[drtechno]

But that brings in the client side: Is the artist, producer, or label demanding this level of loudness? Where does the "dumba$$" factor really lay?

the mixer mixes, the mastering engineer makes it loud and does some fine eq tuning and roll offs. That a way the mix can stay at DR12-14 to be press correctly to an album. Otherwise, if you tried to cut the proof for a pressing with a digital delivery specified signal, you end up with low distorted audio from the cutting coil getting saturated with the signal.

Bottom line: the master engineer sets the loudness. And is allowed to work in the dumb-ass-zone to make the music work for the masses that have cheap audio equipment.

 

[gregorio]

My idea of 'using all the bits' implies keeping all or most of the audio signal as hot as possible, wih peaks at 0dBfs and average levels just below. Typical of modern pop releases.

Ah, I should have guessed you were back on your agenda again but then that is the exact opposite of what you are now stating! If you're talking about the amount of compression used in modern popular music (your agenda), then the lower the dynamic range, the fewer bits are effectively required, which is the exact opposite of "using all the bits"!

I know a rather well known mastering engineer who prefers truncation to redithering as he likes the “edge” (a.k.a. artifacts) it creates, allowing his product to stand out during distribution. I’m sure he’s not alone.

No, he's not alone, although I don't know many who don't dither because they prefer truncation distortion. Those I know who just truncated (don't dither), do it because they feel truncation distortion is inaudible, which generally it is. Certainly it's inaudible when truncating down to 24bit but it's usually inaudible when truncating down to 16bit as well.

Savvy engineers never do 0 dBfs anymore thanks to the almost universal adoption of (slightly) more conservative True Peak measurement standards, which prevent DAC overload on playback.

I'm not sure you should say "never", certainly TV sound engineers never do but then they never did, even before TP and loudness normalisation. Music engineers though, it depends on client wishes, I would always advise TP rather than sample peak if there's any possibility the master is going to be converted to a lossy codec at any point in time (inc. by consumers themselves) or if the TP is substantially higher than the sample peak.

I'll awnser this: We call those people mixing in the "dumb-ass-zone" (-4 to 0dbfs) . Master engineers bring it up full scale and into a limiter for the digital delivery that has poor signal to noise ratio. The dacs are so bad a reproducing the signal, we have to have the mastering engineer make it loud. Is that where we mix at? no! -12 to -10dbfs for most.

If you're going to "answer" something in this particular forum, then you EITHER have to be agreeing with the known facts OR present supporting evidence for disagreeing with them. You cannot just disagree with the facts with no supporting evidence, that can only be taken as nonsense you've just invented in order to troll this forum, which obviously is completely unacceptable! This is the second warning in as many posts, either put up or shut up! IE. "Put up" the supporting evidence of "DACs being so bad at reproducing the signal". Incidentally, it makes absolutely no difference whatsoever mixing to levels higher than -4dB, as long as clipping is avoided.

G

 

[TheSonicTruth]

the mixer mixes, the mastering engineer makes it loud and does some fine eq tuning and roll offs. That a way the mix can stay at DR12-14 to be press correctly to an album. Otherwise, if you tried to cut the proof for a pressing with a digital delivery specified signal, you end up with low distorted audio from the cutting coil getting saturated with the signal.

Bottom line: the master engineer sets the loudness. And is allowed to work in the dumb-ass-zone to make the music work for the masses that have cheap audio equipment.

Just don't always blame the ME for that. Most MEs will educate their clients but in end try to accommodate their project desires.

 

[drtechno]

Just don't always blame the ME for that. Most MEs will educate their clients but in end try to accommodate their project desires.

We all know the money man has the last say....So yes its the Executive Producer/Label's Fault.

 

[drtechno]

If you're going to "answer" something in this particular forum, then you EITHER have to be agreeing with the known facts OR present supporting evidence for disagreeing with them. You cannot just disagree with the facts with no supporting evidence, that can only be taken as nonsense you've just invented in order to troll this forum, which obviously is completely unacceptable! This is the second warning in as many posts, either put up or shut up! IE. "Put up" the supporting evidence of "DACs being so bad at reproducing the signal". Incidentally, it makes absolutely no difference whatsoever mixing to levels higher than -4dB, as long as clipping is avoided.

G

First off I will answer the way I want to. I don't have anything to prove, and it doesn't matter, I usually watch the like of you and Pinnehertz bicker like to old ladies usually...

 

[TheSonicTruth]

"! If you're talking about the amount of compression
used in modern popular music (your agenda), then the
lower the dynamic range, the fewer bits are effectively
required, which is the exact opposite of "using all the bits"!"


