Recording Levels - What's Right?

[bimach]

I've been recording pretty much with trial and error for a while now and I still cant figure out the method for recording the right levels so that I can get a loud, crisp sound on my finished WAV product. I recently bought an EMU 0404 sound card and when I have recorded a track to Cubase I can manipulate the sound to be loud and clear at the level that I want it, but when I go to mix it down to a wave file and then play through a program like iTunes, the sound is dull and soft compared to other music files. Any ideas how I get that loud finished product without clipping? I feel it may be something to do with the levels of my preamp, but I have tried (I think) every combination and it still does not seem to give me what I want. Any help with this would be great. Thanks.

 

[DuoToneBand]

eq the finished product in a program like sound forge

 

[bimach]

Thanks, but could you please tell me more? That means absolutely nothing to me in the sense that I don't know what that program is or how to EQ something or what EQing means.

 

[Monkey Allen]

Where is the sound hitting on your input meters when you record it? I have heard that about -6db is where you want it.

 

[Fishybob]

If you are recording at 24bit then don't worry about getting the loudest track while recording. I think the process that you're talking about is something to worry about once the whole mix works.

Once it all sounds lovely, then you can eq, compress, limit, dither... and/or whatever you like to get the final volume up there with the commercial releases.

Just track a good sound at a decent level, mix with headroom and worry about the other stuff later.



Here are some tips that'll explain everything you'll ever need to know (well for a while). Yes they may be confusing for a newbie but it will all come together pretty quickly and the difference is huge!


http://www.soundonsound.com/articles/Technique.php

especially
http://www.soundonsound.com/sos/aug04/articles/computermastering.htm
http://www.soundonsound.com/sos/Mar03/articles/moreofeverything.asp

best for eq
http://emusician.com/mag/emusic_equal_time/index.html


Stick with it man.

 

[LRosario]

well this is the deal,


Setting the right levels is a situational thing, it depends on the gear and how you calibrate it. The usual approach is to get the incoming signal as close as you can to 0db, or just before clipping. It sounds absurd, but that'll give you your best signal to noise ratio. In other words, it will insure you have more usable signal than unwanted noise.

 

[ecktronic]

Once your track has been mixed and bounced to a stereo wave then you need to master it. Certain things to do at mastering could be EQ (boost lo and hi end and play about with mids, if needed), compression (maybe), Finally boosting the level to get it as loud as a commercial CD (or however loud quiet you want it), to boost the level you need a good limiter (waves L1 or L2 is a good one).
Hope this helps. You could always speak to MASSIVE MASTER or MASTERING HOUSE they are both professional mastering engineers who could master your tracks for you or even give you some more detail into the mastering process.

 

[David Katauskas]

When tracking, I always look at the signal levels both with the instrument playing, and without the instrument playing. Then I try to get those levels as far apart as possible (signal : noise ?) via knob twiddlin'. Many times it's easily corrected at the source.

 

[Farview]

The usual approach is to get the incoming signal as close as you can to 0db, or just before clipping. It sounds absurd, ....

Because it is absurd, as long as you are recording at 24 bit.
Recording levels should hover around -12dbfs. When you mix, the song should hover around -12dbfs. The mastering stage is where you get everything loud.

 

[sdsi]

well this is the deal,


Setting the right levels is a situational thing, it depends on the gear and how you calibrate it. The usual approach is to get the incoming signal as close as you can to 0db, or just before clipping. It sounds absurd, but that'll give you your best signal to noise ratio. In other words, it will insure you have more usable signal than unwanted noise.

Although I don't always practice this...this person does make a valid point. It does indeed give you the best signal to noise ratio...whether you are recording at 8 bit or 32 bit.

 

[warble]

well this is the deal,

Setting the right levels is a situational thing, it depends on the gear and how you calibrate it. The usual approach is to get the incoming signal as close as you can to 0db, or just before clipping. It sounds absurd, but that'll give you your best signal to noise ratio. In other words, it will insure you have more usable signal than unwanted noise.

I have to disagree with this concept. I usually track at about -12db and maybe get a couple peaks up to -6db when I start mixing all the tracks. It's nice to have a little headroom when mixing what's been recorded. I don't experience any noise problems either. How can you possibly record everything at around 0db, and then try to mix it all? Maybe it works for you - if so, that's cool.

