On Sun, 29 Jun 2003 01:07:46 -0500, Lord Hasenpfeffer
wrote:

I spent time today browsing websites and reading FAQs, etc. to find out
more information about the Psycho-Acoustic Model employed by MP3. It
is indeed similar to MiniDisc's ATRAC compression scheme in that it
removes not only "masked frequencies" but also frequencies which
are determined to be "too quiet to be heard" by common human ears.

This proves to me that MP3s encoded from "quieter WAVs" which have
been ripped from CDs mastered with low, average amplitudes will suffer
more at the hands of a lossy audio data compression algorithm than
will MP3s encoded from WAVs which have first been "appropriately
normalized" (i.e. Linux: $normalize -ba -10dBFS ) prior to being encoded.

There is a similar and very lengthy conversation going on at

http://www.hydrogenaudio.org/index.p...=1&t=10637&hl=

There is at least one lame developer (Gabriel) who hangs out at this
forum, so you might be able to catch his attention.

As others have already posted, newer versions of Lame have a variable
ATH, which depends on how close the signal is to full scale.

There are a couple of questions this raises:

1) Is it better to try to bring up soft music near 0dBFS and then
apply mp3gain afterwards? How much quality is audibly gained by doing
this?

2) Should loud music be brought down so that the resulting encoded
bitrate is smaller? How much quality is audibly lost by doing this?

ff123

"ff123" wrote in message
There are a couple of questions this raises:

1) Is it better to try to bring up soft music near 0dBFS and then
apply mp3gain afterwards? How much quality is audibly gained by doing
this?

Yes certainly. Very sensible idea and no audio damage, as opposed to RMS
normalising with/without limiting or compression in most instances.

2) Should loud music be brought down so that the resulting encoded
bitrate is smaller? How much quality is audibly lost by doing this?


Depends how urgent your filesize reduction is I guess.

geoff

ff123 wrote:

There is a similar and very lengthy conversation going on at

http://www.hydrogenaudio.org/index.p...=1&t=10637&hl=

Oh really? Hmmm... Is there like one really arrogant asshole there who
continuously proclaims to know it all and believes that he can teach the
world to sing (or at least drink a bottle of Coke)? If so, I've really
go a mind to slap his fat face! ;-)

Thanks. I'll check it out ASAP.

There is at least one lame developer (Gabriel) who hangs out at this
forum, so you might be able to catch his attention.

As others have already posted, newer versions of Lame have a variable
ATH, which depends on how close the signal is to full scale.

There are a couple of questions this raises:

1) Is it better to try to bring up soft music near 0dBFS and then
apply mp3gain afterwards? How much quality is audibly gained by doing
this?

2) Should loud music be brought down so that the resulting encoded
bitrate is smaller? How much quality is audibly lost by doing this?

Two more excellent questions. And if it is true that "it doesn't really
matter after all", I'm *still* interested in knowing the answers to such
questions just for the sake of knowing more truth about the process.
Whether such truths are evident in the practical or only in the
strictest of abstract philosophical terms, truth is truth and should
always be pursued.

Myke

--

-================================-
Windows...It's rebootylicious!!!
-================================-

On Thu, 03 Jul 2003 00:52:51 -0500, Lord Hasenpfeffer
wrote:

ff123 wrote:

There is a similar and very lengthy conversation going on at

http://www.hydrogenaudio.org/index.p...=1&t=10637&hl=

Oh really? Hmmm... Is there like one really arrogant asshole there who
continuously proclaims to know it all and believes that he can teach the
world to sing (or at least drink a bottle of Coke)? If so, I've really
go a mind to slap his fat face! ;-)

Thanks. I'll check it out ASAP.

There is at least one lame developer (Gabriel) who hangs out at this
forum, so you might be able to catch his attention.

As others have already posted, newer versions of Lame have a variable
ATH, which depends on how close the signal is to full scale.

There are a couple of questions this raises:

1) Is it better to try to bring up soft music near 0dBFS and then
apply mp3gain afterwards? How much quality is audibly gained by doing
this?

2) Should loud music be brought down so that the resulting encoded
bitrate is smaller? How much quality is audibly lost by doing this?

Two more excellent questions. And if it is true that "it doesn't really
matter after all", I'm *still* interested in knowing the answers to such
questions just for the sake of knowing more truth about the process.
Whether such truths are evident in the practical or only in the
strictest of abstract philosophical terms, truth is truth and should
always be pursued.

Myke

Apparently the ATH is lowered (made stricter) for soft music, meaning
that if you turn up the volume for soft passages, the encode should
still sound ok (in previous versions of lame people heard problems
when they did this).

There is also a time constant involved in the adaptive ATH, taking
into consideration the fact that the ear does not adjust immediately
to changes in volume.

I think the consensus of the hydrogenaudio.org thread was that it is
best for most people to use --scale for dynamically compressed music
(if they listen to it at normal volumes) to save bits and to still
remain transparent.

