more Bob Katz

[dangeorge6]

In Bob Katz book on mastering he points out that many poor D/A converters in consumer CD players make 0 dBFS+ signals +3 dBFS+ signals, effectively overloading a perfectly good level. He suggests that recordists should "never exceed -3 dBFS on a simple digital meter" p64.

Do you guys abide by this? I've always been told that one should try to get as close to 0 dB without clipping. Help me out.

Thanks.

[paterno]

Originally Posted by dangeorge6

In Bob Katz book on mastering he points out that many poor D/A converters in consumer CD players make 0 dBFS+ signals +3 dBFS+ signals, effectively overloading a perfectly good level. He suggests that recordists should "never exceed -3 dBFS on a simple digital meter" p64.

Do you guys abide by this? I've always been told that one should try to get as close to 0 dB without clipping. Help me out.

Thanks.

OK. It's time to put that book away for now. Experiment and see what works best for you...

JP

[dangeorge6]

of course, experimentation would probably allow me to conclude that I should record close to 0 dB. But the argument, I think, is that since 0 dB clips on crappy consumer D/A convertors, one should compensate by recording at lower levels.

[paterno]

Originally Posted by dangeorge6

of course, experimentation would probably allow me to conclude that I should record close to 0 dB. But the argument, I think, is that since 0 dB clips on crappy consumer D/A convertors, one should compensate by recording at lower levels.

Yeah, but when you do the mix, and the A&R person or band hears it from the MP3 you had to give them so they could listen on their iPod, and it's not as loud as your competitors, and you don't get the gig because of it, you won't really be worried about whether or not it is clipping.

Coming from a music/recording school background myself I can appreciate where you are coming from, but getting caught up in these kinds of issues will just make your head spin. And with all due respect to the Bob Katz book, don't take it as the final word. Use it as a springboard to develop your own working methods...

John

[Andreas G]

Yes it's true, peaks at -3db is the save way to not provoke distorsions from the DA-replicators of consumer CD-Player.

Thomas Lund from TC makes a research in I think 2002/03 about this phenomen, especially with consumer CD-player. There was a big report in the German Studio-Magazin than.

In general it has to do with the samplingrate and the 0db fequency-peaks which are not even divisible to the samplingrate. For example 997Hz has no easy aquisition to 44,1khz. The replicator trys to complete the "sinuscurve" and this is in the +db aera.
(sorry, my tech english is horrible... but I attache you the report in German )

What you need to be save and to get close to the 0db is an interpolating peakmeter, like in TCs system 6000.

Andreas

[Jim Williams]

What's wrong with giving the tired mastering engineer a couple of db's to work with?

It's not like you're going to drop a bit or something.

Always amazes me when folks complain about the mastering process when they don't give them anything to work with.

Use some common sense.

Jim Williams
Audio Upgrades

[dangeorge6]

Thanks everyone: Two questions:

Andreas: according to Lund's find, do you think one should not accede -3 dB when tracking individual instruments or just not accede -3 dB on the master fader of a session?

Jim: I see what you mean. So you're saying that as a mastering engineer you try to boost the signal of the overall mix as close to 0 dB as possible? If so, you don't worry about possible distortions on consumer CD players?

Danny

[paterno]

Originally Posted by Jim Williams

What's wrong with giving the tired mastering engineer a couple of db's to work with?

It's not like you're going to drop a bit or something.

Always amazes me when folks complain about the mastering process when they don't give them anything to work with.

Use some common sense.

Jim Williams
Audio Upgrades

I'm not complaining. I'm just saying that there are more things to worry about than someone's CD player output occasionally clipping due to meter discrepancies. And unfortunately, whether any of us like it or not, decisions about mixes ARE made by how loud they are, and the output distortion on the listener end may be [with some types of music, [and for better or worse]] a contributing factor to the sound.

I like the mastering process, and I have never had any negative comments about the levels I send to them, which tend to more conservative than the you may have assumed with my previous post. My issues with mastering have always been about how the feel of the record changes after they go through their process. Yes there are technical concerns, but when you're not getting as excited by the mix after it's mastered, it's not a good mastering job.

Getting a feel for what the limits are will be much more valuable than sticking to some hard and fast rule that you read in a book. That was my point...

John

[jonnyclueless]

Originally Posted by Jim Williams

What's wrong with giving the tired mastering engineer a couple of db's to work with?

It's not like you're going to drop a bit or something.

Always amazes me when folks complain about the mastering process when they don't give them anything to work with.

Use some common sense.

Jim Williams
Audio Upgrades

Maybe it's my flashbacks of Bob Ludwig politely scolding me for not printing up to 0dBFS and wasting bits.

But I think for most people it's the A&R people. If another person's mix is louder thn they will percieve it as better, and you could potentially lose the gig. Part f the problem stems from people with no musical or audio knolege calling important shots.

