If you've reached redline, you have clipped. There is no way to undo that. There are ways to mush it down so it's not as objectionable. Not an ideal fix. So this is the argument for recording at 24-bit (32-bit float). This way if you're recording levels are not as ideally high, your noise floor will be greatly extend. So no problem. If like myself you are stuck on working in 16-bit, levels have to be more greatly optimized. So when the peak light comes on, peak does not always mean at its best, which can be misconstrued in the English language as to meaning best or, being at ones peak. What it really means is that you have been caught running a red light, in this application.

I've talked about the use of some creative clipping but that's another conversation. Clipping in another English-language culture could also be indicating stealing and/or rip off. The rude experience of one being clipped. This has nothing to do with a bad haircut. It's accomplished after-the-fact also when you do have a clean track to work from first.

I'm clipping coupons. Is that bad?
Ms. Remy Ann David

record in 24bit and stay away from 0db and you'll be okay!

Well, as you can see, there are brave souls like Remy around who record in 16bit. While you are recording 24 bit you can have your peaks at -48dBfs and still have the same S/N ratio as someone recording in 16bit. So while that looks like a very low signal on your track, there is still a lot of musical information there - in 24 bit recording, there is a lot more room for low recording than you might think. I keep guitars and vocals well away from clipping. I'll push drums a little harder - as Remy says, the occasional clipped drum transient isn't going to kill a rock drum track.

Sweet 24 it is. I just realized I didn't clarify in my original post though. I was talking about the track volume level, the slider not the incoming signal.

If I set the slider to about -30dB will I have a problem getting the signal louder during mixdown? Or will it be the same as if I had it set at 0dB. Or if I set the slider to -30dB and set the signal volume accordingly, will that give me an extra 30dB to work with while mixing?

And what do you tend to use on certain instruments? I only ask this stuff because I lack sufficient preamps to get a signal loud enough to track at levels lower than 0dB. Unless I am totally off, in which case I am prepared to get owned :O

I can't remember for certain but I thought you were using a Presonus unit and Cubase LE. If that is the case where you set the sliders in Cubase does not affect the recording levels at all just the monitor mix. The recording levels are set by the preamp levels only, unless your unit is different than mine.

jg49 wrote: I can't remember for certain but I thought you were using a Presonus unit and Cubase LE. If that is the case where you set the sliders in Cubase does not affect the recording levels at all just the monitor mix. The recording levels are set by the preamp levels only, unless your unit is different than mine.

That sounds right to me. The way it should be. Faders only effect the master bus not the individual channel. Yes it will make the channel louder but only for the mix. The recorded volume is adjustable via the preamp only during tracking. Once the track is recorded, the faders come into play for the mix.

yeah I think you guys are right, the level slider is only for monitor level in this case and does not dictate the incoming signal/level in any way. I just A/B two tracks with same level going in and set one at 0dB and one at -30dB. While the level indicator did change during tracking, when I put them both up to 0dB during playback they were the same volume.

JG you are right about the PreSonus Firebox but I am on Logic Express, close enough my friend.

A couple quick comments since there always seems to be a bit of confusion on the topic - although much of the stuff already stated is right on.

1 - As already mentioned, the only thing that makes a difference as far as actual level coming into the machine is what analog signal hits the A/D converter and how that A/D converter handles it. If there is clipping at the A/D, there will be clipping in the resultant file. If there is no clipping, then you can do whatever you'd like to the fader and you'll be fine.

2 - Once the signal is in the computer and you start messing with the faders, you can push the faders on your virtual mixer (built into the DAW) as hard as you'd like presuming you're mixing on a 32 bit (fixed or floating) bus. However, be careful how you monitor. If your D/A converter (or soundcard) gets a signal that is beyond 0.0dBFS it will represent it as clipped to the analog signal and the sound will not be good and could potentially damage your equipment.

3 - If you push the faders this hard into the mix bus and then bounce down, you will get some indication from your DAW software that you've exceeded 0.0dBFS but will still likely have the information represented as being above 0dBFS. If you've done this, you should reduce the amplitude of the track until it is below 0dBFS. You can do this manually (drop the master fader down or each respective fader), automatically (normallizing - not a big fan) or by use of a digital limiter (not always a hot idea but not always a bad idea either - it depends on the material and how bad/often it's clipping).

Otherwise, if you bounce the track as is and render a 16bit or 24 bit file out of it, you will get digital clipping - you will not like this.

Personally, when I run into a scenario like this, I prefer to lower each individual track's fader until I get the mix that I like. I don't suppose there is a supportable argument for not using the master fader (despite my attempts in the past), so you can also use that tool as well. My biggest gripe with using digital faders is that to alter the amplitude, you're messing with the bits themselves. I'm really getting into using analog summing and mixing lately and have been thoroughly enjoying the results!

Cheers-
Jeremy

what i do in logic, if my master bus is clipping, is putting the "gain" plugin (found in the utilities) in an insert bus of the master track and reduce the signal in the plugin. this way you don't have to adjust each individual track in your project.

ouzo77 wrote: what i do in logic, if my master bus is clipping, is putting the "gain" plugin (found in the utilities) in an insert bus of the master track and reduce the signal in the plugin. this way you don't have to adjust each individual track in your project.

Couldn't you just lower the master slider the same dB? Or are you saying there is a benefit to lowering the gain? IDK, I try to not stress my CPU too much by putting unnecessary plugins.

it depends. if you use a limiter to catch occasional peaks pulling the slider down will not prevent the signal to overload the limiter. you'll just have the same "clipped" signal, just a few db's lower.
of course you could lower the input gain in the limiter itself (at least in most limiters) but for me it's more flexible this way.
if you don't use any limiting on your master bus you can use the fader.

i like to leave the master fader at 0db and reduce gain with the gain-plug (if needed). don't ask me why, it's just the way i do it.

and don't worry about cpu-usage. it's not even worth mentioning...

ouzo77 wrote: and don't worry about cpu-usage. it's not even worth mentioning...

Admittedly, it's one single multiplication on every sample that passes through that gain stage.

If the CPU usage is anywhere near that of a deep and thick reverb plugin with a long decay, something is broken.

BobRogers wrote: Well, as you can see, there are brave souls like Remy around who record in 16bit. While you are recording 24 bit you can have your peaks at -48dBfs and still have the same S/N ratio as someone recording in 16bit. So while that looks like a very low signal on your track, there is still a lot of musical information there - in 24 bit recording, there is a lot more room for low recording than you might think. I keep guitars and vocals well away from clipping. I'll push drums a little harder - as Remy says, the occasional clipped drum transient isn't going to kill a rock drum track.

While that is true in digital theory (144dB-96dB) recording at -48dBfs is definitely overstating the case in practise.

Most (decent) 16 bit A-D's have a dynamic range of around 90-93 dB while 24 bit A-D's are around 115-120 dB so the true improvement is usually more in the region of 25 -30 dB but NEVER 48 dB. This still gives plenty of extra scope to record at lower levels without losing dynamic range (or S+N/N if you prefer). I normally advise around -18 to 20 dB peaks if your source is likely to give big peaks. I don't think I've ever clipped when setting to around - 18 dB. Of course you don't have this luxury with 16 bit so it might be prudent to put a compressor in the record circuit (set properly of course) to eliminate occasional clips.