Author Topic: What are the best settings to normalize? (Read 5981 times)

phishn · « **on:** October 19, 2003, 12:37:08 AM »

Hi all,

What is the best percentage to use to normalize? Should this be done independently with the channels or together?

Sean Gallemore · « **Reply #1 on:** October 19, 2003, 02:18:13 AM »

do the channels together at 100%/0dB

dklein · « **Reply #2 on:** October 19, 2003, 04:03:11 AM »

...unless your levels weren't even to begin with. You want the average levels the same. Sometimes just going with peaks is bad for independent L/R adjustment. Most editing software can do 'analysis' and tell you your average power levels - run it on a section with just music (so you don't get thrown by the loud clapper on your left or right). If they're within 1 dB, I'd lock L/R together. If one channel is consistently louder, this is your chance to restore the balance and image to the centre.

gonads · « **Reply #3 on:** October 19, 2003, 04:02:19 PM »

I'd suggest getting good levels in the field.

dklein · « **Reply #4 on:** October 19, 2003, 05:00:41 PM »

gonads · « **Reply #5 on:** October 19, 2003, 05:13:03 PM »

Quote from: dklein on October 19, 2003, 05:00:41 PM

sorry but I don't like to mess with audio in a computer anymore than is absolutely necessary, things start to sound unnatural.

what is so hard about getting good levels in the field that everyone needs to normalize?

F.O.Bean · « **Reply #6 on:** October 19, 2003, 05:19:23 PM »

Quote from: gonads on October 19, 2003, 05:13:03 PM

Quote from: dklein on October 19, 2003, 05:00:41 PM

sorry but I don't like to mess with audio in a computer anymore than is absolutely necessary, things start to sound unnatural.

what is so hard about getting good levels in the field that everyone needs to normalize?

i get great levels in the field......it is said that if you record 48k in the field(as I do), then when ya resample to 44.1, you will have a better reproduction of the music because when ya resample, ya lose quality...

im a lil puffy, that might sound weird.... 8)but i read that in wavelab help

Sean Gallemore · « **Reply #7 on:** October 19, 2003, 05:19:57 PM »

this is true, I do it for other reasons tho

gonads · « **Reply #8 on:** October 19, 2003, 05:24:04 PM »

http://groups.google.com/groups?q=normalizing+group:rec.audio.pro&hl=en&lr=&ie=UTF-8&oe=UTF-8&group=rec.audio.pro&selm=84afro%24uhu%241%40nnrp1.deja.com&rnum=1

gonads · « **Reply #9 on:** October 19, 2003, 05:26:03 PM »

also:

In article <84a96q$qa7$1@nnrp1.deja.com>, MANDRAKE <rbcoen@my-deja.com> wrote:

> Shoud one normalize all his samples (peak value)?
>
> I remember craig anderton advising against this in a magazine, but
> can't remember why. Anybody does?

Normalizing has a place, but in general it's bad more often than good in
common practice. Every DSP operation can potentially add noise and
degrade audio quality, and add truncation distortion if not dithered.
Also, successive DSP processes really can degrade sound quality. On the
other hand, if you are doing a small gain change and dithering correctly,
and it's the last step of the process (besides dithering of course), that
is, not going to mastering, not having any other processing done, not had
a gain change done already, then it can be OK.

gonads · « **Reply #10 on:** October 19, 2003, 05:31:44 PM »

>1. What is so BAD about normalizing my individual tracks?

In itself: nothing. In the whole context of computer multitrack recording
and mixing, it may well be a step that is superfluous, and it's good
practice to avoid unnecessary processing, because it *might* degrade your
sound. (Yes, that can be fun too...)

>2. Would it not make sense to have full dynamic range at every stage?
>(example...come in hot as possible with slight outboard compression and then
>normalize internally each track)

Normalizing doesn't increase the dynamic range of a recording. Sure, the
peaks are upped, but so is the minimum level, including any
hiss/hum/whatever that was recorded. "Coming in as hot as possible" already
makes sure you get good dynamic range and use the full resolution of your
system.

gonads · « **Reply #11 on:** October 19, 2003, 05:34:19 PM »

all of these quotes are from rec.audio.pro and have been posted to share the opinions of some "pros" on this issue. Normalize if you want, they're your tapes after all, but I think it's good to know what you are potentially doing to the audio.

and you can roll your eyes all you want but if you get good levels in the field there is no reason to do potentially damaging dsp work.

just offering an experienced and alternative point of view... you don't have to like it.

