26 April 2015
A Question of „Loudness Correction” in Perception
Some time ago author’s colleague stated to him a question why some of today’s amplifiers have push button „Loudness”, other „Contour” and some „Selective Tone” whilst most including those expensive ones lack any of them.
Author usually was not too much about it – his best experience in audio 40 years ago happened to be with tube amplifier of radio Stradivari 2 working with developed then by him speakers using one wide range speaker and also alternatively with two way also projected by him Pioneer/ Tonsil boxes. . In Stradivari 2 producer had provided may be most sophisticated „loudness” correction seen by him, and no any deactivation was applied. Six outlets from special potentiometer while normal potentiometer has three proves the difference. Each lead produced own frequency correction. Author at this time had no idea how important it was for obtained result.
For comparison average circuitry found today also in „High End” stuff.
Another really impressing circuitry producer applied in even more surprising radio Beethoven where tube was covered by special „Loudness” feedback instead of parallel RC filters. In this solution, with one „loudness” lead, correction characteristic was not fixed but flowing type according to volume position.
Next author’s choice was stereo amplituner Chopin produced by Videoton – Hungary. „Chopin” was using also potentiometer with 3 lead for „loudness”, three characteristics cooperating with volume, and also with no possibility to deactivate.
[02 May 2015 - schematic drg. updated.]
In Stradivari 2 and Beethoven loudness tone correction had been up beating both – low and higher bands. In case of Chopin only low band had been emphasized with volume down. This type of correction effected in very soft sounding but with exceptionally good low range it was pleasant and appreciated, and it is just one more confirmation that it is low range responsible for overall reception.
Today most stuff offers deactivation of loudness control function. Why?
First think tells: It is no more appreciated by demanding listener, so producers in order to honor customers, followed their choice. Hmmm…..reality seem to be more complicated. Any way why demanding customer should have to hate sophisticated improvement? Honestly ….sometime he will .
Let’s start with what is in practice loudness correction?
It is device assuring that range of perceived tones will not decrease when lowering the levels of produced acoustic pressures. Humans hearing sense has this ability to differ the silent sounds from loud ones, by changing perception of it’s quality. Silent sounds are classified as less important, taking less attention regarding details, and loud are heard with all nuances. This way, when we are talking in a crowd, actually we understand nearest and loudest voices instead of all. In case when we listen in silent room and at low level, to sounds of instruments recorded with closely positioned microphone, we need to amplify nuances to have full impression of what is recorded. That is because when we are listening to sounds close to us, we experience naturally their bigger loudness. In other words with strength of sound comes applicable timbre. By comparison: in piano – strong hit incorporates different timbre then normal. Listening to whisper includes close distance, and relatively loud sound. This assures quality of details. Shouting from distance opposite. Replay of such materialwith lower than natural loudnessfools listener and removes clarity.
To make things more confusing, but in appreciation to truth, it is worthy to understand that recorded material also incorporates „loudness” information. Let’s take gig recording- musical part is done in remote position to catch the „impression”. It will be not detailed close recording. So it is not only humans ear dependency, but also that bass is dispersing in all directions, middle is directional and high are…highly damped by all surfaces . . Audience also should be „remote”. Then, for listening to gig, true will be to perceive not very much nuances because the best transportation for gig sound is middle band. Thus, in such places acoustic engineer has to upbeat low and high bands. In studio singer produces sound in direct position to microphone with no reverberations – „true” replay should have maximum details. As we see we are talking about distance of recording microphone, because microphone responds to distance too. In author’s opinion we need to regain quality associated with original sound volume to perceive true transmission. Without correction, with turning down volume, we perceive a quality of distant sound – ex standing outside, and hearing gig from two blocks away. Or standing outside concert hall, with door opened. With loudness correction we can move inside.
May be not all readers will go along with author, he must also admit that it is not always clear even for him, until he perceives difference between flat sound and the same with proper compensation. Having even best flat sounding stuff without correction, we tend to push knobs ahead to get what we want . Unfortunately „bass – treble” adjustments are not capable to solve it. Equalizers are non accurate too, and settings are highly demanding and time consuming. Following any theoretical curves is misleading, as it is directly not applicable. Sadly no any support is available. This brings author to another phenomena - audited by him outside the listening room not corrected music, sounds usually in way more natural. May be one of reasons is that in such situation we simply get the equality between what we hear (sound from distance ) with what we perceive – sound IS actually in some distance.
