A Bunch of Noob Questions

[Crow_]

Alright, I have a bunch of recording questions, so I figured it'd be better to make one thread rather than a bunch in different forums - I hope that's alright.

My recording setup is pretty different than most that I've read about. So to start off with, I'm wondering if this is alright, or if it's terrible. I use my Peavey PV14 mixer's tape-outs into my computer's line-outs. My sound card is a Creative Soundblaster X-FI Xtreme Audio - it works very well for this. The software I use is Audacity, 48KHz at 16bits.

So usually I plug my guitars into their effects processors, which have limiting to remove static and the sound of the pick, which outputs to my mixer. Even though I'm using the mixer, I only record one track at a time, so everything but one input is muted. I volumize it so that whatever I'm recording peaks, on average, at -6 dB.

Recording the drums and the bass is fine, but I think my problems involve miced recordings. My vocals, even if they peak at about -6 dB on my mixer, are always really quiet. My horns are a lot louder, so the volume sliders are quieter, but they're still quiet in the recording. Yet, neither the vocals nor the horns can be amplified very much without clipping. Recently I've been using a condenser to limit peaks so that I can amplify more, which helps a lot, but I have a new problem.

I was recording a song for my band yesterday, so I'm trying to fix it up today. The problem is that even though none of the individual tracks clip, when they're mixed together into a single stereo track they do. Is this Audacity's fault, or is it just because I have everything too loud? I lowered the gains on everything, and manually deamplified some extreme peaks within some of the tracks, but I still have clipping. And now the song is quieter than the average rock song. What am I doing wrong?

This is what the song sounds like right now. The drums sound kind of iffy, but they were recorded with a cheap drum-mic set, so I don't need any advice on those. But any other advice would be appreciated.

http://qshare.com/get/886144/Were_Alright.mp3.html

 

[Massive Master]

Digital levels (except when you're clipping) mean nothing. Analog input levels mean everything.

-6dBFS is actually pretty hot for vocals at the input... Assuming your converters are calibrated to -18dBFS -- *maybe* -15dBFS -- Well, long story short, you're tracking too hot.

This might help: http://www.massivemastering.com/blog/index_files/Proper_Audio_Recording_Levels.php

After that - Your recordings aren't supposed to "be as loud" as a typical 'off the shelf' CD. There's probably a post about that on the blog also.

Properly calibrating your monitoring chain would be a rather huge help - That's on the blog for sure. That would take an awful lot of 'chance' out of the whole volume thing. Input and output.

And *of course* you're going to be clipping the buss all over the place if you have a bunch of files hitting -6dBFS at the same time. That goes back to 'proper recording levels' -- And even when using 'normal' tracking levels, you're still very likely going to need to attenuate to some extent to prevent clipping at the 2-buss.

But this is the wrong phase to be worried about final playback volume... And almost anything you do to worry about it early (other than keeping wonderful amounts of headroom at every conceivable stage in the game) will come back to bit you in the a$$ later.

 

[Dani Pace]

Don't worry, you are not alone. It takes a little practice to get used to recording much "quieter" than what you want to hear. I started recording on analog systems where recording hot and the occasional hitting the red line was acceptable, sometimes even preferred. It took me a while (and a bunch of really bad recordings) to get used to recording with lots of headroom. Finally I realized that getting that volume I wanted to hear was part of what mastering does. The easiest way I found to work around the lower recording levels was to increase the volume in my monitoring chain. It is much better to monitor "hot" than to record "hot."

 

[Crow_]

Alright, thanks for the feedback. I read your post about recording quieter (Massive Master), and I have one additional question.

After I've recorded at a lower level of about -20dB, will I want to amplify my tracks as much as possible without clipping before using limiters and EQs and whatnot, or should I mess with everything while it's still quiet?

 

[philbagg]

Alright, thanks for the feedback. I read your post about recording quieter (Massive Master), and I have one additional question.

After I've recorded at a lower level of about -20dB, will I want to amplify my tracks as much as possible without clipping before using limiters and EQs and whatnot, or should I mess with everything while it's still quiet?

