DIYHiFi.org

For the sake of audio
It is currently Fri Feb 23, 2018 2:28 pm

All times are UTC




Post new topic Reply to topic  [ 9 posts ] 
Author Message
PostPosted: Mon Jun 10, 2013 6:22 am 
Offline
Muriel
User avatar

Joined: Thu Jul 26, 2007 5:45 pm
Posts: 1179
Folks,

For over two decades my audio systems have been without a proper physiologically corrected Volume control.

Before '89 in old commie kraut land I used a volume control rotary switch (> 30 poles) with a physiological contour build in that pretty well matched the old Eckmiller W88 Fader (it's network was derived from that). One thing that this Volume control allowed was to mix/master music late at night at much reduced volume and still have the results translate well at high volume.

Image

(from http://german-modules.de-blog.jp/blog/2009/03/eckmiller_w85_s.html)

It was part of a custom build "preamp" with gain controls for all Inputs, metering for average loudness (VU meters) and additional gain trim during playback AND multiple also gain controlled outputs that allowed the signal to be send to several different Amplifier/Speaker combos while compensating the gain/spl differences...

Well all that gear was lost and while I always wanted to sort out something, it never quite happened.

As I often listen at realistic levels, I find that High End Audio's "fix" for the lack of a correct physiological volume control, namely to permanently boost the Bass in speakers by 6dB and call that "baffle step correction" (it should be obvious why there is no need for baffle step correction when listening in an enclosed space - though it is needed for listening in anechoic chambers of free space suspended several dozen meters above ground) causes the sound to be excessively bass heavy at realistic volumes.

If you do not believe me about the "baffle step correction" being just Loudness Correction, look at what the baffle step correction does to the power response of the speaker and then compare this against the required loudness correction posted further down - a Speaker with 6dB "Baffle step correction" is actually the same as the Loudness correction for around 70dB average playback SPL instead of 85dB

But, if I then listen to the same system late at night - quietly the sound is predictably thin and clearly has a imbalance with a severe lack of bass. In fact, this still happens even with current fashion of non-flat, bass-boosted loudspeakers if you back off the Volume too much. For ages I have shied away from the obvious solutions, for reasons of "signal purity" and lack of time and lack of a good template to follow in building one (BTW, if anyone has the schematic for the Eckmiller W88, I'd greatly appreciate it).

As I just decided to pretty permanently (well, for the next few month, until a more interesting project will go to fruition) add a box to my system that included a Pot, which in my case was redundant - I had a major BFO (BFO - Blinding Flash of the Obvious). So I decided to cobble in a first order approximation of a proper loudness...

First, proper loudness is not that "bathtub" bass & treble boost applied to a tap on the volume control, invented in America and foisted upon the world by the Japanese consumer giants... The shape of a proper loudness compensates not the so-called "Fletcher-Munson Curve" (incidentally what is usually shown as FM is in Robinson-Dadson), but the DIFFERENCE between the curves, usually no treble boost should be applied! As both Fletcher-Munson and Robinson-Dadson are now outdated and replaced by ISO 226:2003, so we should probably use this as basis. Now what does the required loudness contour look like, taking 90dB SPL as "reference"?

The 90dB reference is German Studio Standard and assumes 14dB headroom, it is in effect the same as THX Standard 85dB SPL @ -20dBFS signal... within 1dB

Image

If we do not require too large a range this can be easily implemented by adding a series RC in the ground connection of the Pot.

I eyeballed my one for a range of 0 - -30dB (or 90 to 60dB SPL).

I'll not make the values I used public, working out the needed RC from the curves shown is a good exercise to stop the brain from aging, or just fire up your favourite Sim...

With such a "loudness" control the Amplifiers Volume control becomes the gain control - so with the loudness set to all the way up/flat set 85dB for -20dBFS at the listening position and then use the "loudness" to set playback level. Voila, no more thin sound at late night listening. Any differences in recording level are controlled using the (remote operated) original volume control BTW...