When listening to flat, my colleague even complains that it must be something wrong. He wish to hear more bass, even if it would be not proper. Most of listeners just gave up, and accepted already decreased range in moderate levels, or alternatively suspects their stuff is not enough. Another situation is when listener has speakers with characteristic not flat, but type loudness curve. It may happen with resonating boxes, and something author described here as common fault in multiway systems. . For special given conditions sound may be close to desired, but depending on spectrum, it will work not for all recordings. Actually applying to such speaker loudness control can bring „over control”, and it is unfortunately usually happening. . But let’s forget this other problems.
Nothing had changed in physiology so why „Loudness Control” is not applied 100% today?
To answer this question we shall have to look at what has changed within last 60 years. . First problem is already pointed – today few band speakers with crossover filtering, and phase distortions, pose danger of „over- control” . Many users experience that pushing „loudness” button makes significant changes not always welcomed, or effect is partial. For satisfactory „Loudness Control” must be assured parallel both 1)speaker with defined dedicated to provided „loudness correction” characteristic and with known efficiency. 2) More or less defined distance to listener( listening conditions).
Was it done in old 50-60 radios? Yes! The speaker system was permanently fixed inside enclosure and matched to loudness correction by the manufacturer. Usually manufacturers of luxurious radio receivers with loudness control were the same experienced manufacturers of dedicated to them speakers – 100% match!. Now distance from listener: in that times no remote control was used and tuning to radio stations was a duty to listener. So he was expected to stay close to the speakers. And he was. No need to mention also, how positive it was to listen to direct sound from radio with little room reverberation. Audiophiles should understand what I am now saying. .
That last note is in authors belief one of factors why some demanding listeners, even today, prize old luxurious radios timbre. Pure in base of operation tube amplifiers, simple and phase coherent speakers with eventually wide and soft frequency distortions, not too high expectations and…. loudness control! That simple but charming sound is usually underestimated by most. And author thinks that beauty of not powerful sound do not require too much.
How about today? Change has came with stereo system. At start, in it’s beginning in 60 – 70ties, producers still followed good old tradition to produce amplituners and speaker boxes sold together, in common design… But in case of damage danger of replacement by buyer and loosing precision quality had grown – original speakers were usually supplied only by authorized service, and no cheap. Damaging of free standing type was easy by using with other stuff, overloading, falling down. And this speakers had to be best sounding – they were responsible for attraction to radio - (today deaf amplifiers attract only with instruction manual and prospect ). For such very expensive production any modifications or troubles to get replacement posed danger of damage of manufacturer image. Specialized speaker manufacturers have easier life – they did not take responsibility for all. In this way the idea of free attached brand speakers made by electronics manufacturer failed, killed by lousy users and time .
Example author’s tube/transistor „Chopin” from Videoton. Equipped with permanent „Loudness Control”, priced highly and the same blessed by owners for deep soft and detailed sound radiated by made by Videoton unique speakers. Blessed… until speakers were finished by their weak point – , evaporating within few years soft sponge suspension of woofers. Any try to adopt another speaker producer serial shelf speakers of to unique loudness control and whole adjustments was impossible – „Chopin” sounded with them poor. It was the case which happened to author, who purchased second hand Chopin already with no original speakers. He was very successful in learning a lot about speakers and „Chopin” construction until he matched pair made by him . Too big problem as for so expensive system.
Today amplifiers must operate acceptable with standard speaker boxes made by another specialists. Is Level „acceptable” a perfection? Nooo.. In this case matching the „Loudness” with speakers is out of reach. It requires much more accuracy of resistances and capacities then standard 10%. Author found in his practice, that for top sound quality sound responsible elements value tolerances should be about 0,2% !! May be that is the reason why some buyers wonder why they were auditing amplifier, were amazed, took new in box and…? Even for the best market speakers differences in their efficiency, and distance from listener make „loudness” problematic, and turn to nonsense. If composed with perfect matched speakers then ok, but if impression is bad then better turn it „off”. And at this point (lucky!) we find button to switch our „Loudness control” off. No obligation to use. Some producers simplify that option in a way of only two characteristics – „flat” and fixed „loudness” As it strongly upbeats sides of band, we must be careful not to use it with full power – our ears may not warn us of total overload of tweeter, power end and woofer. In old design it was secured by volume knob control cooperation . So today we can have it in our amplifier but due to bad match with speakers it may be remote from desired perfection.
