I'm hoping this might make a change from a mega-thread about "Audio
Precision" that is actually wandering all over the shop!... :-)

I've been experimenting with 'sox' to deemphasise audio CD data, and the
flexibility of sox set me wondering about using it to eq old EMI recordings
that sometimes have an edgey quality to the strings, and lack low bass
because of their old fear of bass causing 'groove jumping'. Not
experimented yet, but wondered if anyone else has had a play with this.

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

On Fri, 10 Feb 2012 17:21:06 +0000 (GMT), Jim Lesurf
wrote:

I'm hoping this might make a change from a mega-thread about "Audio
Precision" that is actually wandering all over the shop!... :-)

I've been experimenting with 'sox' to deemphasise audio CD data, and the
flexibility of sox set me wondering about using it to eq old EMI recordings
that sometimes have an edgey quality to the strings, and lack low bass
because of their old fear of bass causing 'groove jumping'. Not
experimented yet, but wondered if anyone else has had a play with this.

Slainte,

Jim

Don't know what sox is, but yes, I've done this kind of thing before
with Audition. I found a useful way to get it right, which was to take
a spectrum of the same piece, but a recording I liked and compare it
with the spectrum of the deficient one. Audition lets you draw
filters, and it was an easy task to construct one that corrected the
deficiencies. I did find that it was a good idea not to try and
correct too much, though.

d

In article ,
(Don Pearce) wrote:

On Fri, 10 Feb 2012 17:21:06 +0000 (GMT), Jim Lesurf
wrote:

I'm hoping this might make a change from a mega-thread about "Audio
Precision" that is actually wandering all over the shop!... :-)

I've been experimenting with 'sox' to deemphasise audio CD data, and the
flexibility of sox set me wondering about using it to eq old EMI recordings
that sometimes have an edgey quality to the strings, and lack low bass
because of their old fear of bass causing 'groove jumping'. Not
experimented yet, but wondered if anyone else has had a play with this.

Don't know what sox is, but yes, I've done this kind of thing before
with Audition. I found a useful way to get it right, which was to take
a spectrum of the same piece, but a recording I liked and compare it
with the spectrum of the deficient one. Audition lets you draw
filters, and it was an easy task to construct one that corrected the
deficiencies. I did find that it was a good idea not to try and
correct too much, though.

Sounds like descriptions I've read of Har-Bal. Does anyone have
experience with that program?

Stephen

In article , Don Pearce
wrote:
On Fri, 10 Feb 2012 17:21:06 +0000 (GMT), Jim Lesurf
wrote:


I've been experimenting with 'sox' to deemphasise audio CD data, and
the flexibility of sox set me wondering about using it to eq old EMI
recordings that sometimes have an edgey quality to the strings, and
lack low bass because of their old fear of bass causing 'groove
jumping'. Not experimented yet, but wondered if anyone else has had a
play with this.

Jim

Don't know what sox is,

http://sox.sourceforge.net/Docs/Features

The above gives a summary of sorts. But basically it is a command for doing
all kinds of things/with to audio files. I use it mainly for simple
conversions like Wave = Flac or changing sample rates or depths. But more
recently I've used it to deemphasise 'eq'd audio CD tracks. e.g. by using

sox infilename.wav -b 24 outfilename.flac deemph

It will take an input file (wave) ripped as 44.1k/16 and generate a flac
output that has had the Red Book deemphasis applied. The '-b 24' gets it to
output the flac as 24 bits per sample to avoid any losses that might arise
when dithering down the output back to 16 bit.

Since then I've also experimented with doing things like

sox infile outfile equalizer 3k 2o -6.0 bass +6.0

[N.B. that is '2 followed by a lower case 'Oh', not a zero as it means
'octave'.]

which would apply a 2nd order band filter centered at 3kHz that is 2
octaves wide with a dip of -6dB at 3k and also apply a 'baxadall' bass lift
rising to 6dB at LF. You can actually chain a series of such effects if
required. Or have complex effects specified by a script file. You can
define the filters by time-domain or FFT if required. Or biquad
coefficients, etc, so it is (bewilderingly!) flexible.

Being a command util it is then easy for me to write a simple drag-and-drop
wrapper or use it for batches of files if required.

The real question, though, is what - if any - such alterations may be an
'improvement' when it comes to something like the more 'scratchy' old EMI
recordings.

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

On Sat, 11 Feb 2012 10:07:53 +0000 (GMT), Jim Lesurf
wrote:

In article , Don Pearce
wrote:
On Fri, 10 Feb 2012 17:21:06 +0000 (GMT), Jim Lesurf
wrote:


I've been experimenting with 'sox' to deemphasise audio CD data, and
the flexibility of sox set me wondering about using it to eq old EMI
recordings that sometimes have an edgey quality to the strings, and
lack low bass because of their old fear of bass causing 'groove
jumping'. Not experimented yet, but wondered if anyone else has had a
play with this.

