can whisper and talk regularly into their mic and still have it sound perfect & crystal clear.
Example: ww w. youtube .c om /watch?v=V91ByDQpIdQ

At the beginning, Tech N9ne is just talking, but it is perfectly loud and clear to understand him. If I were to do that, it would sound totally drowned out by the beat if my vocals stayed at -6db. That's why I thought you had to try to get to 0, my fault. What can I do to get the louder, but clear sound that professional artists do in their mics?<br>

Another example of not raising your voice for the surrounding loud sound:
ww w. youtube .c om /watch?v=TtA4swl7gI4

did you normalise your beat track back to -6dbfs like I suggested or did you think that was optional?

If you have everything at the same level to begin with then you should be able to hear everything, if not turn the beat track down (turn it's input gain down or lower it's channel fader/slider/volume knob) until you can hear the vocals. Everything should be running along fine at around -6dbfs peak or maybe -4dbfs if you want to hear it knock a bit.

When you've got everything ready to mix down, export it without changing a single knob or level, as a wav file.

Now re-import the wav file into a new cool edit project and start the mastering process, which is about getting it to knock without distorting. Read the sticky on mastering and mixing secrets again and again and then decide what you need to do - don't just stick a hard limiter on it 'cause some fool who's been making beats less than a month says do it - I been doing this shit since 1977 and I'm still learning how to do it right. There are folks here who would disagree and say that it's gotta be knocking before you master it - they don't know they asses from they mouths and often put things in the wrong place.

If you want some further ideas look up some threads that moses and dvyce and sleepy have contributed to in the area of mixing and mastering

 

[LilAyUTP]

Compression and make up gain ^^

 

[bandcoach]

won't solve the initial problem.......

---------- Post added at 09:56 AM ---------- Previous post was at 09:42 AM ----------

So, I'm left wondering at what level do you actually listen to our playback and through what?

This may address some other gain-stage issues, but will also allow me to judge whether you are truly listening to your project or just listening without disturbing folks.

 

[waterise]

did you normalise your beat track back to -6dbfs like I suggested or did you think that was optional?

Yeah, on the first page, the beat track is at -6 volume. I didn't use the "normalize" setting though, just used the volume in the picture.

Here, I made a video showing what I do.
NOTE: That acapella I used the effects on was already mixed & mastered. I'm just showing what I do to MY vocals. I don't normalize now to -3 or at all, but that's what I meant by normalizing makes it louder and so I can hear myself over a beat.
Pause when needed.

ww w. youtube .c om/watch?v=bYJQqALjTRs

 

[waterise]

So, I'm left wondering at what level do you actually listen to our playback and through what?

This may address some other gain-stage issues, but will also allow me to judge whether you are truly listening to your project or just listening without disturbing folks.

I listen back through these headphones. You can hear every detail with them. I play it back from the computer I'm on at 100% volume.

 

[bandcoach]

go back and read the sticky on using compressors and what I wrote initially - your current compression process is too much - you suck all of the dynamic variation out of your signal by doing what you do to begin with and all you are really doing is dropping your signal down to the level of the quietest material rather than raising your quieter material to the level of your loud material - ever hear of makeup gain?

I would still strongly suggest that you are trying to do some things in the wrong place at the wrong time.

Record the vocals so that they don't overload the input on cool edit - once it's distorted it is always distorted

Don't compress the vocals until all other mix issues are addressed.

Can you not add each level of processing as a separate insert fx - I seem to remember that I can do this in Audition 4.0 which of course is the younger, yet more stylish sibling of Cool Edit.

Fix your backing track levels by normalising them outside of the program; use audacity it has a very good almost transparent normalise plugin as standard.

And I will say it again: if you do not receive your backing tracks as wav files then convert them (again using audacity) to wavs that you only use from that point in your project.

---------- Post added at 10:45 AM ---------- Previous post was at 10:39 AM ----------

I listen back through these headphones. You can hear every detail with them. I play it back from the computer I'm on at 100% volume.

View attachment 37955

here you miss my point - you are listening through those when you record your vocals only? Or to everything?

If the latter then you are listening without disturbing folks.

Do yourself a favour and get yourself some good quality audio monitors and a proper audio interface that has both ins and master outs and control room outs (the latter are so that you can listen on whatever playback system you connect it to).

 

[waterise]

go back and read the sticky on using compressors and what I wrote initially - your current compression process is too much - you suck all of the dynamic variation out of your signal by doing what you do to begin with and all you are really doing is dropping your signal down to the level of the quietest material rather than raising your quieter material to the level of your loud material - ever hear of makeup gain?

I would still strongly suggest that you are trying to do some things in the wrong place at the wrong time.

