Loudness Wars - some actual data/evidence!
 

People might be interested in seeing

http://jcgl.orpheusweb.co.uk/temp/vickers.gif

This shows a graph of the results of some research published by Vickers who
has reported on 'Loudness Wars' at a recent AES convention.[1]

The blobs on the graph represent the best selling CDs over a recent period
of time. The vertical scale scores them by how well they sold. The
horizontal scale assesses them by a measure of peak-to-mean ratio in dB we
can call 'DR'. [2]

In brief, if 'loudness and compression' boosted sales then you'd expect CDs
with a low DR to score higher and high DR to score lower. i.e. You'd see a
clear sloped relationship between the blobs with the ones at the lefh-hand
end being higher than at the right-hand end.

The red line is the trend line though the data. No sign of much slope.
Vickers reports the correlation is poor. i.e. no clear sign of any
relationship between compression and sales.

Vickers also reports on the rise in hearing damage and wonders if this is
associated with relentless loudness and peak-mean compression.

Given that the people who routinely compress music on CD *don't* seem to
have made public any assessable evidence I've yet been able to find to
support their behaviour, the above seems worth considering.

Points to ponder. :-)

Slainte,

Jim

[1] Alas, the print of this seems only available from the AES at present.
Can't find a copy of the pdf of his presentation openly on the web. But you
can find more from his page at

http://www.sfxmachine.com/docs/loudnesswar/

[2] Based on the 'meter' method given by pleasurizemusic.com

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

Loudness Wars - some actual data/evidence!
 

"Jim Lesurf" wrote in message
...
People might be interested in seeing

http://jcgl.orpheusweb.co.uk/temp/vickers.gif

This shows a graph of the results of some research published by Vickers
who
has reported on 'Loudness Wars' at a recent AES convention.[1]

The blobs on the graph represent the best selling CDs over a recent period
of time. The vertical scale scores them by how well they sold. The
horizontal scale assesses them by a measure of peak-to-mean ratio in dB we
can call 'DR'. [2]

In brief, if 'loudness and compression' boosted sales then you'd expect
CDs
with a low DR to score higher and high DR to score lower. i.e. You'd see a
clear sloped relationship between the blobs with the ones at the lefh-hand
end being higher than at the right-hand end.

The red line is the trend line though the data. No sign of much slope.
Vickers reports the correlation is poor. i.e. no clear sign of any
relationship between compression and sales.

Vickers also reports on the rise in hearing damage and wonders if this is
associated with relentless loudness and peak-mean compression.

Given that the people who routinely compress music on CD *don't* seem to
have made public any assessable evidence I've yet been able to find to
support their behaviour, the above seems worth considering.

Points to ponder. :-)

Slainte,

Jim

[1] Alas, the print of this seems only available from the AES at present.
Can't find a copy of the pdf of his presentation openly on the web. But
you
can find more from his page at

http://www.sfxmachine.com/docs/loudnesswar/

The slides used as part of the AES presentation are available on his
website.

Two things that emerged from it that interested me:

Firstly that the "loudness war" is the result of competition between
producers to have *the loudest* record. This kind of competition rapidly
results in extreme behaviour. A couple of other examples of that behaviour
that I can think of are the cold-war competition between the USA and the
USSR to have the most deadly nuclear arsenals, and the competition between
senior business executives and bankers to be paid more than their peers. In
all three cases the motivation is not to do something inherently correct or
justifiable, but by the need not to let the other guy get the better of you.

The other point was that the loudness war may have harmed music sales by
producing records that people tire of listening to, even if they are unable
to understand exactly what it is about the records that makes them feel that
way. I am reminded that I first heard "Switched-on Bach" (Bach played on a
Moog synthesiser) my initial reaction had been "wow!", but I very soon got
bored by it.

David.

Loudness Wars - some actual data/evidence!
 

