Record levels

Ya , so record the track..Don't go above zero.

set the limiters there? Them who? -12..and smash it against the wall?
I could record with -40 peaks, but that would mean more work down the line, and it would make setting monitor levels harder. So there's more to it than "don't go above zero."

Limiters? Why?
 
Lazer is the spinal tap of record levels. Limiters, compressors, hit it hard, then turn up to 11 - it solves all his recording needs. He's said this zillions of times. If it's quiet, it's bad. I get that, but my needs are very different. Some where between our two approaches is normal, maybe?
 
My apologies! I missed out a " ' " ... "Gnat's" A well known contraction here in Limeyland for "Gnat's Knacker! i.e. a microscopic level or quantity. Often 'cleaned up' to "Gnat's underpants".

Dave.
 
The (passive) pickups at 11, set the gain at the interface at just below solid red light. Then the characteristic growl of the GZR PUs is noticeable. It's there at lower gain settings, just can't hear it.
 
Last edited:
The (passive) pickups at 11, set the gain at the interface at just below solid red light. Then the characteristic growl of the GZR PUs is noticeable. IT's there at lower gain settings, just can't hear it.
Ah, in that case you are driving the FRONT END of the interface into mild distortion. Not a lot you can do about that and another interface probably wouldn't show the same effect. Do note however that this is an entirely ANALOGUE effect?

What make and model of interface please?
Dave.
 
I don't actually record at those levels. I keep it around -18dBFS, that's with the interface gain knob between 12:00-1:00.
 
No worries. I wasn't describing a problem with recording, just the behavior of the interface gain as it appears to bring out the characteristics of the EMG-GZR pickups when set just below red light. It's far from clipping, so no problems there. I get a similar sound running the bass through an Ampeg amp sim while the interface gain is set to produce -18dBFS at Reaper's fader meter.
 
Lazer is the spinal tap of record levels. Limiters, compressors, hit it hard, then turn up to 11 - it solves all his recording needs. He's said this zillions of times. If it's quiet, it's bad. I get that, but my needs are very different. Some where between our two approaches is normal, maybe?
1000

Youre just too loud... Now get outa here kid you got no future..Slackers!
 
My apologies! I missed out a " ' " ... "Gnat's" A well known contraction here in Limeyland for "Gnat's Knacker! i.e. a microscopic level or quantity. Often 'cleaned up' to "Gnat's underpants".
I know what a gnats underpants are. I meant what do you mean by "Digital recording systems sound exactly the same whether recording [a touch] under 0dBFS* or way down at -50. Obviously you have to increase the GAIN of the monitoring system to hear the same SPL" ?
What are you saying there ?
 
I know what a gnats underpants are. I meant what do you mean by "Digital recording systems sound exactly the same whether recording [a touch] under 0dBFS* or way down at -50. Obviously you have to increase the GAIN of the monitoring system to hear the same SPL" ?
What are you saying there ?
Ah, right. Well, digital recording is fantastically linear both in level and in level v frequency. That is it does not matter if your signal is in the -18dbFS peaking to -10dBFS area or -40dBFS peaking to -20dBFS , distortion is extremely low in both cases. Lower than anything else in the recording chain, especially speakers and their power amplifiers.

Of course, you get a lower output signal recording at -40dBFS and to hear the same level on speakers or cans as the -18dBfs level you need to increase the gain of the monitoring chain.

It is unlikely that anyone would choose to record at -40dBFS but should they do so it would not matter to the overall sound quality. This is of course in total contrast to analogue tape which is not only very non linear around -10VU and gets progressively worse at 0VU and beyond but also compresses (HF squash) as frequency rises and the lower the tape speed, the greater both forms of distortion. But I guess that is why some people love it!


Dave.
 