[Reptil]

We need a new format. one with more headroom. not less. (not lossy)
24 bit gives us A LOT of headroom to play with. and a lot of dynamics..
we are waiting for something that won't happen, unless there is action,
from within the music business. (Isn't this what the AES is for??)
we need to educate people/consumers on what is dynamics.
this will not end the loudness wars. but it will help.
giving up on the development in audio is lame.

percieved loudness is cool when used as a stylistic thing.
but is not cool when your carefully balanced mix is smashed to bits.
in a "finalizer", by people listening to "reveals".
just before it is sent to the pressing plant.
and it doesn't help saying that they are messing up
because they have no idea what I'm talking about.
so yes, wrong people for the job.

don't get me wrong: dynamics squashing is not a bad thing when done with care and preserving what is there. but that is mastering, not messing up..

[u b k]

well, the L2 and every other digital brickwall limiter is a CLIPPING GENERATOR. it may be a bit more polite (pun!) than square-wave hard clipping, but it's clipping nonetheless.

most d/a's can handle +3 overs. the ones that can't are so cheap, so bottom-feeder, that anyone who is using them will in all likelihood not hear the difference.

i have cd's mastered at modern levels and they sound wonderful. it can be done.


gregoire
del ubik

[Andreas G]

Originally Posted by dangeorge6

Thanks everyone: Two questions:

Andreas: according to Lund's find, do you think one should not accede -3 dB when tracking individual instruments or just not accede -3 dB on the master fader of a session?

Hi Danny,

from a technical aspect... yes I think so, even with -3 dB there is a lot of dynamic room left at 24bit, no disadvantages here. And for the A&R people aspect... well that's your decision... why not a special presentions mix up to 0 dB or check your level with a interpolating peakmeter: if no overs appears... well than, make it loud.

Beside that, I wouldn't overrate this theme, there are tons of very loud mastered CD out there (succesful major CDs!!!) with lots of "overs" and nobody seems to care about distorsions, you only hear from time to time: "uhhh what a hard digital sound". The problem is, on one CD player the distorsion appears more on others less. sooo???

Andreas

[jdunn]

Originally Posted by dangeorge6

Jim: I see what you mean. So you're saying that as a mastering engineer you try to boost the signal of the overall mix as close to 0 dB as possible? If so, you don't worry about possible distortions on consumer CD players?

I think you misunderstood. What he's saying is that if you give the mastering engineer a mix that's already pegged at -0.1 dBfs, they cannot increase the level at all without peak limiting. The more room you give them, the more they can boost the level without flattening the peaks.

See, most people are still trying to use digital like analog. Slam that f@cking signal as hot as it will go every possible chance you get! Digital just does not work like that. Not if you want it to sound good that is.

[sleepwalker]

Originally Posted by paterno

OK. It's time to put that book away for now. Experiment and see what works best for you...

JP

In my opinion, these are the most unhelpful responses ever. That's like a teacher telling a student to quit reading art books and draw. Maybe it's a good idea, but it's useless advice.

[thethrillfactor]

Originally Posted by u b i k

i have cd's mastered at modern levels and they sound wonderful. it can be done.


gregoire
del ubik

Please name some of these if you don't mind.

I ask because i've heard nothing in the last 5 years(since 2000 when i feel things really changed) that i could say that about.

[lucey]

It may be true that some players clip but on the whole it's a moot point. The modern standard is a peak level that not only goes to 0dbfs but RMS levels above -10dbfs, some even -6 from the peak.

As a mixer, you need not worry about your peak level, anything peaking around -10 is fine in a 24 bit world and as a practical concern for mastering.

I'll assure you that a tiny percentage are printing at -3dbfs, like one in 10,000.

Mastering to -0.3 is possible but not -3.0.




((As an aside, the best sounding loud masters I've made from digital sources were uncompressed or barely compressed on the 2 mix and had lots of headroom in the mix. The masters were hot but the punch remained and the distortions were far better then slamming the mix aiming for levels and as Jim says, leaving nothing for mastering.))

[GilWave]

Originally Posted by sleepwalker

In my opinion, these are the most unhelpful responses ever. That's like a teacher telling a student to quit reading art books and draw. Maybe it's a good idea, but it's useless advice.

How can you say that? Paterno is spot-on.

Any book, even a reference manual, is a guide - period.

People used to say "if it ain't broke, don't fix it". Flood worked under Eno, and came away with the concept of "if it ain't broke, break it!"

You going to argue with Flood's success? Listen to "Zoo Station" on Achtung Baby - if Flood had "gone by the book" we'd never have heard that song that way, in fact many of the songs and productions that we hold dear wouldn't have happened if someone with an ounce of creativity didn't go against the grain.

-g

[paterno]

Originally Posted by sleepwalker

In my opinion, these are the most unhelpful responses ever. That's like a teacher telling a student to quit reading art books and draw. Maybe it's a good idea, but it's useless advice.

First of all, my comment is a little out of context because there was another thread regarding a quote from the previous page of the cited book right before this one.

Secondly, in my experience, you can get bogged down by reading everything about a subject and not actually applying it. I think it's pretty good advice. And I've had teachers say something very similar during my music school years -- ' Don't think - just play'.