Record the vocals so that they don't overload the input on cool edit - once it's distorted it is always distorted

Don't compress the vocals until all other mix issues are addressed.

Can you not add each level of processing as a separate insert fx - I seem to remember that I can do this in Audition 4.0 which of course is the younger, yet more stylish sibling of Cool Edit.

Fix your backing track levels by normalising them outside of the program; use audacity it has a very good almost transparent normalise plugin as standard.

And I will say it again: if you do not receive your backing tracks as wav files then convert them (again using audacity) to wavs that you only use from that point in your project.
here you miss my point - you are listening through those when you record your vocals only? Or to everything? If the latter then you are listening without disturbing folks.

Do yourself a favour and get yourself some good quality audio monitors and a proper audio interface that has both ins and master outs and control room outs (the latter are so that you can listen on whatever playback system you connect it to).

So compress at 1.5:1 instead of 4:1 like you said on page one?
Or I guess until it sounds right to my ears?


I listen to the beat through those headphones while I record vocals, then play it back through the headphones with the beat after I edit the vocals. All the sound I hear will be from those headphones. Vocals only, the beat, and both together.


What do you mean by "backing track?" Do you mean the beat? I'm new to these sort of terms, sorry. Never really used or heard them before.


When you say, "Can you not add each level of processing as a separate insert fx" do you mean have effects on a track already there without going to the "Effects" menu like I do? Yeah, you can, but I haven't done it.

---------- Post added at 07:12 PM ---------- Previous post was at 06:59 PM ----------

Re-cap so I got it all:


*Record at about -6db - check
*Compress after all the other effects (right?) - will do from now on
*Do not normalize vocals - check
*Try compressing at a 1.5:1 ratio - will do from now on
-*Do not use limiter - Not sure if I should or not. It seems to help each time
*Make beat volume -6db - will do

You said, "Everything should be running along fine at around -6dbfs peak or maybe -4dbfs if you want to hear it knock a bit." Do you mean beat AND vocals together all should be at -6db or beat and vocals both at -6db when I mix?


Question: On page one you mentioned a "master bus." What exactly is that and how can I access it through Cool Edit?

 

[waterise]

Gotcha.

I thought I had to compress after everything, though? Or was it just a suggestion?

Because the way I do it as showed in the video, it sounds good to me; it just isn't loud/as surrounding like a professional song is. Didn't know if it's the mic, pre-amp or both. I'll change up the way I add effects and not do it in a destructive way.

Question: If I do all this to my vocals and the beat is at -6db and I STILL can't hear them as loud as I think they should be over the beat, should I make the Volume higher on the track or what from there? Make the beat lower?

 

[bandcoach]

Compress to whatever works for you but seriously you need to be less heavy handed in applying fx of any sort.

Get yourself some studio grade audio monitors; don't spend less than $1k and make sure you can hear what they sound like before you buy

The beat is a backing is it not - it is what is in the background when you are recording your vocals and remains there afterwards.

Add each level of effects so that you do not radically change the audio is what I meant here, rather than apply and then process the result of the application of a particular effect (which may or may not be undoable depending on how often you save, how many levels of history (undo) you have available, etc).

I do not mean to be rude but please get a hold of some basic reading material in this area and apply what you are reading as you go.

Insist that any backing track that you purchase is delivered as a -6dbfs mixdown wav or aiff file not a pre-mastered hard-limited piece of crud in mp3 format that makes it difficult to do anything with.

read this article from weiss sound on how to overcome but not completely survive the problems such crud delivers to you.

Above all be patient and try each thing one at a time - do not try to do everything in one go.

Compress your vocals - now check out how they are sounding.

Add some reverb to your vocals, now check out how they are sounding.

Do one thing, evaluate the impact it has.

Undo it and retry it if it failed to do what you wanted it to.

This where I am talking about non-destructive techniques with processing your audio - insert the effect into the processing chain rather than applying it directly to the audio file in an unrecoverable fashion.

---------- Post added at 11:33 AM ---------- Previous post was at 11:26 AM ----------

Re-cap so I got it all:
*Record at about -6db - check
*Compress after all the other effects (right?) - will do from now on
*Do not normalize vocals - check
*Try compressing at a 1.5:1 ratio - will do from now on
-*Do not use limiter - Not sure if I should or not. It seems to help each time
*Make beat volume -6db - will do

You said, "Everything should be running along fine at around -6dbfs peak or maybe -4dbfs if you want to hear it knock a bit." Do you mean beat AND vocals together all should be at -6db or beat and vocals both at -6db when I mix?


Question: On page one you mentioned a "master bus." What exactly is that and how can I access it through Cool Edit?

those numbers should be -6dbfs meaning -6db full scale - measured against your daws/systems absolute 0dbfs level.