F.O.Bean · « **Reply #12 on:** October 19, 2003, 05:39:06 PM »

i wasnt trying to say anything, just what i knew, thanks for shedding new light...

its a good thing i have flacs untouched of the 48k masters!!!

gonads · « **Reply #13 on:** October 19, 2003, 05:45:19 PM »

I wasn't you I was directing that at...

and flacs of your masters is a smart move... more people should be taking such a step before they engage in dsp work that they don't really comprehend.

F.O.Bean · « **Reply #14 on:** October 19, 2003, 05:53:31 PM »

Quote from: gonads on October 19, 2003, 05:45:19 PM

I wasn't you I was directing that at...

and flacs of your masters is a smart move... more people should be taking such a step before they engage in dsp work that they don't really comprehend.

no doubt

Simp-Dawg · « **Reply #15 on:** October 19, 2003, 06:20:32 PM »

Quote from: gonads on October 19, 2003, 05:45:19 PM

I wasn't you I was directing that at...

and flacs of your masters is a smart move... more people should be taking such a step before they engage in dsp work that they don't really comprehend.

regardless of how i do in the field (getting better after 1.5 yrs at it) i always keep an untouched 24bit/48khz flac fileset just in case....plus if i do have to do any processing on the file it gets done in this realm and thne later gets resampled/dither so all processing is done at the highest possible accuracy.

mterry · « **Reply #16 on:** October 19, 2003, 07:07:22 PM »

Quote from: gonads on October 19, 2003, 04:02:19 PM

I'd suggest getting good levels in the field.

what a cocky ass comment, hehe, I love it!

phishn · « **Reply #17 on:** October 19, 2003, 08:14:36 PM »

"I'd suggest getting good levels in the field. " First off, I'll be running a DMIC-20 if I can ever get the damn thing to work. The DMIC-20 only adjusts by 5db so there may be cases where the levels aren't perfect. I know a Todd Mod may help but I dont dare screw with the unit since its not functioning properly untouched.

Ed. · « **Reply #18 on:** October 20, 2003, 05:36:54 AM »

what would be better, normalizing your recording or upping the volume a few db. considering both are bad, but realizing you need to do something because you didn't get good enuf levels in the field. i usually normalize, that way i know nothing got clipped, but would the volume increase be better?

ed

dklein · « **Reply #19 on:** October 20, 2003, 09:13:56 AM »

Re: rolling eyes...

Quote from: diskobean on October 19, 2003, 05:39:06 PM

i wasnt trying to say anything, just what i knew, thanks for shedding new light...

Quote from: gonads on October 19, 2003, 05:45:19 PM

I wasn't you I was directing that at...

It was me he was directing that at. It just seemed that anything I put into words may have been taken as hostile and I didn't want to get into it...but yes, of course you should try and get good levels in the field. I thought the questions was about what to do if you don't.

And if you're in the 24 bit realm, good levels become less of an issue. No point in risking clips with that much resolution available. You can be 12 dB down and still get an effective 22 bit resolution.

mterry · « **Reply #20 on:** October 20, 2003, 09:48:51 AM »

Usually in the field you can get pretty god leveles, but monitoring your levels is a hard task sometimes with souch loud music being played. There was only one time I didnt have to do any editing, but for the most part, editing is always the case

Chuck · « **Reply #21 on:** October 20, 2003, 04:50:01 PM »

I record 48 kHz to DAT. If I'm going to burn a CD, I resample in WaveLab to 44.1 kHz. I've found very few instances where digital editing was necessary.

I have noticed "audible artifacts' from digital editing. Sometimes it adds a harshness to the recoding that wasn't there before.

To me, the best thing is to make a good 48 kHz tape to start with. Make a bit perfect transfer to the HD. Resample using quality software only if necessary. Cut your tracks with CDWav and make SHN/FLAC files or burn some audio CD's from those WAVs.

The less you mess with it, the better it'll sound.