Ok, and I'll try to word this as carefully as possible. I do not know if linear or logarithmic scale applies to the following, but here goes:

Reddbook/CDDA accommodates 16bits - 96dB dynamic range, or 6dB amplitude per bit. So, assuming 0dBfs = full scale = top of that 96 dB dynamic range:

If you have a sound, any sound - a 3kHz pure sine for instance - peaking at 0dBfs on that CD, wouldn't it be logical to assume that all 16bits/96dB DR are being continuously used?

So I am at a loss as to how such a signal, or a music track with both RMS and peak just below full scale is "not using" all the bits. Assume bit #1 = -96 to -90dBFs, and bit #16 = -6 to 0dBfs. For something to be not using all the bits, all the time, it would have to either be extreeeemely dynamic, or, be not all that dynamic, and for whatever reason peaking nowhere near full scale. IE: That same 3kHz sine wave tone peaking at only -12(14bits) or -24dBfs(12bits).

I knowI have probably worded this horribly, so please forgive me!

 

[drtechno]

Here is something that will make you think @TheSonicTruth , one of my instructors of the trade, (Alan Parsons) told me once that loudness is a perception. A tone generator, that outputs one frequency, that is a continuous waveform, will have a louder perception because the brain compares the signal to nothing (soft part), so it will appear louder. The only correlation I've witness is that the distance between loud and soft effects loudness, but too close will lead to a not so desirable loud signal. Too much dynamic range will make the signal set back in the mix, and make it appear to be weak.

btw the numbers don't mean so much, and meters are not used often as the ears are the primary thing that is used when mixing. CD is only marginally better than cassette tape, which in Alan Parson's opinion, was the worst format to come out.

 

[TheSonicTruth]

Here is something that will make you think @TheSonicTruth , one of my instructors of the trade, (Alan Parsons) told me once that loudness is a perception. A tone generator, that outputs one frequency, that is a continuous waveform, will have a louder perception because the brain compares the signal to nothing (soft part), so it will appear louder. The only correlation I've witness is that the distance between loud and soft effects loudness, but too close will lead to a not so desirable loud signal. Too much dynamic range will make the signal set back in the mix, and make it appear to be weak.

btw the numbers don't mean so much, and meters are not used often as the ears are the primary thing that is used when mixing. CD is only marginally better than cassette tape, which in Alan Parson's opinion, was the worst format to come out.

"CD is only marginally better than cassette tape,... "


Sonic snorts Pepsi Cola all over his tablet!


There's a 'wrong' statement for ya, Gregorio!
Ave Maria..

 

[bigshot]

CD is only marginally better than cassette tape, which in Alan Parson's opinion, was the worst format to come out.

 

[Glmoneydawg]

Oh dear...horrific how quickly you found that...did you pull it out of your AMC pacer?

 

[TheSonicTruth]

What I have to say about comparisons of Red Book CD to cassette and eight-track could get me banned, so I'm just zip it!

 

[pinnahertz]

Here is something that will make you think @TheSonicTruth , one of my instructors of the trade, (Alan Parsons) told me once that loudness is a perception. A tone generator, that outputs one frequency, that is a continuous waveform, will have a louder perception because the brain compares the signal to nothing (soft part), so it will appear louder.

In context, he's right. But the example is completely out of context!

The only correlation I've witness is that the distance between loud and soft effects loudness, but too close will lead to a not so desirable loud signal. Too much dynamic range will make the signal set back in the mix, and make it appear to be weak.

Those statements speak to lack of experience and familiarity. They're so vague there's no way to respond.

btw the numbers don't mean so much, and meters are not used often as the ears are the primary thing that is used when mixing.

Again, the statement screams "lack of experience". Meters not used often? Well, there's one on every output channel of every mixing desk, on every input channel of every recording device, every output channel of every recording device, every input/output of every effects device (hardware or software), but sure, they're not used much. How silly. And the "numbers don't mean much"? Yeah, that would only be true of someone without any experience. No, you don't mix with meters, but you sure don't ignore them either!

CD is only marginally better than cassette tape, which in Alan Parson's opinion, was the worst format to come out.

You must have not understood him. He's no idiot, but that statement is fully idiotic. There is no performance metric in which the CD doesn't completely crush the cassette. Perhaps the problem is the definition of "margin".

The cassette beat 8-track in every possibly way, and beat the 78rpm record too...and...well, no need to go on, it's hardly the worst, just hardly comparable to the CD either.

 

[bigshot]

Meters aren't used to come up with relative balances or to create the sound of a mix. They are used to calibrate and keep things in spec. Keeping stuff out of the red is the easy part of a mix. The creative aspects are a lot more time consuming and important. And the creative aspects don't show up on meters.

