Bounce window to default to no normalizing

[jtees]

Im sick of reading "my song sounds too quiet when bounced" threads

[jtees]

But then again it would probably result in "my song sounds like its getting played through a guitar amp" threads

[David]

Agreed!! I've been lobbying Apple to default Normalizing to off ever since the feature was introduced. Doesn't make sense to have Normalize pre-selected in a pro app IMO.

[ski]

The CURSE of Normalizing. CURSE!

[Eriksimon]

Extra annoying is the realization that to default it to off probably just 1 byte out of the 789.355.082 (my) Logic Pro consists of, needs to be changed... if I only knew which one it is...

[Eriksimon]

Agreed!! I've been lobbying Apple to default Normalizing to off ever since the feature was introduced. Doesn't make sense to have Normalize pre-selected in a pro app IMO.

They could give it some sophistication by also being able to set a ceiling for it...

In fact, the code is already there...

pic 2012-05-26 at 09.18.19.PNG

[redlogic]

They could give it some sophistication by also being able to set a ceiling for it...

In fact, the code is already there...

Yep.

[jtees]

http://www.logicprohelp.com/forum/viewtopic.php?p=438173#p438173

Now I'm stuck in an anti-normalize by default infinite loop

[ski]

LOL!

ceiling... code...

Yep. But the reality of it is that when bouncing tracks, it's highly unlikely that you'd want their level to be arbitrarily made higher than the original source; but that's exactly what Normalizing does. So I would argue that even if a ceiling was put in place, the amount of level increase (or decrease?) that would occur for every bounce would be entirely arbitrary, and therefore useless.

Case in point: bouncing out "unity gain" tracks to give to a mix engineer, and here's what that means... I want every track I bounce out to reflect the level it's at in my arrangement. No louder, no softer. Then, when the engineer lays those tracks up on his/her system and puts all the faders up at zero, they're hearing my mix. Unity gain tracks (or "unity gain stems" when creating stems).

That's common practice.

But... if you were to normalize all of the tracks upon bouncing, regardless of any pre-set ceiling, every track I bounce is going to have its gain changed by an arbitrary amount. So when the engineer puts those tracks up at unity, the overall blend won't resemble the balance I created for my arrangement. Ergo, in light of how absolutely common this practice is, having normalizing on is entirely useless.

Now...

Let's say I have a few tracks that are so low in level that they need a bit of a gain boost just so that the engineer can see some semblance of a waveform! In those cases (and they do happen), I'll raise the gain of those tracks by an amount I determine (say, 6 dB). When I send the tracks to the engineer, I include a note that says, "All tracks at unity except tracks x, y, and z which need to be brought down 6 dB". Done.

[Eriksimon]

Yes, Ski, that all makes sense for a professional workflow, but don't forget many of us do it all ourselves, and never deliver anything to others. What I especially dislike about this "checkbox only" option is that the user has no influence over, and is not even informed about how much deciBel are going to be added or subtracted... that makes it "Expressy" or "Garagebandesk", not Pro-like.
But in the end, the Normalize function is already there (twice even, in the Wave editor, see my and Redlogics' post), so it may as well be removed from the bounce window alltogether.

[ski]

Yes, Ski, that all makes sense for a professional workflow, but don't forget many of us do it all ourselves, and never deliver anything to others.

Yes, but even if you don't deliver your tracks to other people, there's still no reason to have normalizing on all the time. Well, almost none... the only time I can see this being of use is if you're making sample libraries, and you want to bounce out a whole bunch of normalized regions for the purpose of having a consistent level between samples. Other than that, there are very few reasons to normalize anything, even for the non-professional.

What I especially dislike about this "checkbox only" option is that the user has no influence over, and is not even informed about how much deciBel are going to be added or subtracted...

Still, I have to ask, what purpose would it serve to set a fixed ceiling on the normalization process? Say you have two tracks, virtual instruments: electric guitar and synth pad. The guitar part peaks at -12 and the pad at -6. You want to bounce out those parts and bring them back into Logic as audio regions. So let's say you had a ceiling of -6 on the normalizing feature. After the bounce you would lose the difference in level between the guitar and pad because they're now both peaking at -6: the pad hasn't changed at all because it was peaking at -6 before the bounce, but the guitar is now 6 dB louder than it was before. See what I'm saying?

[Eriksimon]

Yes, Ski, that all makes sense for a professional workflow, but don't forget many of us do it all ourselves, and never deliver anything to others.

