Overall approach / Normalize

  • Thread starter Thread starter gibsonsgharp
  • Start date Start date
LOL, Massive Mastering, you are completely correct!!!

Mixing in Reaper with levels from Reason and right off the bat, mix seems to have a lot more "air"

I actually didn't know there was a sound difference, as long as nothing clipped :confused:
I just assumed it was a bad idea due to the amount of potential problems
it could cause in relation to headroom and clipping :eek:

Learn somethin new every day :rolleyes:
 
Never had any major clipping problems. Usually normalized all files to -3db. But at the same time, I sometimes had intermittent EQ clipping issues when freqs were cut/gained. Usually just turned down output stage a few db. Unsure whether or not overall sound quality is better, but seems to give me more "headroom" for eq/comp. Common sense things, but when you learn or are taught a different way to do things, you just go blind...
 
Because I have made it a personal practice not to mix above 0bd on ANY fader.

I don't quite understand.

So you raise the levels of your tracks through normalisation so that you do not have to raise them with faders?

Firstly, the gain you apply through normalisation is no different to that which you would apply by pushing the track levels up.

Secondly, why do you want the track levels up in the first place?

If you feel a track needs to have its fader pushed above unity to be at the right level for the mix then stop what you are doing, pull all the other tracks down and mix at a lower level so that you don't.

Then, raise the level of the mix later on.
 
Even -3... Two tracks hitting together at -3dBFS = CLIP. If you have stacks of tracks, you're in for a world of hurt.

-10... -12... Something within reason *maybe* - But still, there's no advantage to it.

But tracks are at different levels. There's nothing wrong with that - It's natural for that matter.
 
But tracks are at different levels. There's nothing wrong with that - It's natural for that matter.
And exactly because it *is* natural, you're going to wind up with instrument tracks at different levels anyway. It's called "mixing" :)

G.
 
Even -3... Two tracks hitting together at -3dBFS = CLIP. If you have stacks of tracks, you're in for a world of hurt.

This depends on the bus architecture of the DAW, it's not an absolute. Not that I condone normalization.
 
This depends on the bus architecture of the DAW, it's not an absolute. Not that I condone normalization.

yeah. along these lines, clipping in 32 bit floating point is far beyond 0dBFS. But yes, normalizing a track is a pointless step, but does no harm. You have pretty close to infinite headroom with 32 bit floating point architecture.
 
yeah. along these lines, clipping in 32 bit floating point is far beyond 0dBFS.

Also in the Pro Tools TDM systems 48 bit mixer. From their white paper:

"In the Pro Tools 48-bit mixer however, this is not the case. By shifting the original 24-bit word to roughly the middle of a 56-bit register, there is enough headroom to sum 128 full-code tracks at +12 without internally clipping, as well as enough low level resolution to pull channel faders down to -90 dB without losing any of their original bits."

http://akmedia.digidesign.com/support/docs/48_Bit_Mixer_26688.pdf

Where you do have to watch out in PT TDM systems is how the 24 bit interface between plug-ins work.

As I have said before "know your DAW" as well as your signal chain.
 
Using floating point ans an excuse to exceed 0dBFS is a useless cheat. There is absolutely zero reason to ever have to go above 0dBFS, and if one is doing that it's likely because they've made a mixing or staging mistake, not because they actually have a purpose for doing so.

Chances are these days your project is going to wind up 16 bit in playback anyway, where 0dBFS is a wall. Respect the wall regardless of your DAW's capability.

G.
 
Back
Top