Jim

Don't know what sox is,

http://sox.sourceforge.net/Docs/Features

The above gives a summary of sorts. But basically it is a command for doing
all kinds of things/with to audio files. I use it mainly for simple
conversions like Wave = Flac or changing sample rates or depths. But more
recently I've used it to deemphasise 'eq'd audio CD tracks. e.g. by using

sox infilename.wav -b 24 outfilename.flac deemph

It will take an input file (wave) ripped as 44.1k/16 and generate a flac
output that has had the Red Book deemphasis applied. The '-b 24' gets it to
output the flac as 24 bits per sample to avoid any losses that might arise
when dithering down the output back to 16 bit.

Since then I've also experimented with doing things like

sox infile outfile equalizer 3k 2o -6.0 bass +6.0

[N.B. that is '2 followed by a lower case 'Oh', not a zero as it means
'octave'.]

which would apply a 2nd order band filter centered at 3kHz that is 2
octaves wide with a dip of -6dB at 3k and also apply a 'baxadall' bass lift
rising to 6dB at LF. You can actually chain a series of such effects if
required. Or have complex effects specified by a script file. You can
define the filters by time-domain or FFT if required. Or biquad
coefficients, etc, so it is (bewilderingly!) flexible.

Being a command util it is then easy for me to write a simple drag-and-drop
wrapper or use it for batches of files if required.

The real question, though, is what - if any - such alterations may be an
'improvement' when it comes to something like the more 'scratchy' old EMI
recordings.

Slainte,

Jim

Well, I suppose you try a few by ear to see how it goes. I don't envy
you a command line interface for this job though? The graphic
environment of Audition makes the whole thing a good deal more
intuitive.

d

In article , Don Pearce
wrote:
On Sat, 11 Feb 2012 10:07:53 +0000 (GMT), Jim Lesurf
wrote:
[snip]

The real question, though, is what - if any - such alterations may be
an 'improvement' when it comes to something like the more 'scratchy'
old EMI recordings.


Well, I suppose you try a few by ear to see how it goes. I don't envy
you a command line interface for this job though? The graphic
environment of Audition makes the whole thing a good deal more intuitive.

But it brings me back to my main question, which I've left unsnipped above.
I am wondering if others have already experimented and formed a view wrt
what kind of alteration may be systematically beneficial for various
classes of material?

In part this is prompted by the BBC CDs I found need Red Book deemphasis.
But also by having bought the 'Decca' box set of 50 classical recordings -
and now also the new 'Living Presence' set - and comparing them both with
old EMI classical recordings. I wonder to what extent each label has its
own 'sonic signature' simply in response terms.

I recall that EMI back in the 60s had a habit, for example, of rolling away
low bass with the reason being "to avoid groove jumping" or increase how
much could be fitted onto an LP side. And 'multi mic' recordings do
sometimes give me the feeling of 'too much presence band' compared with the
'sparse' arrangements favoured by Decca or Mercury. Or is this more a
matter of different mic characteristics?

Whatever the reasons, I was set wondering if anyone else had already
experimented along these lines and come to any conclusions.

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

On Sat, 11 Feb 2012 12:45:33 +0000 (GMT), Jim Lesurf
wrote:

In article , Don Pearce
wrote:
On Sat, 11 Feb 2012 10:07:53 +0000 (GMT), Jim Lesurf
wrote:
[snip]

The real question, though, is what - if any - such alterations may be
an 'improvement' when it comes to something like the more 'scratchy'
old EMI recordings.


Well, I suppose you try a few by ear to see how it goes. I don't envy
you a command line interface for this job though? The graphic
environment of Audition makes the whole thing a good deal more intuitive.

But it brings me back to my main question, which I've left unsnipped above.
I am wondering if others have already experimented and formed a view wrt
what kind of alteration may be systematically beneficial for various
classes of material?


I thought I had covered that. Yes I've done exactly that, but I would
say the results were not uniformly good. Using my method I found that
it was a bad idea to go beyond about half of th shortfall for any
frequency band. Beyond that, an unpleasant degree of unnaturalness
overwhelmed things.

In part this is prompted by the BBC CDs I found need Red Book deemphasis.
But also by having bought the 'Decca' box set of 50 classical recordings -
and now also the new 'Living Presence' set - and comparing them both with
old EMI classical recordings. I wonder to what extent each label has its
own 'sonic signature' simply in response terms.

