[[email protected]]

On Monday, October 12, 2015 at 9:31:32 AM UTC-4, Scott Dorsey wrote:
wrote:

Actually, I was criticizing the need to have
all tracks on a CD peak within 1/10 of full scale,
for WHAT EVER reason.

Why? There's nothing wrong with that. If the D/A converter is properly
designed, it's going to reproduce what went into it exactly.

Peak levels have nothing to do with actual loudness. I can have a very
quiet CD with one triangle hit that goes up to 0dBFS without even
sounding very loud because it's so brief.

So you _want_ your peak levels to be at 0dBFS. The question is where
you want your _average_ levels to be, because that's where loudness comes
from.

Now, there IS a reason to leave plenty of headroom when you're going to
be doing processing afterward, or on an original recording when you can
never be sure something loud and unexpected won't happen. But that has
absolutely no application to a final release CD.

I know we have all explained this to you many many times and you don't get
it, but I am going to try again.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."


Thanks for that explanation Scott.


My issue with peak normalization(which is what
you describe above is called), is that a song
with a lower peak-to-average ratio, peak normalized
alongside a song with a higher peak-to-average,
WILL SOUND LOUDER than the one with the higher
peak to average.


OTOH: If I use my ears to loudness-normalize 3-4
songs of different genres or production eras
(stuff from the 1970s, 1990s, and last month),
the final result is that they will all sound
equally loud to my ears, but some of them
my not peak at 0dB full scale - when I look
at the meters during playback. And in my
mind there's NOTHING WRONG with that; as long
as they all sound about as loud as each other
and sound GOOD.