On Fri, 27 Jun 2003 22:07:08 -0500, Lord Hasenpfeffer
wrote:

Bob Cain wrote:

What is your measure of that whooping?

1 (one).

Waveform pictures tell nothing.

I was accused of clipping, limiting, compressing, and reducing the
dynamic range of my original WAVs by "normalizing" individual files or
"batch normalizing" whole groups of related files with the Linux
application known as "normalize".

My waveform pictures *do indeed* reveal that not one of these
accusations are true. Every single one of them is *false*.

You appear unable to understand the basics. You *cannot* increase the
average level of a piece of music (while avoiding clipping) without
affecting the dynamics of the piece. This is not simply a matter of
comparing the highest peak with the lowest trough, but of observing
the *transfer curve* which determines the relationship of loud and
quiet passages. In order to 'normalise' to a different average level,
you must be changing something, yes? If it's not a simple matter of
increasing the level of the entire CD (aka turning up the volume
control), then you are clearly messing with the internal dynamics, and
introducing a form of 'soft limiting'. This should be obvious.

If *you* like the sound of that, then fine, but please don't jump onto
this forum crowing that you've 'whopped the ass' of MFSL, because that
just makes you look like a brain-dead ten-year-old with a new toy.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

"Stewart Pinkerton" wrote in message


You appear unable to understand the basics. You *cannot* increase the
average level of a piece of music (while avoiding clipping) without
affecting the dynamics of the piece.

Stewie - you miss the pont. He is *improving * the dynamics, because the
original mixing and mastering dudes were totally incompetent.

geoff

Stewart Pinkerton wrote:

You *cannot* increase the average level of a piece of music
(while avoiding clipping) without affecting the dynamics of the piece.

If the levels of the original peaks are as far from Full Scale as the
peaks of MFSL's DSoTM CD are I can. As can be clearly seen in my
screenshot, I easily increased the average level of the WAV encompassing
that entire album by +4.5dB and introduced no limiting, compression, or
clipping to the waveform of the work.

This is not simply a matter of comparing the highest peak with the
lowest trough,

I didn't compare the highest peak to the lowest trough. If anything, I
compared the levels of the highest and lowest peaks; both of which are
located above 0dB.

but of observing the *transfer curve* which determines the relationship
of loud and quiet passages.

I *think* I've just used different words to describe the same thing.

In order to 'normalise' to a different average level, you must be
changing something, yes?

Correct. It's called "the amplitude" (and nothing more). This is
important when working with lossy compression schemes which employ, in
part, methods of psychoacoustic filtering designed to remove frequencies
from sound sources which are "too quiet to be heard" by the human ear.

If it's not a simple matter of increasing the level of the entire CD
(aka turning up the volume control),

My screenshot reveals that that is in fact all that I have done with the
original WAV.

then you are clearly messing with the internal dynamics, and
introducing a form of 'soft limiting'. This should be obvious.

Where in the screenshots that I have provided do you find any evidence
that 'soft limiting' has indeed been applied to the original WAV? I
don't believe it has. If it had been, your point would be valid.

If *you* like the sound of that, then fine,

No, I like the sound of my MP3 after I've amplified the original WAV as
opposed to the sound of an MP3 encoded directly from the original WAV.

but please don't jump onto this forum crowing that you've 'whopped
the ass' of MFSL, because that just makes you look like a brain-dead
ten-year-old with a new toy.

Whoops! Looks like I've been outed!

Myke

--

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

Lord Hasenpfeffer:
Stewart Pinkerton wrote:
In order to 'normalise' to a different average level, you must be
changing something, yes?
Correct. It's called "the amplitude" (and nothing more). This is
important when working with lossy compression schemes which employ, in
part, methods of psychoacoustic filtering designed to remove frequencies
from sound sources which are "too quiet to be heard" by the human ear.

Got an idea to test your psychoacoustic threshold theory..

1. Encode an MP3 from the normalized wave.
2. Encode an MP3 from the original source.
3. Use MP3Gain to do an MP3-native 'scale factor edit' to increase the volume of the quiet MP3
to match that of the normalized MP3 as closely as possible (it'll help if you know exactly how
much you raised the volume of the normalized wave).
4. Compare.

-StArSeEd
--
dchub://tsphub.dyndns.org:1979
IRC EFnet #smashing_pumpkins
Email: am
ICQ UIN: 1711589

StArSeEd wrote:

Got an idea to test your psychoacoustic threshold theory..

Whoops, it's not a theory just yet; merely a hypothesis
(and a source for curiosity).

1. Encode an MP3 from the normalized wave.
2. Encode an MP3 from the original source.

Hey, I know! I'll use MFSL's "Dark Side Of The Moon" for this!

3. Use MP3Gain to do an MP3-native 'scale factor edit' to increase the volume of the quiet MP3
to match that of the normalized MP3 as closely as possible (it'll help if you know exactly how
much you raised the volume of the normalized wave).
4. Compare.

That's a very, very good idea. As soon as I get this "MP3Gain" and
learn how to use it, I'll let you know what happens.

Then again, before I just go about trying to "learn how to use it", got
any suggestions on which college I should attend so that I can come here
afterwards and talk about my findings? :-)

Myke

--

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

Stewart Pinkerton wrote:


--

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