Gain Structure

Quite by happenstance, on a selected wave form in the editor, I dragged the wave-form itself upward inside the wave form bar. I noticed that the waveform shape increased inside the waveform bar from its current size to a waveform that was larger in amplitude. The results seem to look like a weak signal compared to a very strong signal without going thru the recording process again.

My question is this: What exactly is that process doing? Specifically, does it really remove the recording step if you have a weak signal. So if you recorded a weak signal, just stretch the wave form up and you’re done. To do that it must manipulate the signal to noise ratio. Otherwise, a constant signal to noise ration doesn’t provide a better product. Or is this just aesthetics! Thoughts?

Did the sound level increase? I think all that does is change the display scale of the waveform, rather than actually applying any processing to the audio file.

If you dragged up on the top part of a clip, you did apply gain. The amount of gain displays in the info line above the project arranger window.

It’s simply adding gain. Exactly as if you raised the fader, except this is all before the signal hits the channel.

The ratio of signal to noise won’t change, but the noise floor will rise.

It doesn’t replace getting a healthy level to tape, because, if you record too quiet, the nose will become apparent when you raise the clip gain. As always, one needs to get a nice, solid level to tape.

Thanks man. makes sense.

I prefer to normalize instead of using the clip gain. I received a project from a friend wanting me to do a quick mix. In that project there was a guitar track recorded at -25dB then clip gain 30dB up, and the fader pulled down to -15dB. So I got a guitar track playing that is clipping but nothing going into the red, and to make things even more difficult the clip was shared with 2 other tracks. I really wish every clip had an obvious indicator on it that showed the clip gain value in big bold red color :slight_smile:
It’s immensely useful but requires some discipline or you can get completely lost.

I didn’t initially understand what you were saying. But after some reading I think what you are saying is that you rather adjust the gain of the entire track as opposed to each individual clip. Is that correct?

correct,

Let me ask another way. If there are two clips (A and B) that make up an audio file on track C, assume also that clip A is 3 db louder than clip B, when C is normalized, what exactly happens? Does both clips A and B increase by a set amount of db irrespective of whether or not A is louder than B or does the normalization process take into account the difference in gain of A and B and only increase B to match A?

Some reading on the subject: http://www.learndigitalaudio.com/blog/normalize-audio

Ok good stuff. Good read. Clears up my questions.

Makes sense but never thought about it. Noted.

I will usually, in the project window, adjust the gain structure in this way so when it hits the compressor it’s already kind of evened out, makes it easier on the compressor. It’s nice also because you get a visual representation of what is louder or softer than the other parts.

I use the split tool with KCs for gain up/down, can work pretty quickly that way. There is a pencil tool for that purpose too, but I never could work smoothly with that.

There are some plug-ins that do this, Waves Vocal Rider actually being the only one I know that might work in Cubase. I was drooling for a few years hoping that Quiet Arts’ Wave Rider would be made available to Cubase users, but I’ve long since given up on them.

I’ve never checked how adjusting the gain of the clips might be same/different from automating the “Pre-gain” in the rack … anyone know?

I find myself doing that a lot too.

Wouldn’t the pre-gain have the same effect as adjusting the gain on audio events, but globally for the whole track?

The pre gain does not discriminate. The event boost is per event or per clip. If u select all events or all clips and boost behaves globally per track as you indicate.