For those not wanting a full Loudness implementation it may be best to use a 3-Way toggle to implement "bypass", -15dB and -30dB Loudness compensated attenuation, for (respectively) Lynyrd Skynyrd at Mach 2 on Spinal Tap (bypass), Eve Cassidy at "audiophile approved" levels and anything you fancy late at night without making too much of a nuisance out of yourself...

Now I think I need to work out something with a wider range (say > 40dB), using a 24-Position switch and a better curve fit than my "quick and ugly" hack on existing gear... This is where that Eckmiller W88 schematic comes in - anyone please?

Oh and do try to remove the bass bost from your speakers (if they have it) and make it adjustable...

Ciao T

PS, while active, overly complex and not blessed with enough steps for my taste etc., the circuit of the Neumann W482ST may be found on-line and may serve as information on how things can be done. The original Eckmiller that I cannot find was easier/simpler to implement and fully passive.

_________________
"For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." Richard Feynman


Top
 Profile  
Reply with quote  
PostPosted: Mon Jun 10, 2013 4:58 pm 
Offline
Cow

Joined: Mon Feb 09, 2009 2:36 am
Posts: 149
Not this one (W86):

Attachment:
ECKMILLER W86A.JPG


From "Vintage German modules:Neumann, Siemens, TAB, IRT, Lawo, ANT, Eckmiller, Adis, Monitora and Telefunken"

http://audio.kubarth.com/rundfunk/index.cgi#w


You do not have the required permissions to view the files attached to this post.


Top
 Profile  
Reply with quote  
PostPosted: Mon Jun 10, 2013 9:48 pm 
Offline
Cow

Joined: Mon Feb 09, 2009 2:36 am
Posts: 149
Doh, of course you know the 'vintage German module' site since you referenced it here:

http://www.diyaudio.com/forums/analog-l ... ost2942825

Here's an update that'll get you to ian millar's referenced circuit (McIntosh type 'loudness'):

http://ianamillar.com/blog/poqx/

Here's a pretty good normalized Fletcher Munson

Attachment:
Normalised Fletcher Munson.jpg


You do not have the required permissions to view the files attached to this post.


Top
 Profile  
Reply with quote  
PostPosted: Mon Jun 10, 2013 10:00 pm 
Offline
Cow

Joined: Mon Feb 09, 2009 2:36 am
Posts: 149
And here's the cute little circuit:

Attachment:
McIntosh type loudness circuit.JPG


You do not have the required permissions to view the files attached to this post.


Top
 Profile  
Reply with quote  
PostPosted: Mon Jun 10, 2013 10:03 pm 
Offline
Cow

Joined: Mon Feb 09, 2009 2:36 am
Posts: 149
And here's the curve:

Attachment:
LoudnessCurves.gif


You do not have the required permissions to view the files attached to this post.


Top
 Profile  
Reply with quote  
PostPosted: Tue Jun 11, 2013 9:18 pm 
Offline
Muriel
User avatar

Joined: Thu Jul 26, 2007 5:45 pm
Posts: 1179
Hi,

JDC wrote:
Here's an update that'll get you to ian millar's referenced circuit (McIntosh type 'loudness'):

http://ianamillar.com/blog/poqx/

Here's a pretty good normalized Fletcher Munson

Attachment:
Normalised Fletcher Munson.jpg


Fletcher Munson is 80 Years old. And it is not the most accurate or realistic

Ian Millars circuit is better off with deleting R1/C1 (the HF part) and re-adjusting the time constants for R2/C2to ISO 226:2003 (shown earlier).

I would still like to look at what makes the W88 work (again), it seems by far closer to ISO226:2003 than FM or RD. I guess I am just very lazy.

Ciao T

_________________
"For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." Richard Feynman


Top
 Profile  
Reply with quote  
PostPosted: Wed Jan 01, 2014 5:05 pm 
Offline
Cow

Joined: Tue Jan 15, 2008 12:09 am
Posts: 280
@JDC - Variations:
http://diyaudioprojects.com/mirror/members.aol.com/sbench/freqres3.html
Image
Image


Top
 Profile  
Reply with quote  
PostPosted: Thu Jan 02, 2014 11:18 am 
Offline
Board jester
User avatar

Joined: Mon Dec 13, 2004 3:08 pm
Posts: 2042
It depends on your loudspeakers and acoustics. Most loudness circuits work too strong.
I defeated it in my car radio.