Read especially DickD's post on page 4:

http://www.hydrogenaudio.org/index.p...t=10637&st=75&

ff123

David Morgan (MAMS) wrote:

1) Is it better to try to bring up soft music near 0dBFS and then
apply mp3gain afterwards? How much quality is audibly gained by doing
this?

Yes certainly. Very sensible idea and no audio damage, as opposed to RMS
normalising with/without limiting or compression in most instances.

In all of the cases I've presented so far in defense of my use of
Normalize, I would surmise that simply employing its default behaviour
(i.e. normalizing to -12dBFS instead of -10dBFS) would be all that's
required to achieve said effect of "bringing soft music near 0dBFS".

I've described since the beginnings of this discussion in the other
newsgroup(s) my use of -10dBFS as a means I've discovered by which I can
achieve "slightly hot" levels which (at least to me) are similar to
allowing the meters to "lightly touch red" when recording to analog tape
which, AFAIK, is a perfectly sensible analog recording technique which I
have sorely missed being able to do since making the switch into the
digital realm.

I agree totally. But this places the manipulation in the hands of the end user -
which is admittedly, the best idea. However, as Myke is planning some serious
business as a webmaster, he does need to encode the best way possible so
that the largest number of end users get the best possible product without
having to deal with, "Oh this sucks... but I'll 'fix' it with my toys".
Casual listeners won't be up for taking the time. If it isn't good,
they'll move on.

Thanks, David!

Myke

--

-================================-
Windows...It's rebootylicious!!!
-================================-

"Lord Hasenpfeffer" wrote in message

In all of the cases I've presented so far in defense of my use of
Normalize, I would surmise that simply employing its default behaviour
(i.e. normalizing to -12dBFS instead of -10dBFS) would be all that's
required to achieve said effect of "bringing soft music near 0dBFS".



In the circumstances that is quite a reasonable thing to be doing.

A lot of bandwidth would have been saved if you had clearly explained that
(last week) and had been able to explain that "Normalise (application)
prevents clippiong when RMS normalising.

I've described since the beginnings of this discussion in the other
newsgroup(s) my use of -10dBFS as a means I've discovered by which I can
achieve "slightly hot" levels which (at least to me) are similar to
allowing the meters to "lightly touch red" when recording to analog tape
which, AFAIK, is a perfectly sensible analog recording technique which I
have sorely missed being able to do since making the switch into the
digital realm.

Most semi-pro-and up Windows (or Mac) audio software has plugins available
to achieve pretty much just that.

geoff
-=========================-
Linux. Attracts-those-inclined-to-quasi-religious-fanatiscm-and-does-little-
-to-help-their-credibility-lyicious
-=========================-

Geoff Wood wrote:

In all of the cases I've presented so far in defense of my use of
Normalize, I would surmise that simply employing its default behaviour
(i.e. normalizing to -12dBFS instead of -10dBFS) would be all that's
required to achieve said effect of "bringing soft music near 0dBFS".

In the circumstances that is quite a reasonable thing to be doing.

Good. In the meantime I'll start exploring my options with MP3Gain as
well as looking into upgrading Notlame (not that I have a problem with
the one I have already but... 2 years is quite a long time in this
arena). And, of course, more reading and testing and posting of results
as they become available.

A lot of bandwidth would have been saved if you had clearly explained that
(last week) and had been able to explain that "Normalise (application)
prevents clippiong when RMS normalising.

Well, the asinine stupidity continues over in the other thread even to
this day so I'm not so sure the *******s over there are really all that
interested in really being all that much help to anyone, regardless. It
seems their only goal is to compete and see who can be the biggest
*******. They're like little Chihuahuas who incessantly bark at all
passersby in an effort to try and make themselves appear bigger than
they and everyone else knows that they really are.

I've described since the beginnings of this discussion in the other
newsgroup(s) my use of -10dBFS as a means I've discovered by which I can
achieve "slightly hot" levels which (at least to me) are similar to
allowing the meters to "lightly touch red" when recording to analog tape
which, AFAIK, is a perfectly sensible analog recording technique which I
have sorely missed being able to do since making the switch into the
digital realm.

Most semi-pro-and up Windows (or Mac) audio software has plugins available
to achieve pretty much just that.

Well, when my professional and/or personal requirements extend beyond
that of simply needing to make "better" MP3s from my low-level CDs,
you'll be the first I'll call for advice on what to use. I'd much
prefer to slap something on my iMac so as to make that stupid thing a
bit more useful to me than it is right now, seeing as how it helps to
have one when you design pages which you want the whole world to be able
to see see equally well - rather than just the "ordinary Windows users"
who're out there in the world.

-=========================-
Linux. Attracts-those-inclined-to-quasi-religious-fanatiscm-and-does-little-
-to-help-their-credibility-lyicious
-=========================-

Try learning how to use it someday. You might find the cost-free,
crash-free, virus-free, anti-virus-free, constant-reboot-hassle-free,
Microsoft-bull****-free life to be kinda nice for a change.

Myke

--

-================================-
Windows...It's rebootylicious!!!
-================================-