how loud to record???

[AlfredB]

hi there, folks:

i am pretty new to home recording - but with lots of enthusiasm...


one beginner Q:

how loud should I record INDIVIDUAL trax?

- roughly as loud as they will appear in the final mix (so, a background track would be recorded at lower volume than a lead track)????

- or record roughly all tracks at the same level (-6 db) independently if they will be louder or less loud in the mix?

thx for clearing this up
alfred

 

[Ian Lund]

how loud

Recording engineers use something called Gain structure. Proper gain structure will put your level where it is not too hot to cause clipping but not to low where some of the audio will not be audible. Proper gain structure is a happy meduim between both of those...It depends really on what medium you are using. If you are using digital than defienlty keep it well below clipping. on analog you can acutally "hit" the tape with a lot of signal (and for lack of better words) clip the tape which is called saturation and is a desired effect. do some experimenting. you will find out fast in the mix if you have your desired recording levels. Cheers Ian

 

[Seeker of Rock]

Generally speaking, the stronger you can get the signal to record without overloading the better. You set the actual volumes of each track during mixdown. Until then, record each track as loud as the next.

 

[jonhall5446]

Generally speaking, the stronger you can get the signal to record without overloading the better. You set the actual volumes of each track during mixdown. Until then, record each track as loud as the next.

I agree, It's easy to lower a fader, but if you record it too low you might not get all the good stuff

 

[crankz1]

This all depends on the quality of your soundcard. If it's a 16 bit card .... by all means record as loud as you can without clipping. This will give you the most dynamic range of the 16 bits. With a 24 bit card you can safely record slightly lower levels and still maintain a good dynamic range.
Main point is to record at levels that use as much of the bit depth as possible without clipping (going over 0db).
Whatever program you are using to do your multitrack sequencing should give you the ability to adjust track volumes for playback. This would be the very most basic form of creating the mix.

 

[fenix]

Generally speaking, the stronger you can get the signal to record without overloading the better. You set the actual volumes of each track during mixdown. Until then, record each track as loud as the next.

Not necessarily. Depending on the preamp, your peaks at 0db on digital meter wil be somewhere around +10 to +20db on your preamp. So if you want to keep it clean as possible, you have to record a little softer. Everything is a compromise in audio.

 

[Seeker of Rock]

There are always exceptions, but again, GENERALLY SPEAKING, record your signals normally (you may end up having to re-track if you don't) and set your volumes at mixdown.

 

[Cloneboy Studio]

- roughly as loud as they will appear in the final mix (so, a background track would be recorded at lower volume than a lead track)????

NO! This is what I have termed 'mixing while recording' and it's a mistake I see a lot when I mix other people's material. If you record too low you don't capture enough signal versus the inherent amount of noise to get a clean recording, if you set it to loud you risk clipping and loss of heardroom.

When you are tracking--TRACK; when mixing... MIX. Don't try to put the cart in front of the horse.

- or record roughly all tracks at the same level (-6 db) independently if they will be louder or less loud in the mix?

Yes record them at the same levels. I aim to get a -6db peak with a rough signal average no lower than -15db when I record to 24 bit digital.

 

[chessrock]

Recording engineers use something called Gain structure. Proper gain structure will put your level where it is not too hot to cause clipping but not to low where some of the audio will not be audible.

This is a very good observation.

I'll take it a step further and point out that good recording engineers will use what is called Optimal Gain Structure. This is kinda' what sets apart the pros from the pretenders and wannabes.

When you hear phrases like: " ____ (insert name of recording engineer) could record in your basement with your gear and get a better sound than you could if you were to record at Abbey Road (or similar studio)."

Basically, the idea behind the theory is that just about any piece of gear nowadays, no matter how modest/cheap . . . as long as it isn't broken . . . is capable of very good results if used within it's limitations. Take a Mackie mixer for example. I've found that their mic pres sound quite impressive . . . until you start cranking the gain past 12 or 1'oclock or so. Then they kinda' start sounding like poo. Armed with this bit of knowlege about it's gain structure, I know that it might be a good idea to avoid them on quiet sources that require loads of gain. There -- See? I just used my noggin' for the betterment of my recordings.

