Normalize?

How important is this process? I'm working on a track as we speak, found a good sample: chopped it up, laid it out, then added drums (drum hits, not breaks...yet), now my original sample is lost in the mix. My concern is loss of dynamic range.

Is it a natural approach to Normalize everything?

i normalize everything and then adjust volumes later. i've never experienced bad results of any kind

I normalize most of my samples (usually to -1dBfs) to have a good signal to noise ratio on the sampler.
There are two different forms of normalize. The one I use is thee peak normalize. this process simply analyze the signal for the highest peak and raise the complete sample equally to the point where the highest peak hit the amp value you choosed. This process is free from quality loss cause its a simple equal addition on the signal.

The other one is a process more like a limiter. I don't use it cause it rellay changes the sample and therefore chaangess quality. I don't know exactly how it is calculated but the inteenet knows

When I sample/record I use 24bit and raise the record level so the levelmeter peaks at something like -12dBfs to -7dBfs.
After the recording I cut away all the stuff I dont need and normalize to -1 dB. Then I change bitdepth to 16bit. By doing this I get the most out of the 16 bits.

peace

peterpiper wrote:I normalize most of my samples (usually to -1dBfs) to have a good signal to noise ratio on the sampler.
There are two different forms of normalize. The one I use is thee peak normalize. this process simply analyze the signal for the highest peak and raise the complete sample equally to the point where the highest peak hit the amp value you choosed. This process is free from quality loss cause its a simple equal addition on the signal.

The other one is a process more like a limiter. I don't use it cause it rellay changes the sample and therefore chaangess quality. I don't know exactly how it is calculated but the inteenet knows

When I sample/record I use 24bit and raise the record level so the levelmeter peaks at something like -12dBfs to -7dBfs.
After the recording I cut away all the stuff I dont need and normalize to -1 dB. Then I change bitdepth to 16bit. By doing this I get the most out of the 16 bits.

peace

Cool mate, sounds like your doing most of your processing on a DAW as opposed to the MPC. Before hardware I was recording everything into Pro Tools, but now I'm trying to do as much as possible in the MPC before mix down in Pro Tools.

A quote from Bob Katz - Mastering Audio:

"Technically speaking, normalization adds one more degrading calculation and level of quantization distortion. And since the material has already been mixed, it has already been quantized, which predetermines its signal-to-noise ratio - which cannot then be further improved by raising it."

Based on that, I'm thinking to stay clear of normalizing and approach the whole thing with my audio engineer hat on.

astral_Signal wrote:
A quote from Bob Katz - Mastering Audio:

"Technically speaking, normalization adds one more degrading calculation and level of quantization distortion. And since the material has already been mixed, it has already been quantized, which predetermines its signal-to-noise ratio - which cannot then be further improved by raising it."

Based on that, I'm thinking to stay clear of normalizing and approach the whole thing with my audio engineer hat on.

If he relly ment that peak normalizing degrade the sound he is wrong. At least in my case. I can normalize a recorded 24bit sample a 100 times from "normlize to -126" or whatever back to "normalize to -1" and it still will sound the same. No dynamic change, no extra noise, of course no aliasing.


peace

EDIT: I probably should add that the calculation in my DAW (Audition) works with 32bit floating point. If I normalize to -300dB for example and save it as 24bit (type 1), reload it and normalize to -1dB I got nothing but noise. But when I save in the default WAV format (32bit float) I get still the same clean sample after reload & normalize.

So I have no doubt that Mr Katz knows what he is talking about. You just can't use his statement for my post In which context did he say that?

peterpiper wrote:

astral_Signal wrote:
A quote from Bob Katz - Mastering Audio:

"Technically speaking, normalization adds one more degrading calculation and level of quantization distortion. And since the material has already been mixed, it has already been quantized, which predetermines its signal-to-noise ratio - which cannot then be further improved by raising it."

Based on that, I'm thinking to stay clear of normalizing and approach the whole thing with my audio engineer hat on.

If he relly ment that peak normalizing degrade the sound he is wrong. At least in my case. I can normalize a recorded 24bit sample a 100 times from "normlize to -126" or whatever back to "normalize to -1" and it still will sound the same. No dynamic change, no extra noise, of course no aliasing.


peace

EDIT: I probably should add that the calculation in my DAW (Audition) works with 32bit floating point. If I normalize to -300dB for example and save it as 24bit (type 1), reload it and normalize to -1dB I got nothing but noise. But when I save in the default WAV format (32bit float) I get still the same clean sample after reload & normalize.

So I have no doubt that Mr Katz knows what he is talking about. You just can't use his statement for my post In which context did he say that?

"...since the ear responds to average levels, and normalization measures peak levels, the result can totally distort musical values. A ballad with little crest factor will be disproportionately increased and so will end up louder than a rock piece with lots of percussion!".

you guys confuse normalizing with compression it seems.

**** wrote:you guys confuse normalizing with compression it seems.

No, they aren't. Both techniques involve volume and impendence of the signal being affected. But they are wholly different.

Metatron72 wrote:

**** wrote:you guys confuse normalizing with compression it seems.

No, they aren't. Both techniques involve volume and impendence of the signal being affected. But they are wholly different.

I normalize most of my samples (usually to -1dBfs) to have a good signal to noise ratio on the sampler.

normalizing AFTER sampling/recording doesn´t improve the s/n ratio.

**** wrote:
normalizing AFTER sampling/recording doesn´t improve the s/n ratio.

Yes s/n was the wrong term. Its more "I normalize to use most of the dynamic range without distortion / quantize noise." I just tried to explain it in english but I erased it cause I thought its to much blabla and can't be understand the way I described it.
Ok. what I really want to say is: normalizing a 24bit sound (and then converting it to 16bit) is better than normalizing a 16bit sound.

peace

Rather than reading articles on it. Use normalise the next time you sample some shit. If it works for you, great, if not at least you tried.

Different samples give different results in my experience.

mastasteez wrote:Rather than reading articles on it. Use normalise the next time you sample some ****. If it works for you, great, if not at least you tried.

Different samples give different results in my experience.

Different samplers as well.
Seems like some samplers are designed for you to use normalize after you sample to get the best results.