I recall that EMI back in the 60s had a habit, for example, of rolling away
low bass with the reason being "to avoid groove jumping" or increase how
much could be fitted onto an LP side. And 'multi mic' recordings do
sometimes give me the feeling of 'too much presence band' compared with the
'sparse' arrangements favoured by Decca or Mercury. Or is this more a
matter of different mic characteristics?


The problem here is that if you try and recover the bass on vinyl, you
very quickly discover the joys of groove warp and rumble. The
nastiness of multi-mic recording, of course, is just something you
have to endure if that is the performance you want to hear. I'm
willing to bet that they recorded a safety track from an X-Y pair or a
Jecklin disc that is languishing on a reel of half inch somewhere. A
re-release of the old material based on those tapes would be very
welcome.

Whatever the reasons, I was set wondering if anyone else had already
experimented along these lines and come to any conclusions.


My conclusions? If the problem isn't severe, go for it. But you won't
get any joy out of trying to fix something really nasty.

d

On 11/02/2012 14:08, Don Pearce wrote:

My conclusions? If the problem isn't severe, go for it. But you won't
get any joy out of trying to fix something really nasty.

My conclusion is that you should get a Linsley Hood amp from the 1970s
and play with the switched frequency bass and treble controls,
and the switched frequency variable slope treble filter until you get bored,
then put everything back to flat and just enjoy the music.

--
Eiron.

In article , Don Pearce
wrote:
On Sat, 11 Feb 2012 12:45:33 +0000 (GMT), Jim Lesurf
wrote:

In article , Don Pearce
wrote:
On Sat, 11 Feb 2012 10:07:53 +0000 (GMT), Jim Lesurf
wrote:
[snip]

The real question, though, is what - if any - such alterations may
be an 'improvement' when it comes to something like the more
'scratchy' old EMI recordings.


Well, I suppose you try a few by ear to see how it goes. I don't envy
you a command line interface for this job though? The graphic
environment of Audition makes the whole thing a good deal more
intuitive.

But it brings me back to my main question, which I've left unsnipped
above. I am wondering if others have already experimented and formed a
view wrt what kind of alteration may be systematically beneficial for
various classes of material?


I thought I had covered that. Yes I've done exactly that, but I would
say the results were not uniformly good.

That addresses my specific point of interest more directly. I've had the
impression that there does tend to be a 'label balance' which often
distinguishes pre mid 1970s EMI from Decca, for example. Here I'm focussing
on the spectral response differences which seem to me to often make the EMI
more forward in the few kHz region but lacking in low bass compared with
Decca.

So your view is that any such general tendency varies so much in detail
from example to example that there isn't a single alteration that might be
useful as an option for an 'EMI sound'? That is one of the points I've been
pondering. One reason for this I outline below.

Using my method I found that it
was a bad idea to go beyond about half of th shortfall for any frequency
band. Beyond that, an unpleasant degree of unnaturalness overwhelmed
things.

[snip]

The problem here is that if you try and recover the bass on vinyl, you
very quickly discover the joys of groove warp and rumble.

I'm not interested here in altering what I hear on LP. Indeed, one of the
things that prompted my wondering about this is that I keep having the
impression that - for EMI in particular - there is a difference in spectral
balance between their LP and CD issues of some (if not many) works.

Indeed, I've been wondering if they used to 'tweak' what was cut to LP in a
way that does this, but haven't (always) been taking that into account for
later CD re-issues. I am thinking of materal from many years ago.

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

In article , Eiron
wrote:
On 11/02/2012 14:08, Don Pearce wrote:

My conclusions? If the problem isn't severe, go for it. But you won't
get any joy out of trying to fix something really nasty.

The problem for me is EMI recordings of excellent performances of classical
music by favourite artists that have a sound quality I don't like. This can
certainly be endured and 'listened though'. But given the flexibility of
what I could try, I have become curious. So wondered what others may have
tried already and decided.

My conclusion is that you should get a Linsley Hood amp from the 1970s
and play with the switched frequency bass and treble controls, and the
switched frequency variable slope treble filter until you get bored,
then put everything back to flat and just enjoy the music.

I already use a Quad 34 whose 'tone controls' I modified some years ago. I
do alter the settings depending on the source material. Although most of
the time I keep to the same settings that give a better response from
fairly flat material and measure closest to flat averaged over a reasonable
frequency bandwidth. The problem is that I don't think normal 'tone
controls' really address what I was curious about.

When I get a chance I'll experiment with this and see if anything emerges.
But Don's comments do chime with my feeling that this may vary too much.
Although I have the feeling that a fairly simple common alteration might be
useful for some old EMI reissues on CD. It may well be that I'm just
hearing the effects of messy multi-micing, mic resonances, etc, and so
impractical to deal with. But I am wondering, for example, if a dip in the
few kHz region may make some of the poorer examples sound better.

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