Sorry you took offense...

john

[Ged Leitch]

if you take a look at some of your favourite CD's impoted as a wav into your DAW then you will see how hot those levels are.But does it sound bad to you when listening on a cd player? probably not since their your faves.
It goes to show that there is obviously a place for these technical concerns but in the end it all comes down to whether or not it sounds good to you.
.I think the music is the most important aspect - not whether the CD player your listening on is occasionaly clipping.
cheers

[eyesore]

This thread doesn't make any sense. What does a CD player's ability to playback certain levels have to do with tracking levels? This implies some sort of digital 'memory' that doesn't exist. The player is only concerned with the level of that final master that it is actually playing - it has no way of knowing what level the individual tracks were recorded at and it simply doesn't care - unless they are exceeding that threshold in the final master which then would force the whole thing to exceed it.

Thanks.

[Youn]

eyesore,
I don't believe anyone mentioned tracking levels, perhaps you jumped in from another thread and didn't realize it?

You are right though, nevertheless...

[u b k]

Originally Posted by thethrillfactor

Please name some of these if you don't mind.

I ask because i've heard nothing in the last 5 years(since 2000 when i feel things really changed) that i could say that about.

i don't mind at all, i love turning people on to new sounds. among my favorites from the last 5 years:


zero7: simple things ('when it falls' sounds good too, but not tapey)
elbow: cast of thousands
radiohead: amnesiac
amos lee: amos lee
air: anything by air, but 'virgin suicides' sounds particularly lovely


enjoy!


gregoire
del ubik

[gwailoh]

Originally Posted by eyesore

This thread doesn't make any sense. What does a CD player's ability to playback certain levels have to do with tracking levels? This implies some sort of digital 'memory' that doesn't exist. The player is only concerned with the level of that final master that it is actually playing - it has no way of knowing what level the individual tracks were recorded at and it simply doesn't care - unless they are exceeding that threshold in the final master which then would force the whole thing to exceed it.

Thanks.

See this, this, and this.

[norman_nomad]

Originally Posted by dangeorge6

Do you guys abide by this? I've always been told that one should try to get as close to 0 dB without clipping. Help me out.

Thanks.

You don't need to record as "close to 0db without clipping". You won't lose fidelity by leaving some headroom in a 24 bit world. Tracking with peaks reaching -14 to -6 dbfs is a good practice.

Leave your mastering engineer some headroom!

That Bob Katz book is great...written in a simple language and fairly practical... I read it through a couple times when I first started getting into recording ... are you finding it useful?

[jonnyclueless]

If you print up to 0, they can just as easil bring it back down and you get the same result as printing at -6, only you can have it either way. I've printed with headroom only to have the mastering engineer ask why I didn't bother to print at full level or use compression. I think you're fine either way.

[jayfrigo]

A couple of points - Know the rules before you break them. Take all that info from books like Bob's and get an understanding of how things work, and then you can make an informed decision when and to what extent to ignore the "rules" so as to accomplish your particular goals. The info can't hurt, as long as you remember to use your own head and those two things on either side of it.

And as far as the other Bob telling you to "use all the bits," what was that, like 1986? Back when we only had 16 bit converters that maybe provided 14 bits of useful resolution, that was good advice. With modern converters, it really isn't anymore, and for many years Bob L has not been saying such things, at least when I've heard him speak.

[u b k]

Originally Posted by jonnyclueless

If you print up to 0, they can just as easil bring it back down and you get the same result as printing at -6, only you can have it either way. I've printed with headroom only to have the mastering engineer ask why I didn't bother to print at full level or use compression. I think you're fine either way.

there is a difference, namely the behavior of the analog input stage of the a/d. every a/d i've ever worked with has a sweet spot where the input likes to be driven. less than that and all i do is record more noise, more than that and it starts to choke.

in general, to my ears, converters seem to sound the most open when peaks hit around -14, -10 at most. i have yet to hear any converter that sounds its best when levels are peaking past -6. it's not an abvious, distorted, clipping thing, it's a subtle thing, along the lines of changing the loading on a mic pre.

a game of inches, etc...


gregoire
del ubik

[jonnyclueless]

That probably depends on what level it's calibrated to. I have never had any problems myself like that, but if the source is straining to put out enough ignal, then i can imagine it not sounding as good at full scale.

[jonnyclueless]

That probably depends on what level it's calibrated to. I have never had any problems myself like that, but if the source is straining to put out enough ignal, then i can imagine it not sounding as good at full scale. But that's just me, calibration levels are the most important area to me (as opposed to just always using -18dBFS).

[Bob Olhsson]

FWIW Bob Katz is talking about recording and not mastering.

If you are recording a real performance, there is never time to experiment. I find I get consistently better results trying to never go over -6. In 24 bit recording I've found it's always better to error on the side of too low as opposed to too high.

Digital clipping doesn't sound that bad but subsequent processing can sound really ugly.