Now about the conclusion. Should we today forget „Loudness Control”? In author’s opinion not. Only this correction provides comfort of silent listening or normal listening with great impression without fatigue by noise. Even he would go further – it is impossible to experience total pleasure of listening. Obvious proof of that, is observed today tendency to master digital music in a way to decrease original dynamics range to minimum. That is considered by demanding customers as nonsense and is called „loudness war”. The result is amplifying l silent parts in order to keep the frequency spectrum constant . To get rid of it, and to apply right spectrum correction, author sees few problems to be solved. First is how to obtain signal to control curves. Before it was volume knob position – now, often when volume is controlled by electronic, we can reach more precision by measuring voltages on output and using the electronic controlled RC filters. Better way would be even to use the measuring microphone in listener’s location but it seems complicated and vulnerable for miss operation. Second problem it is of course production of dedicated to amplifier speakers and later assure its supply for replacement. This is critical -author’s experience shows that required accuracy of system is really more than suspects anyone who had audited it. And that is why it only can be done by the same producer. Third is to work out matching between correction and speakers. According to author’s actual knowledge that matching has rather little to do with „equal loudness curves” They are complicated and describe only our hearing sense. Why little? Proof is while our ears are still the same we don’t need to have „loudness correction” in natural listening! We need it only when incorporated spectrum of material „do not belong” to actual listening loudness.
It shows surely additional coast of production and grow of price tag. But thinking about it as an alternative or market gap between standard speaker-less amplifiers, making for buyers hard times to match best for them speakers, and top quality High End with very highly priced speakers, seems not so impossible. Author remembers such stuff done by many producers in 70-80ties usually appreciated by users . Then market divided in cheap for all always happy and very expensive for few always unsatisfied audiophiles . In authors opinion final result is exciting and justifies the idea.
Update 28 May 2015 – some information altered for better explanation.
27 April 2015
Tuning of Loudness Control in Pioneer SX-34
At the beginning of use , in order to avoid misleading in time of search for original, „as recorded” sound, author disconnected „Loudness” circuitry and spared it as shown on diagram in „Modification” with 200 kOhm resistance. This allowed setting up speaker systems and judgement of amplifier’s timbre. Another reason was that author considered implemented loudness result „useless” . It wasn’t a producer’s fault – as mentioned before producer matched this control to original or advised speaker boxes, which are not around . After job was done, author returned back to loudness circuitry to make it working in new circumstances.. Original circuitry diagram is shown below.
Operation of original volume control was troublesome. Author described possible reason in previous part regarding amplifier modification
Unwanted capacities marked in red create the RC feedback with input resistances. The higher resistance the lower affected frequency. Due to this feedback only one position of potentiometer was trusty – maximum, where input resistance is reduced down by previous stage. But in this situation was a problem of input devices not having own volume control to adopt to maximum volume in Pioneer. Also a trace of noise was heard.
All this forced author to make serious approach to adjust band spectrum outcome at low positions of volume potentiometer. Lucky provided by manufacturer in potentiometer loudness connection have made it possible. With original values of R and C elements in loudness circuit switching it „on” was for author not acceptable and for that reason it was primary disabled.
Developed by author upgrade of loudness tone control for Pioneer SX-34
In project author compared basic data to exceptional (as for him) results perceived with Stradivari 2 mono amplifier equipped with complicated loudness . Facing actually no availability of double potentiometer with 4 sections the original one found in Pioneer with two only had to be used.
For Pioneer SX -34 low cut off frequency was evaluated to
RC= 5nF x 68K = 340,0.
In author’s concept are different R and C values and
RC = 9,86nF x 33,5 K = 330,1
It’s not much difference in „cut off” but double more value of compensation. .
For high frequencies originally RC is 0,05nF x 800K = 40 . In new concept
RC=147pF x 800K = 117,6
- a big difference to Pioneer. Let’s look at Stradivari 2 data – Capacity C for top regulator is 150pF, probably R=800K then our RC is 120 which means this band is almost according to Stradivari 2 and no change in volume of compensation comparing to both diagrams . Feedback signal from next stage was apparently elaborated by author by following an idea found in Beethoven. The reason of applying two devices in one circuitry was existing value of partial resistances in potentiometer . For standard Pioneer circuitry loudness, when turning down, begins to activate in half position.The strongest action is observed in lowest positions of potentiometer. Upper is then degraded in sopranos. With idea of feedback the strongest action and improvement is observed just in half of turning where feedback signal is at maximum. Coupling two allowed better sounding for upper range of potentiometer turn. . Additionally this solution gave more room to possibility of adjustment.
Tests of all kinds of sources and different sound material had presented not recognizable outcome in comparison to Stradivari 2 as desired.
Example of frequency spectrum in loudness 50% action ( app 50% of volume)
100 200 300 600 1K 3K 6K 11K 22K
in 75% loudness action (app 25% volume)
In 25% action (app 60% of volume)
Amplifier with loudness "off" - (100% volume)
29 April 2015: Update 1 - Schematic of loudness circuitry modification altered according to actual version.
02 May 2015 - added notes to graphs.
28 May 2015 - some alterations done for better understanding.