Discussion in 'Working with Sound' started by krakdhaus, Oct 31, 2018.
...asking for a 'friend'. Any non-obxious benefits?. Downside?
Normalizing is nothing special. If you want your tracks hitting hot levels then go for it, otherwise don't normalize and use clip gain to achieve the desired levels on the meter. I generally never normalize tracks but use clip gain and try to get -12dBFS to -18dBFS on each individual tracks.
The benefit to this method is that you have plenty of headroom to work with and also the analogue modeled VSTs are made to work effectively in this region.Hope this answers your question.
Use audio vitamin structure, it is an automatic normalizing vst which also gives lots of headroom.
I don't use it, since I'm older and come from an era when it was a destructive operation to normalize. I rather use proper gain-staging.
Normalizing what? The recorded single tracks (channels) or the final mix/master? And to what level?
Actually normalizing means nothing more than to adjust a track's peak level, not a destructive operation (nowadays? @Baxter). You can normalize to 0dBFS, -6dBFS or any other value below 0dB.
For single tracks/channels it depends on your workflow. If you use a lot of analogue emus 0dB is not a good idea (too high). The final mix/master I would normalize to -1dB (LUFS standard).
Last edited: Oct 31, 2018
Never, ever, normalize. If you're using EBU standards your tracks should peak at -6dBFS, and if you're mastering you should probably be limiting. Normalizing is identical in every respect to a simple change in gain, and offers no meaningful benefit.
None of this makes any sense to me.
If you want to normalize, be it a single track or a full mix, do it. It means the software automatically ups the volume of the track so the highest peak hits 0db. This is very handy in many situations. I have no idea why this should never ever be done like you say. I have never seen anyone making this claim before. This supposed -6dbfs ebu standard is also news to me. -6db = 50% less signal. This is not how media is delivered on CD's, on online streaming services or really anywhere.
One thing I will say. More often than not the end product is not delivered at 0db peak level either.
I've seen people use -0,1, -0,2 and even -1db settings when printing. This is mostly because DA converters are not perfect and there is a theoretical possibility that something somewhere someday will fuck up something at -0db. I tend to assume this is mostly superstition, but I follow the practice too, just in case.
OP should clarify his/her question.
Riot7 in which genre you are mixing...or do you talk in common..?
he meant -6dbfs for the mix before mastering
do not normalize or do it only the sample dynamics has been processed correctly
pls repeat ...i dont understand... m9cao no offense what do you mean?
edit: ahh now i get it....had similar thought too...
Last edited: Oct 31, 2018
I'm talking ITB. If you are hitting your master bus plugins too loud, just trim the gain down at that stage when you need to. On a modern digital system there's generally absolutely no need to fiddle around and worry about staying at -6db before that, just make sure you are not feeding your DAC too hot and you're fine.
However I've seen studios requesting -12db etc. files. It has never made any sense to me. It's an unnecessarily inconvenience to the client / contractor. Yeah, great, you have fancy analog gear and you need to worry about hitting it with the right voltage. You also no doubt have a perfectly fine volume knob and meters on whatever machine you are playing the files on with. Use it. Don't make your client do it for you.
thx for sharing
In fact it's not theoretical at all. There are real existing inter sample peaks which can be 3dB higher than the peak level and LUFS True Peak measuring is for showing these. It's also the reason why some limiters offer ISP option (although I know of limiters which have this option but don't work properly).
I can only agree with this. Although there are some good reasons to keep the RMS level at -18dB (EBU) or -20dB (SMPTE), every professional working mastering engineer is checking this level and adjusting it anyways. In fact different meters show different RMS values, so there's no reason to bother the client with such demands (apart from staying some dB below 0dBFS to prevent these ISPs, as long as it is no floating point file).
It was destructive in the 16bit days (96dB dynamic range). ofc now with 32bit float internal processing (and export) you have 1680dB of dynamics. But I still don't do it (old habit). I use gain-staging...because I'm a pro.
Edit: Vertical waveform zoom, yall.
Last edited: Nov 2, 2018
Because it does nothing useful. It looks good, it often sounds good, but does nothing useful.
At mixing you should be working to -6dBFS peak / -18dbFS RMS to leave headroom for mixing. If you don't and normalize to 0dBFS all you'll end up doing is either pulling the fader down to bring the track back within the needed headroom or loosing low end bits by over summing into the floating point. So why do it? Better to put a glass ceiling limiter at -6dBFS and level to -18dBFS RMS/VU and start with a good 21bits of sound properly populated for the mix.
At mastering you shouldn't aiming for 0dBFS as you're ignoring intersample peaking and running too close to the digital headroom of your final format. So why do it? What does it get you?
Normalizing is a bit like wearing a St Christophers whenever you go flying. It feels good but it makes no difference at all. But hey, if you insist on wearing one, I don't care.
I normalize anything, since it is much easiert so see whats going on. (using Maschine)
when recording vocals normalizing is not an option for me,
for the simple fact that normalizing makes diffrent takes less coherent,
makes sense if you think about it, the peaks are not the most notible factor about volume.
what would work is some kind of dynamic range normalizing, but I have not seen that to be done by any daw.
the normalize knob does one thing in my projects, rising the volume.
When the regular volume knob won't go any further....
how would these be done ?
isnt gain staging sort of dynamic range normalizing ?
This doesn't make much sense to me either. I go in the red all the time, yet nothing is clipping unless I want it to. I also often for example use samples that have really low volume in my projects. So I normalize them when I bring them to my project. Either that or I would need to perhaps completely crank the sliders, maybe even add extra gain from somewhere else.
Sure, after default normalize function the sample may be now too loud. And you are right, then you have to move the sliders down.
You are acting like moving the mixer faders is a giant hassle. But I do it all the time. I recommend everyone doing any kind of production work also use the sliders. And since I'm not dealing with samples that are dynamically accurate representations of the big bang, I'm not at all worried I lose any kind of dynamic range even if I fiddle with the sliders and create new busses all day.
Also, I specifically said I'm not printing to 0dbfs. You seem to have missed it.
This is true. Inter-sample peaks are kind of a red herring though, if you ask me. Most DAC's can deal with them and even if they can't the distortion should be inaudible and even if it's not, you are probably nevertheless still not not delivering your stuff at -5dbfs or however far the absolute highest 'true' peaks may reach. AFAIK no-one really does that, at least deliberately. It seems to be a non-issue. And if we are talking about pop music it's usually already limited to a point that would make an audiophile cry if he knew. (and of course the peaks don't actually exist in the digital domain)
Last edited: Nov 2, 2018
Separate names with a comma.