Mess with it while it's still quiet. That way you're leaving yourself lots of headroom. Lets say you just turn up your tracks as far as they'll go now, then
you start boosting frequencies with eq. They'll be louder again. And you'll clip.

Or if you turn everything up, and then turn the vocals up and they won't go
any further but they're still not loud enough? You'll have to turn everything else down. This will happen over and over. Not to mention, if your using sends (for reverb/delay etc.) and you've pushed your faders up, you'll probably clip on the send which will produce a digitally distorted (not nice) version being sent to your fx unit/plug. There's also the risk of clipping in plugins, and you won't see it straight off because the plug-in window won't be open. It's just an all round bad idea.

Just turn up your speakers/headphones while mixing. You won't notice it's quiet, seriously. Then when you've done your mix you can add some compression (very small amount) to the master track and get it as loud as possible without clipping.

The loudness of the entire mix is best left til last, ie. the mastering stage (or simple light master track compression if you don't choose to get it professionally mastered).

 

[philbagg]

Double post.

 

[philbagg]

Triple post, stupid internetz...

 

[Massive Master]

Agreed - but let's take something into consideration just because it bugs the hell out of me --

-20dBRMS (ish) is NOT "quiet" - It's NORMAL.

It's the freakishly LOUD recordings on the shelves that are NOT normal.

I'm as guilty as the next guy in churning them out (under protest). But I really think if more people realized that "line level" is "typical" - in some cases, even "too much" (EBU/THX/DTS, etc.), even if it doesn't "begin the healing" at least we'll be thinking clearly about it.

 

[philbagg]

-20dBRMS (ish) is NOT "quiet" - It's NORMAL.

Do you mean pre or post master?

 

[Massive Master]

Both.

Notice I said "normal" and not "typical" --

"Back in the day" (not long ago) a typical "LOUD" disc was riding maybe -15dBRMS. Earlier, really normal - maybe around -18dBRMS. That was typical (and normal at the same time).

What we're doing now might be typical - But it's not 'normal' (with respect to the design and specifications of the playback equipment).

 

[Crow_]

Then when you've done your mix you can add some compression (very small amount) to the master track and get it as loud as possible without clipping.

So instead of amplifying, compress instead with a normalization to 0dB? Would I want to do that to the individual tracks, or combine them all into a stereo mix first?

 

[Massive Master]

*Assuming* you mean *MINUS* 0dBFS ("0dB" means nothing - literally), there is basically no time at any step in the entire recording process that anything - Any individual track, group, aux, buss, (etc., etc. down the line) should ever approach full-scale (-0dBFS, which of course, you should actually leave a half dB or so of headroom - as if a half dB could be seriously called 'headroom') until the very final processes during the mastering phase.

Normalization hasn't been done 'normally' (heh, heh) since the early 90's. But tasteful use of compression and limiting are certainly in there somewhere...

 

[Crow_]

Yeah, I meant near the end of the mastering phase, after everything else is done. I usually go to -1dB. Would I want to amplify my stereo mix to that volume though, or compress to it?

 

[philbagg]

Yeah, I meant near the end of the mastering phase, after everything else is done. I usually go to -1dB. Would I want to amplify my stereo mix to that volume though, or compress to it?

In regards to your individual tracks, only compress them if they need to be
compressed. At that stage, all you're worrying about is the mix and the individual
tracks. Repeat: MIX and INDIVIDUAL TRACKS. Not your final output.

Now,

amplify my stereo mix to that volume though, or compress to it?

Not quite sure what you mean here. When you've done your mix, and you want
to get it louder, use some compression on your master stereo bus.

Let's say you've done your mix and don't compress the master stereo track,
and you normalise it so that it peaks at 0dBFS. There could be a few
peaks throughout the song that are much louder than the average level. So
only they will be up at 0dBFS, and everything else will still be fairly quieter
than this.

So, this is why you compress. If you can reduce the level of those peaks,
and then normalise the track, the average level will be significantly louder.

Does this make sense? (not sure how clearly I actually explained it)

I could get into some maths if you want

 

[Crow_]

Alright, that explanation made sense. Thanks!