Another example would be the Sytek MPX-4A, which is at the very opposite end of the ideal gain spectrum, so to speak. Once you play with one long enough, you start to realize that they actually sound better as you start cranking them past 4 or 5. Lack of headroom at the input seems to be their only downfall, so if you can learn to work within that particular limitation, you've got what amounts to be a world-class mic pre for 200 bucks a channel. And I really don't think that's an exaggeration, either. If you have an ounce of patience to learn it's gain structure, you've got a world-class mic pre in the right hands.

The idea is that everything out there should have it's sweet spot where it will sounds best. Experiment enough and find where those sweet spots are, and you will become a recording legend in no time at all. I say that jokingly, but actually it may not be too far off from the truth. Understanding gain structure can most definitely and without question save you a lot of cash money $ $ $ $.

 

[mshilarious]

Not necessarily. Depending on the preamp, your peaks at 0db on digital meter wil be somewhere around +10 to +20db on your preamp. So if you want to keep it clean as possible, you have to record a little softer. Everything is a compromise in audio.

I presume you are talking about clipping the preamp vs. the converter?

If your preamp clips at +20dBu and especially +10dBu I think it's time for a better preamp.

 

[fenix]

I presume you are talking about clipping the preamp vs. the converter?

If your preamp clips at +20dBu and especially +10dBu I think it's time for a better preamp.

i'm simply pointing out that on average, you will clip a converter when the preamp is around +18db. In order to get a cleaner signal and still retain resolution, you must compensate a little.

 

[Blue Bear Sound]

NO! This is what I have termed 'mixing while recording' and it's a mistake I see a lot when I mix other people's material. If you record too low you don't capture enough signal versus the inherent amount of noise to get a clean recording, if you set it to loud you risk clipping and loss of heardroom.

When you are tracking--TRACK; when mixing... MIX. Don't try to put the cart in front of the horse.



Yes record them at the same levels. I aim to get a -6db peak with a rough signal average no lower than -15db when I record to 24 bit digital.

I disagree -- first of all, with digital recording at 24-bits, you don't have to be anywhere near -6dBFS to obtain a high-quality signal. So this means you CAN set your levels much closer to where you want them in the final mix and not worry about sacrificing any sonics.

There are solid reasons for "mixing while you track" - not the least of which is that depending on how you're mixing (analog/digital/etc), your console's tape returns may not like a very hot signal such as those peaking at around -6dBFS or higher.

Bottom line is that you have to know your rig - the gain structure, you own working methodology, and best practices for the signal chain involved. What's appropriate in one case may not be in another depending on these variables.

 

[Cloneboy Studio]

-6db is peak, the average signal levels are a lot closer to -9 to -12 db for a target. My main argument is that it is easy to decrease gain and get a good sound than it is to increase gain if you need more signal during mixing.

 

[Bulls Hit]

When I'm recording, I hit the drums as hard as I like, and I crank the amp as loud as is reasonable.

This lets me turn the preamps down, and so maximise the signal to noise ratio

 

[fenix]

When I'm recording, I hit the drums as hard as I like, and I crank the amp as loud as is reasonable.

This lets me turn the preamps down, and so maximise the signal to noise ratio

huh? amp?

 

[Bulls Hit]

Guitar amp.

Heh, not drums

 

[mattamatta]

I say mic a drumset in a huge room, play it loud, pick it up with a pair of ribbons in the back call it a recording...

Probably dont even add any other tracks...

Yeah, clients/"clients" will LOVE it!

 

[Cloneboy Studio]

When I'm recording, I hit the drums as hard as I like, and I crank the amp as loud as is reasonable.

This lets me turn the preamps down, and so maximise the signal to noise ratio

I'm not sure if this is as sound a principle as you may think. For one guitar amps produce a lot of noise in the last 10% of their volume travel. Hitting the drums hard is a good idea but not because of lack of preamp gain, but the real tone of the drums become exposed.

 

[mattamatta]

Hitting drums loud is fun!

My drummer friend and I were laughing about how we'd prefer to jam with my girlfriend on drums (and him on bass or guitar, we all play all the instruments these days) than our other friend, even though he has more experience (I just tought her some basics the day before), because she "rocks harder"...

It was said that it's as if she's taking out her rage upon his drumset... in a musical manner of course.

Moral of the story is, it's just plain fun to hit anything loudly with a stick!

 

[Blue Bear Sound]

-6db is peak, the average signal levels are a lot closer to -9 to -12 db for a target. My main argument is that it is easy to decrease gain and get a good sound than it is to increase gain if you need more signal during mixing.

No argument on that point!