_________________
Elso


Top
 Profile  
Reply with quote  
PostPosted: Thu Jan 02, 2014 5:38 pm 
Offline
Muriel
User avatar

Joined: Thu Jul 26, 2007 5:45 pm
Posts: 1179
Elso,

Elso Kwak wrote:
It depends on your loudspeakers and acoustics. Most loudness circuits work too strong.
I defeated it in my car radio.


Traditional loudness circuits are way wrong on all fronts.

This was the point of my original post. I have since found W85 details and it uses a SINGLE CAP on a tap of the resistor chain.

Easily used for stepped attenuators but not so easy for pot's, as tapped units are getting rare.

I would suggest that 90dB AVERAGE replay SPL is quite loud, if we set this at 0dBVU (-14dB "Full Scale") it means peaks will hit 104dB (lets allow +/-3dB tolerance at least).

Then, if we attenuate the signal by 20dB (around midpoint on many Preamp's) to 70dB average SPL (a more common level of listening) we need +10dB around 20Hz and +3dB around 250Hz. Maybe now the bass heavy tuning of some american Speakers (which sound extremely bass heavy if played loud) makes some sense?

If we drop the average SPL to 50dB (this is "late night" listening) we need around +20dB at 20Hz and +3dB at 500Hz.

If we make our "Volume" control in effect a physiological compensated "Loudness" regulator (that is it regulates how loud the program is played) from 50 to 90dB average SPL and add a level control with an indication of a slow average at "0dB-VU" (or just good old fashioned needle meters) so we can set the input level correctly, we have the right setup.

Alternatively, if our sources are down to a single one (say PC) and program material we are really interested in has a reasonable even average level (or we use replay-gain on major quiet outliers and re-live on excessively compressed material), we can drop the level control, if we accept that we cannot get attenuation below -40dB (with loudness) and we make sure toasjust the systemgain structure.

For example (re 1KHz):

0dBFS = 2V ->

-14dBFS = 0.4V ->

Preamp "Volume/Loudness" Control at -20dB ->

0.04V to the preamp ->

Preamp gain 14dB ->

0.2V to Poweramp ->

Poweramp Gain 26dB ->

4V to Speaker ->

"Standard" 87dB/2.83V/1m Speaker @ 3m Stereo Triangle ->

Appx. 90dB average SPL at listening position

This is a common Setup, actually often Preamp Gain and and poweramp gain are higher.

So at this point the Frequency should be flat, if we attenuate to a further 40dB, we need 20dB Boost at 20Hz, with 20dB attenuation we 10dB boost at 20Hz.

Of course, we can simply drop the gain of the whole chain by 20dB (or at least 10dB) and just use a 40dB (or 50dB) Range"physioloigical volume control" (which can be managed using just a normal pot or 24-step attenuator), but this may not always be possible, except in a single source setup with source "replay gain".

One way is, we implement our "level control" and "physiological volume control" separate and set the Level control Midpoint to -20dB. Then our physiological volume full up and normal programme material will give us around 90dB. Two separate level/volume control suck big time though (only worse thing I can think of are dual mono volume controls using potentiometers).

A cheapskate 2-control way is to make a "loudness control" with (for example) a 6-way rotary switch, and doing compensation in coarse 5dB Steps including the required attenuation...

Another way, we make a 60dB range volume control and ignore the top 20dB setting midpoint or a little above at -20dB. As a 60dB control with good tracking is not easy using pots and with sensible steps we need > 30 Step switches this seems a difficult proposition using classic methods.

If we are willing to use a MCU and suitable relay/resistor chain setup's or such, we can control level and physiological equalisation separate and have easily >70dB Adjustment range.

Finally, if using PC and digital volume control, we may also apply loudness as part of the gig, for now Iam not convinced of the relative merits of this approach, especially using modern Delta-Sigma DAC's.

Non of these methods seem entirely satisfactory to me and I am still wondering if there is a better way...

Ciao T

_________________
"For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." Richard Feynman


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 9 posts ] 

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group