The problem though is that some music does sound better when its compressed
and some does not. When you listen to the radio, you can clearly hear how
bad it makes some music sound as the folk in radio seem to be aiming for a
0db dynamic range for car listeners

Brian

--
Brian Gaff....Note, this account does not accept Bcc: email.
graphics are great, but the blind can't hear them
Email:
__________________________________________________ __________________________________________________ __________


"Jim Lesurf" wrote in message
...
People might be interested in seeing

http://jcgl.orpheusweb.co.uk/temp/vickers.gif

This shows a graph of the results of some research published by Vickers
who
has reported on 'Loudness Wars' at a recent AES convention.[1]

The blobs on the graph represent the best selling CDs over a recent period
of time. The vertical scale scores them by how well they sold. The
horizontal scale assesses them by a measure of peak-to-mean ratio in dB we
can call 'DR'. [2]

In brief, if 'loudness and compression' boosted sales then you'd expect
CDs
with a low DR to score higher and high DR to score lower. i.e. You'd see a
clear sloped relationship between the blobs with the ones at the lefh-hand
end being higher than at the right-hand end.

The red line is the trend line though the data. No sign of much slope.
Vickers reports the correlation is poor. i.e. no clear sign of any
relationship between compression and sales.

Vickers also reports on the rise in hearing damage and wonders if this is
associated with relentless loudness and peak-mean compression.

Given that the people who routinely compress music on CD *don't* seem to
have made public any assessable evidence I've yet been able to find to
support their behaviour, the above seems worth considering.

Points to ponder. :-)

Slainte,

Jim

[1] Alas, the print of this seems only available from the AES at present.
Can't find a copy of the pdf of his presentation openly on the web. But
you
can find more from his page at

http://www.sfxmachine.com/docs/loudnesswar/

[2] Based on the 'meter' method given by pleasurizemusic.com

--
Please use the address on the audiomisc page if you wish to email me.
Electronics
http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

Loudness Wars - some actual data/evidence!
 

In article , Brian Gaff
wrote:
The problem though is that some music does sound better when its
compressed and some does not.

Brian: As an aside: Occurs to me to ask. Are you able to read OK the PDF of
the slides, etc, that Vickers has on his website? There are also some
videos. if you can't access his slides you may not know the details of what
he found.

That said...

Probably so, for "some" (whatever that turns out to mean if it isn't simply
self-defined). That isn't the point of the results I put up, though.

The results of the work by Vickers (and others he mentions) indicate that
the music biz 'gurus' who apply level compression thereby produce CDs whose
sales show *no sign* that sales *were* increased by the level compression.

Easy to say that one of the horses will win the race. Not always so easy to
say *which* one in advance. :-)

The real question here is regarding the judgement of the people who apply
so much compression, and what basis in evidence they have to support the
idea that they *can* judge this successfully any better than tossing a coin
or using the i ching. Or simply not bothering. :-)

The results indicate that the gurus who routinely apply level compression
seem to do so *without* the judgement needed to reliably pick when that
*will* improve sales. I assume that is their main measure of 'better' so
far as the rock/pop biz is concerned. On that basis, the evidence is that
they fail.

The outcome seems to be that they take for granted that level compression
is often 'required' by default for rock and pop, so they apply it without
any judgement to such categories of music.

The rest of the paper then deals with possible negative results - e.g.
hearing damage in the young, and increased listening fatigue.

Personally I suspect that one of the reasons mp3 at low rates has become
popular is that the CDs of pop music have been level compressed and this
then makes the difference between CD and mp3 less of a concern to people.
So it encourages people to go and download 'free' mp3 rather than buy a CD.

Excessively compressed radio would contribute to that as well, I guess.

If there is any evidence that any of the gurus *do* reliably boost sales by
level compression I've not seen it. All I've heard is others insisting that
the perps possess 'hidden knowledge' us mere mortals must take on trust as
they are so clever they could not possibly be mistaken. The results from
Vickers indicate no sign such faith is justified. Quite the opposite.

So yes, I agree that some level compression is sometimes a good idea. But
that does seem to have morphed into a Loudness War. Vickers has shown the
'game theory' mechanism that leads to this if the participants haven't a
clue about the way this can happen and no-one can challenge their belief
system presented as 'fact'.

Personally, I have the academic scientist's and engineer's suspicion of
'hidden evidence' we are not allowed to see.

I'll look forwards to see real assessible evidence from those who apply so
much compression. But I'll not hold my breath while I wait. :-)

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html