How do I get good audio recording level without clipping?

Cubase
1

Hi,

My problem is I’m trying to record a lead guitar part. Guitar’s going into a Focusrite Voicemaster Pro in order to use some light compression to tame peaks, then into a Focusrite Scarlett 2i4 which is my audio interface/preamp. I’ve used Cubase for years but am new user to Elements. Nearly every time I adjust the Voicemaster and Scarlett to abtain a non-clipping signal, the small input meter on the transport bar shows red clipping at the top of the bar for the first note played. Firstly, is this relevant? Because on the waveform after recording it doesn’t seem to clip…?

Secondly, and more importantly, could someone tell me how to get a reasonably loud recording (fat waveform!) without such high peaks? I know I could compress it really heavily prior to going into Cubase, but I want to keep the dynamics as real as possible and apply compression later (as is often advised).

I seem to either get a waveform that looks too narrow/quiet or a fat one that clips. Any help is hugely appreciated!

2

Not familiar with elements but if it’s like cubase, make sure your waveform zoom is sufficient that your not suffering from the illusion that the recorded signal is too small. Apart from that, I’d suggest you search for audio “gain staging” on youtube and watch some of the excellent videos and tutorials about this - that should help you best.

3

Recording at 24-bit, one way to do it is to do a rehearsal run of your playing into Cubase, and adjust the preamp trim so no peaks are greater than -12dBFS. Then after recording, you can always turn up the track’s volume fader if you want to to make it sound louder.

Like mentioned above, don’t worry about what the waveform looks like, you can make that appear bigger or smaller with the small slider on the side of the project window without affecting how loud it actually is.

4

Hi.

First off, you don’t look at the wfm, and use that as a guide as to whether or not you are clipping on the way in.

Secondly, I don’t know where this trim so no peaks are greater than -12dBFS comes from… :question:

Third:

the small input meter on the transport bar shows red clipping at the top of the bar

Are you talking about the Audio Activity (Default In Ch) to the right of MIDI In Activity? That’s just duplicating the input fader - which is really what you should be looking at.

Four:

Guitar’s going into a Focusrite Voicemaster Pro in order to use some light compression to tame peaks, then into a Focusrite Scarlett 2i4 which is my audio interface/preamp.

Typically, you go into the mic pre first, and then a compressor if you want and/or need to. I am not familiar with either of your two pieces of gear - is it not possible to go guitar > mic pre > comp > soundcard? That’s really what you want to do.

Yes, you want to try and avoid clipping. However, that’s not to say that if you see any red, that it’s all bad. Some may be ok - some. You definitely want to avoid constant red, but if you see a touch here and there, you are probably ok. I am not telling you to definitely go into the red, however. Red does not necessarily mean clipping. Red is a warning. Take it as such.

Again, I am not familiar with your gear, but do either have meters? Let’s assume so. You want your level to be as hot as possible (there are occasions you don’t, but generally you do), meaning that both pieces should have a good level. Again, going into the red a bit is most likely ok. If you see too much red, back off a bit. Next, you may want to have a look at the sound card’s control panel (sw) as there are probably adjustments there. Depending on how the driver was written, what you are seeing is pre-Cubase. Or at least, that is what you should be seeing. Next, you want to adjust the input fader in Cubase so that it too has a good, strong level, without too much (if any) clipping. If the meter goes up into the red (past 0) too much, back off.

That’s about it. Have fun.

5

Hi. …

Secondly, I don’t know where this > trim so no peaks are greater than -12dBFS > comes from… > :question:

It’s a fairly common suggestion that I find very helpful. Leaves lots of headroom for the inadvertent loud bits, yet with 24-bit recording sounds fine on playback when/if the volume fader is pushed up.

I also find it very helpful to keep from overloading some plug-ins that are sent to from multiple tracks.

Finally, I find it avoids having to pull the Cubase 2-bus master fader all the way down to keep from clipping on the way out.

I used to record the way you suggested, recording “… your level to be as hot as possible”, but ran into problems like these that I learned I could avoid by recording with peaks around -12dB or so.

6

Aloha guys and great thread! Some nice info being shared here.
Thanks for that.

In all my years of experience when it comes to stringed/wind instruments (of all sorts)
this issue lies not so much with the gear as with the player.

Not to hit this too hard, but some guys (gals) just have that ‘touch’.
Some players can perform extremely well on their ax and ‘nail’ all the parts but
still don’t have that ‘tone’, that ‘touch’.
And if they don’t have it, no amount of gear or tech will achieve that very special sonic goal.

This is not as big an issue when it come to synths/elec keyboards/plugs etc but
is with guitars. They are a different breed.

In this regard, knowing ‘when to say when’ can also be a very useful tool.
{‘-’}

7

If you use the 32bit floating format, the only places that you have to really worry about levels are:

a) Those going into the soundcard’s analog to digital converters (ADC).
With 24 bits available on most soundcards, one can afford to allow a reasonable amount of headroom. The levels would have to be varying wildly to warrant compression.

The main problem is that performers who are not used to recording will be more variable with their levels. They also tend be rather tame during level checks and let rip more when they get into the actual takes.

I get the performer to play the loudest part of the track and adjust the levels to just avoid clipping, then back off a tiny bit more to allow for their enthusiasm. It is a quick set of trial and error iterations.

However, if peaks are very occasional, I get the performer to try to temper those a little to enable the average levels to be hotter. For vocalists, it may involve getting them to lean back a little during those loud parts. The levels will be better, but the timbre will still give the ‘intense’ sound.

b) Those going into the soundcard’s digital to analog convertors (DAC).
A brick-wall limiter with a soft knee, or some sort of maximiser, at the final post-fader insert slot on the main output will prevent overloads here.

However, none of this is to say that levels shouldn’t be managed throughout, as at a minimum it allows wiggle room to make adjustments.

The basic takeaway here is that input levels don’t have to be taken to risk-of-overload levels. It’s the output levels that really need to be managed well.

I have separate group tracks that feed a master group that feeds the cue mix. That makes it visually easier to do a mix for the performer, but also put on any FX, like a reverb, or a low-latency compressor to even out an excessively varying track so a performer can hear it properly in their mix.

8

You could:

levels.jpg
levels.jpg898×724 64.8 KB