 

[SouthSIDE Glen]

This may be an apples and oranges thing. Here's my take on it, FWIW...

In general, running close to 0dB is good when talking early in the signal chain (e.g. out of the preamp) and good when tracking to disc or analog tape as a discrete mono or stereo recording that will either fly on its own or later be moved (re-recorded) to another deck for multitracking or mastering.

On the other hand, when tracking individual tracks to disc intended to be later mixed by DAW, it's a good idea to leave a few digital dB of headroom for summing peaks and then to digitally mix down to stereo, leaving a few digital DB for mastering elbow room.

The apples are either a discrete mono or stereo recording (e.g. recording to stereo DAT), or early digital signal chain levels such as preamp outputs, or recording to analog tape in which case you need all the S/N you can squeeze an peaking at +3 or more is not always a bad thing. These are the 0dB areas.

The oranges are when dealing with multi-channel digital tracking and mixdowns, in which the 6dB or so of headroom are advantageous.

G.

 

[sile2001]

I tend toward the -12dB range as well. It seems that when I used to shoot for close to 0dB, no matter how much preparation and careful tuning I did ahead of time, and no matter how many times I told the person "play as loud as you plan on getting", something would always happen, and they would be even louder, and I would have a nice little nasty sounding digital clip right in the middle of a great take.

 

[Cylkk]

so lets say you've recorded your signils at an average of -12 and have peaks up to -6ish, then you compress in the track.
The make up gain for the compressor should be what? bringing the average up to -6 with less peaks? or keeping the average at -12 (no gain) and taming the peaks?

 

[sile2001]

There is a huge difference between the peak levels and percieved loudness. A song that is properly mixed and EQ'd but only peaks at -12 can actually seem louder than if it is poorly mixed and EQ'd but peaks at -0.5 simply because the properly mixed song is so much clearer.

I can make my songs average -6 and peak at -0.5...and Flood can make U2's songs average -6 and peak at -0.5. Why do U2's songs sound loud and clear while mine sound quiet and muddy? Because Flood can mix circles around me while he's sleeping with his hands tied behind his back!

Mixing and mastering is where it's at (well, and a good source). The actual average and peak volumes of a song don't really mean a whole lot if it's not properly mixed and mastered.

 

[Cylkk]

yeh i realise that, but doesn't answer my question

 

[Farview]

The simple answer is to use the makeup gain to bring the level up to -12dbfs average.

The convoluted answer is that the makeup gain doesn't matter. By the time you are compressing it, you are mixing. Just bring the level of the track up to where it sits in the mix properly.

 

[Farview]

This may be an apples and oranges thing. Here's my take on it, FWIW...

In general, running close to 0dB is good when talking early in the signal chain (e.g. out of the preamp) and good when tracking to disc or analog tape as a discrete mono or stereo recording that will either fly on its own or later be moved (re-recorded) to another deck for multitracking or mastering.

On the other hand, when tracking individual tracks to disc intended to be later mixed by DAW, it's a good idea to leave a few digital dB of headroom for summing peaks and then to digitally mix down to stereo, leaving a few digital DB for mastering elbow room.

The apples are either a discrete mono or stereo recording (e.g. recording to stereo DAT), or early digital signal chain levels such as preamp outputs, or recording to analog tape in which case you need all the S/N you can squeeze an peaking at +3 or more is not always a bad thing. These are the 0dB areas.

The oranges are when dealing with multi-channel digital tracking and mixdowns, in which the 6dB or so of headroom are advantageous.

G.

The reason your take is apples and oranges is because you are talking about analog and digital as if the scales were the same. 0db on an analog machine is -12dbfs digital so proper gain stage would dictate that you would set your digital levels for -12dbfs and your analog levels for 0db. That works out because they are the same level.

 

[RAMI]

Flood can mix circles around me while he's sleeping with his hands tied behind his back!

Man, I thought these geniuses were weird, but this is a little too kinky for me.

 

[SouthSIDE Glen]

Man, I thought these geniuses were weird, but this is a little too kinky for me.

Yeah..."sleeping"...Yeeesh!

G.