 

[gregorio]

[1] First off I will answer the way I want to. [1a] I don't have anything to prove, and it doesn't matter ...
[2] one of my instructors of the trade, (Alan Parsons) ...

1. Does that apply to everyone or just you? If it's everyone, then I want to answer by calling you a liar and a troll or at the very least, ignorant and deluded. However, we can't "answer the way we want to" here, there are TOS and forum etiquette/rules so I will not answer the way I want. Instead, I'll just say that your failure to provide any supporting evidence for your outrageous claims, plus your rather aggressive response in order to deflect and avoid presenting any, creates a strong indication/implication which others can easily deduce for themselves.
1a. Yes it does matter and yes you DO have something to prove, or at least, you have to provide some reliable supporting evidence for any claim which contradicts the actual/known facts. Failure to do so is effectively a tacit agreement that your assertion was FALSE/INCORRECT, based on a simple misunderstanding/misinterpretation of the actual facts at one extreme or at the other extreme, based on a deliberate lie, delusion and/or intention to troll this forum. In essence, this has already been explained to you, MORE THAN ONCE and yet you continue to do the same thing, strongly indicating "the other extreme"! if you don't want to follow the rules/etiquette of this forum then either leave and find another forum, where made-up nonsense is more acceptable, or stay and have your nonsense refuted, make yourself appear more and more ignorant/foolish and eventually probably get banned anyway! What's your choice?

2. This is the second time you've implied you have some formal qualification in music/audio engineering, is this really your claim?

Assume bit #1 = -96 to -90dBFs, and bit #16 = -6 to 0dBfs.

OK, this statement is where you seem to have gone wrong and stated the exact opposite to what you apparently intended to state. Your assumption is incorrect, in fact it's backwards. Bit #1, more formally called the Most Significant Bit (MSB), is effectively = -6 to 0dBFS and bit #16, which in a 16bit format is more formally called the Least Significant Bit (LSB), is effectively = -96dB to -90dB. A more heavily compressed/crushed signal therefore employs fewer bits, NOT more/all.

There's a 'wrong' statement for ya, Gregorio!

Yep, clearly very "wrong" indeed but others have already pointed it out.

G

 

[TheSonicTruth]

1. Does that apply to everyone or just you? If it's everyone, then I want to answer by calling you a liar and a troll or at the very least, ignorant and deluded. However, we can't "answer the way we want to" here, there are TOS and forum etiquette/rules so I will not answer the way I want. Instead, I'll just say that your failure to provide any supporting evidence for your outrageous claims, plus your rather aggressive response in order to deflect and avoid presenting any, creates a strong indication/implication which others can easily deduce for themselves.
1a. Yes it does matter and yes you DO have something to prove, or at least, you have to provide some reliable supporting evidence for any claim which contradicts the actual/known facts. Failure to do so is effectively a tacit agreement that your assertion was FALSE/INCORRECT, based on a simple misunderstanding/misinterpretation of the actual facts at one extreme or at the other extreme, based on a deliberate lie, delusion and/or intention to troll this forum. In essence, this has already been explained to you, MORE THAN ONCE and yet you continue to do the same thing, strongly indicating "the other extreme"! if you don't want to follow the rules/etiquette of this forum then either leave and find another forum, where made-up nonsense is more acceptable, or stay and have your nonsense refuted, make yourself appear more and more ignorant/foolish and eventually probably get banned anyway! What's your choice?

2. This is the second time you've implied you have some formal qualification in music/audio engineering, is this really your claim?



OK, this statement is where you seem to have gone wrong and stated the exact opposite to what you apparently intended to state. Your assumption is incorrect, in fact it's backwards. Bit #1, more formally called the Most Significant Bit (MSB), is effectively = -6 to 0dBFS and bit #16, which in a 16bit format is more formally called the Least Significant Bit (LSB), is effectively = -96dB to -90dB. A more heavily compressed/crushed signal therefore employs fewer bits, NOT more/all.



Yep, clearly very "wrong" indeed but others have already pointed it out.

G

Ok Greg so I had the bits numbered in reverse. I am diagnosed dyslexic, fyi.

But you suggestion that less dynamic music, brickwalled to - and possibly beyond - full scale, uses "less" of the available bits than either, IE, a tone peaking at only -24dBfs, or, something extremely dynamic(-20dBfs rms with peaks between -2 and 0dBfs) still does not make sense. It goes against my gut instinct.

So please explain how, in cogent, succinct, precise terms, how something loudness processed, both averaging and peaking at or near digital full scale, is "not using all the bits".