Yes, but even if you don't deliver your tracks to other people, there's still no reason to have normalizing on all the time. Well, almost none... the only time I can see this being of use is if you're making sample libraries, and you want to bounce out a whole bunch of normalized regions for the purpose of having a consistent level between samples. Other than that, there are very few reasons to normalize anything, even for the non-professional.

Absolutely; I fully agree with and understand all your points, let there be no mistake about that. I never use the normalize feature, because I know how to mix (close enough to) a given peak value anyway.

What I especially dislike about this "checkbox only" option is that the user has no influence over, and is not even informed about how much deciBel are going to be added or subtracted...

Still, I have to ask, what purpose would it serve to set a fixed ceiling on the normalization process? Say you have two tracks, virtual instruments: electric guitar and synth pad. The guitar part peaks at -12 and the pad at -6. You want to bounce out those parts and bring them back into Logic as audio regions. So let's say you had a ceiling of -6 on the normalizing feature. After the bounce you would lose the difference in level between the guitar and pad because they're now both peaking at -6: the pad hasn't changed at all because it was peaking at -6 before the bounce, but the guitar is now 6 dB louder than it was before. See what I'm saying?

Yes, I see what you're saying and I'm pretty sure you know that I already saw what you're saying before you wrote it down for me to read, right? I'm just being a bit "devils' advocaty"...

It can be handy in (only a few) other scenario's. What you outlined in the paragraph above is very clear. If I would do that, I would simply lower the Guitar fader with 6 dB... and then you could (validly) say "well, what's the point of that?" The only point I can come up with is that it is esthetically more pleasing to see a bolder waveform...

[ski]

Excellent! You fell right into my little trap! (insert evil mad scientist laughter here) <--- evil

Seriously tho, I deliberately left out a key factor... I gave you some facts, of one track peaking at -12, the other at -6. But let's say you were doing this on your own. How would you know that those were the peak levels for those tracks? Well...

You'd have to watch the meters for each track as it played down, making note of the peak levels attained for each one. Only by going through that (tedious) exercise would you know what your peaks are, and then you'd know how your normalizing ceiling of -6 would affect the levels after the bounce.

But...

Logic's meters aren't 100% accurate. They don't sample fast enough to catch every peak there might be. So really, the only way you'd be able to scientifically ascertain that -12 and -6 were the peak levels would be to bounce each track without normalizing, and then open those files in the sample editor. From there, use the Search Peak function or better yet, the sample editor's Change Gain function and then use its Search Maximum button to find the peak levels. Once you know what they are, you could set your normalize ceiling to a value of your choice and do the math afterwards to figure out what the difference in level is going to be between your bounced/normalized tracks.

But without being so intrepid as to go through all that malarky, any level you set for the normalize ceiling would result in an arbitrary amount of gain change for each track.

[Eriksimon]

Excellent! You fell right into my little trap! (insert evil mad scientist laughter here) <--- evil

Seriously tho, I deliberately left out a key factor... I gave you some facts, of one track peaking at -12, the other at -6. But let's say you were doing this on your own. How would you know that those were the peak levels for those tracks? Well...

OK, I must stop you there.
If they are anything else than pluginless audio, I'd bounce those tracks in place and check the bips' peaks in the Wave Editor! Ha! Bips' peaks! O yeah!

[ski]

LOLOLOL!

You get an "A+" on your bippy report card!

[jschmidtiii]

so...is there a setting you can change so that your bounces aren't normalized by default?

[jordito]

so...is there a setting you can change so that your bounces aren't normalized by default?

Just set the Normalize option in the Bounce window to Off. The change will be recorded in Logic's preference file so it will stay like that for each and every project unless you manually change it to something else.

J.

[TequilaKez]

I find it quite convenient personally. I always try give my tracks plenty of headroom and L2 on the master set to barely shave of a db or two for a good listening level. So %99 of the time it should never kick in (I'll bypass the limiter to bounce mostly). In that rare occasion where I've let something slide, there's that safety net there which avoids that email from the mastering engineer 'ummm.. you've got clipped peaks at blah blah'. It's a nice reassurance for bip's for the same reason. If had to choose between %1 of the time, a track being clipped or a track being slightly softer I choice the latter. There's the sound quality agrument but I've never been able to hear any difference other the volume..... small amounts of attenuation @ 24bit I'm talkin.

btw, mine has always defaulted to 'overload protection only'. Is that the same for everyone?