When mixing for final sound before rendering the mix for the next stage in the distribution process (mastering), you should be aiming at -6dbfs to -4dbfs as your peak in the final mix of vocals and beat/backing track.

A Master bus is the master output channel or the mix channel and you control it's level by using channel gain and the channel volume fader/knob/dial. My guess is that when you move into multi-track mode, which you would have to be in to mix your vocals and beat/backing track together, this is an additional channel to the two you get from the vocals and beat channels. I use Audition 4.0 (came with CS5.5 Master Suite), which is the younger yet more stylish sibling of Cool Edit, and this is what I am presented with in multi-track mode.

 

[waterise]

^ My reply just got put above your somehow.

For some reason, "wet" and "dry" are grayed out.




Edit: How come when I import an artist's professionally mixed & mastered song into Cool Edit, it spikes over the 0db? Did the person who mixed & mastered their songs do it wrong?

 

[bandcoach]

^ My reply just got put above your somehow.

that's because I answered the added part to your reply above that after my initial response, so it got a new time-stamp and shows up now as being after yours even though it's time-stamp was before yours. This is the way of teh world of technology and absolute timestamps in database based websites and forums.....

Gotcha.

I thought I had to compress after everything, though? Or was it just a suggestion?

Compression is a do or not do thing. try both and do blind A-B testing with friends - you will know because you made the recordings, but they won't and all you want to know is which mix sounds better, not louder, knocks harder, but better. Enough variation in the way that you apply compression and consistent A-B testing and you will start to get a feel for what really works.

Because the way I do it as showed in the video, it sounds good to me; it just isn't loud/as surrounding like a professional song is. Didn't know if it's the mic, pre-amp or both. I'll change up the way I add effects and not do it in a destructive way.

I would suggest revisiting your input stages, as it seems like what you are really saying is that the level coming in isn't "hot" enough and so you are applying fx to compensate for this. Fix the signal levels as early in the capture process as possible - make the mic more sensitive by increasing the input gain/knob so you don't have to push as much air to get a sound. then adjust all other levels in your signal chain to reflect this new sensitivity.
In a live sound situation, where you are trying to make things heard over other things that you cannot turn down (you can remove them from the mix but if they are loud they will still be heard), your first go-to is compression - it drops the levels of a hot signal to a useful yet controlled level.
However, your focus in studio sound is to make everything heard without overkill:

• signal too hot not a good idea - this means compressing to reduce the louds and making the softs almost inaudible unless you use makeup gain and then you end up back where you were with everything too loud.
• signal too cold not a good idea - this means using post recording gain increases (also known as using an expander, the opposite of a compressor) to bring the signal up to a level where you can actually hear it and every breath and every scratch and anything else you can think of that might make a noise during the recording process.
• properly gain-staged input chain and the signal is audible without being overbearing and can be heard without further alterations to the captured audio file.

Question: If I do all this to my vocals and the beat is at -6db and I STILL can't hear them as loud as I think they should be over the beat, should I make the Volume higher on the track or what from there? Make the beat lower?

This is the art of mixing.

In my book the art of mixing is making things quieter so that other things can be heard. So if you can't hear the vocals, turn the input gain on the beat channel down

The art of mastering, in my book is making the track medium ready: CD quality 44.1kHz 16 bit LinearPulseCodeModulated audio file (more commonly referred to as AIFF or WAV) with a wide dynamic range (ideal is 96db, but actual is usually about 80db from the softest to the loudest sound reproduced). this means processing to create a club/car listening master (knocking and loud) or a radio/home listening master (subtle variations in dynamics are easily heard) or a combination

---------- Post added at 01:34 PM ---------- Previous post was at 01:21 PM ----------

For some reason, "wet" and "dry" are grayed out.

View attachment 37956

the difference s between your version of cool edit and my version of Audition are too large to offer any insight on this one, sorry

Edit: How come when I import an artist's professionally mixed & mastered song into Cool Edit, it spikes over the 0db? Did the person who mixed & mastered their songs do it wrong?

Might be artefacts of the conversion process - check you convert from audio cd processing chain to see if there is a sneaky normalise thrown into the input chain....

Alternatively, and more likely, it probably means that your settings are off in Cool Edit- i.e. 0db is not 0dbfs but some other lesser number like -1dbfs or -3dbfs. Some daws allow you to make such modifications to match your monitoring situation. Some apply them as a default setting which confuses more than it helps

Even more alternatively, it could be that you're thinking the -3dbfs lines are the 0dbfs lines on your tracks

I would need to see the manual to know more about which possibility is correct...... so version would be helpful (tracking manuals down on-line is very easy these days as long you know version numbers)