There is no prize for recording digital levels hot as you can before clipping, and potential for damage. Aliasing and distortion from intersample peaks sounds bad. Leaving headroom gives you a safety margin in case there's transients or something that suddenly gets a bit louder while you're printing the take. Plus, often tracks that are easier to mix. If you have to pull everything down at least 12 dB so you don't light up the master buss, it probably wouldn't hurt to track that much lower. Digital recording craps out when you peg it. Tracking way too low like -40 or 50 doesn't really have that kind of penalty even though it's not ideal.


bouldersoundguy said:
For the most part I want a given track to average -18 dBFS over time. That is, I want to see it spending roughly equal time above and below -18. That said, percussive sources are different. You have to set those by peak, keeping their peaks below 0 dBFS. I typically set them to about -12 dBFS with the expectation that they'll tend to hit a higher level at some point during recording. I'm fine with -6, and if it gets to -1 without going over, I call that a win.

As far as tone changing with record level, that's possible if it's due to something in the chain changing with gain. Most gear we use doesn't do that very much in the normal ranges we use, but it's possible.

I like the distinction between steady state and transient signals. I'd definitely approach levels differently for distorted guitar vs. drums.
 
"As far as tone changing with record level, that's possible if it's due to something in the chain changing with gain. Most gear we use doesn't do that very much in the normal ranges we use, but it's possible."

Possible yes BSG and possibly more common than we think?

First off, some high Z interface inputs use a discrete FET instead of the virtually transparent jfet op amp. Solo FETs are not very 'clean' but some manfctrs use one to proclaim "warmth" or "mojo" for their guitar inputs. Yer pays yer money but I want to make up my own mind thank you very much how I ***k up my guitar's sound! The FET will give different degrees of "mojo" at different signal levels.

Much is also made of transformer mic inputs but again they are a 'frequency/level sensitive' distorting device.

Transformerless mic pres are not always innocent either. Some have an HF 'droop' that intrudes into the audio band as gain is maxed out. Even some (most?) loudspeakers and headphones change character as the wick is advanced!

The human ear is very sensitive to tonal differences as absolute sound level changes (Equal Loudness Contours) but pretty poor at JUDGING absolute sound level. Any tests done for different digital recording levels must be very careful to match monitoring levels precisely and that is not at all easy.

Dave.
 
Just don't clip the input, and don't make the level so low that you can't even see the waveforms... you should be good. If you want to stick to the "keep it in the yellow", that's fine.
 
Just don't clip the input, and don't make the level so low that you can't even see the waveforms... you should be good. If you want to stick to the "keep it in the yellow", that's fine.
Yes, don't clip the input (!??) but "waveforms" don't matter. Most DAWs allow you to scale them. "keeping in the Yellow" is precisely what you DON'T want to do as it gives you next to zero headroom.
Most DAW meters start the Yellow portion at -9dBFS (my AA 1.5 does anyway. Samplitude pro X 6 has rather different meters that go past 0dB to show intersample peaks)

Dave.
 
Yes, don't clip the input (!??) but "waveforms" don't matter. Most DAWs allow you to scale them. "keeping in the Yellow" is precisely what you DON'T want to do as it gives you next to zero headroom.
Most DAW meters start the Yellow portion at -9dBFS (my AA 1.5 does anyway. Samplitude pro X 6 has rather different meters that go past 0dB to show intersample peaks)

Dave.

My Yellow is nowhere near that and even if your DAW can "scale" waveforms, most DAWs by default show them based on actual signal... therefore you wouldn't want them so low you could barely see them because your Signal to Noise Ratio would be quite higher than you'd want.

 
-12 to -16 is about where my tracks peak for tracking; I leave plenty of headroom. At 24bit/48, just relax and "save room for later."

If/when I run through my compressors when tracking, I decide how hot I want to go through them (they have tubes) and then the output into the recorder stays, as mentioned, between -12 to -16.
With the cheap electronics in my Tascam DP24SD, I don't push things; I just try to track at a relaxed sweet spot. The recorder is happier and so is the music.

C.
 
Back
Top