Use of computer software with two way speakers such as the Mackie HR824MK

[David Pinnegar]

Hi!

I have noticed that computer organ simulation users are still enthusing over Mackie HR824MK and similar speakers as though there are no alternatives for Organ reproduction. For the avoidance of doubt, this post applies specifically to organ reproduction only and the circumstance of the desire to create an absolute reality of the sound of an organ pipe and collections of organ pipes. Those are very specialised criteria which do not take away from the generality ofthe overall excellence of reproduction that people enjoy in the course of using such speakers.

I was willing to stake my reputation on the fact that two way speakers, especially the modern incarnations with active crossovers are not best matched to pipe organ reproduction and do not give the computer organ simulation signal the best medium to shine through.

"Proof! Proof! Proof!" - the Hauptwerk forum asked and rotten tomatos were thrown at me.

Yesterday afternoon an organist turned up to practice, as they often do, and as he played I crawled around the instrument to check everything was operating as it should.

What I have been saying on the forums was absolutely confirmed: the Diapason and in fact a Posaune had been playing through a two way speaker. It had sounded plastic. (On the computer organ simulation forum I mentioned that this occured on some stops and not others, and bad eggs were hurled at me.) I replaced this two way speaker with a dual concentric two way speaker. It still sounded plastic. I rediscovered yesterday that I had resorted to putting that speaker onto a sub-woofer filter and used a different type of speaker for everything above tenor C . . . and those stops are now alive.

One only has to look at
or
to see that whatever the brand of speaker, Behringer, Genlec or whatever, speakers that have a disposition that looks like this simply cannot adequately and realistically reproduce the organ range and continuum of freqencies. Just visualise the pipework and the size of the pipes, of course directly relating to physical requirements of sound and imagine which pipes are coming out of which speaker and which parts of which pipes are coming out of each speaker.

Speakers like this are designed for instance to shine for singer and percussive bass, melody and accompaniment - or one might even argue listening to modern electronic music on radio 1, percussive accompaniment with low tones and percussion treble, and a melody that doesn't matter if it's only somewhere in the middle.

Over 99% of all music sales are non-classical, so any comments that one sees written about such speakers in relation to classical music will be in the tiniest minority.

Proof? http://www.dogsonacid.com/showthread.php?threadid=570199 .
See the comment of "The Inspector" -
- "they are just not critical enough" but particularly also the comments of "Feeding Cone"
- "sound like most of the speakers in the world"
- "I thought that the upper mid-range was a bit iffy"

Of course Diapasons need absolute clarity in the upper-midrange. They don't have a complex structure of dissonant harmonics and so their output might look like the graph of the 8767 054 Hammond registration on
http://www.stefanv.com/electronics/hammond_drawbar_science.html

The upper midrange is critical to the relationship of harmonics in the sound.  If the peak near 2kHz is disrupted by coming out of the two units at once, then the sound won't sound as real as other solutions offer.

Incidentally, that graph is demonstrative of the use of technology in replacing the use of thought . . . because by definition, the Hammond is incapable of producing those frequencies as shown by the graph. Rather than a bell-like peak, the graph should show a spike at 440Hz and its harmonics. Why? The Hammond is incapable of producing 439Hz and 441Hz nor any frequency in the one note between 440Hz and 880Hz. Having said that, the Fast Fourier Transform that produces such graphs examine a set of frequencies within a spectrum of time, that time being limited and thereby leading to the infill of frequencies  we see in the graph. But perhaps that may also relate to the way in which our ears perceive sound? Bearing in mind that our continuity of vision is fooled at the point of 12-15 frames per second, time limitations within our mental processing will give rise to perceptions in the frequency domain are not directly apparent.

Why bother to urge people to use speakers better designed for organs?

If we are to encourage enthusiasm in organs, whether pipe or electronic simulations, the result must be captivating. That will only occur with instruments that are nothing short of aurally top rate. Second best is not good enough. (By this I don't imply expense - expense and suitability do not always equate)

One would think that computer organ simulation users might desire the direction of organ excellence rather than in any way a brand worship.

If one is not allowed to put pointers along the way that suggest that other solutions give better excellence without being suspended from the forum, the forum and the product associated with it becomes only a must-have fashion accessory instead of serving a quest for perfection.

Best wishes

David P

POSTSCRIPT: http://greygum.net/sbench/sbench102/caps.html is interesting on distortions introduced into systems by capacitors

[KB7DQH]

I have had the opportunity to listen to a variety of "canned" music reproduced on what one would consider "high-end" audiophile speakers, along with appropriate equipment feeding them.  I noted that the mid-range frequencies reproduced extremely well, vocal pieces recorded in studio reproduced through this equipment sounded like you were standing in the studio.  Among the pieces of music auditioned were some with real pipe organs, some accompanied by orchestra-- and the system made a good accounting of itself in that important range of frequencies required for pipe organ reproduction, as was intended by the builder of the system!

What I did notice was the inadequacy of the "powered subwoofer" employed in the setup... One kilowatt  in a bit over a foot square cube feeding a single "long-throw" foot-diameter speaker.
Yes, you could sort-of hear 32' pedal departments, but that was about all,  my four twelve- inch drivers mounted as an infinite baffle in the corner of my listening space, much larger than the one
occupied by this "boom cube" and being fed with a quarter of the power and properly balanced against the rest of the speakers in my system, allow one to actually feel more than hear those notes...

The two-way system in my kitchen sorely lacked adequate midrange output so I added a pair of 6X9" cabineted automotive units which also have separate tweeters. With no crossover available I simply wired them in series with the two-way units in the hopes that enough audio would be left over to drive the smaller speakers enough to do the job, and , remarkably it worked, but, I think another 3dB fed to the 6X9 pair couldn't hurt.  Simply a matter of calculating the right value shunt capacitor to place across one of the other speakers to "shift over" the right amount of power at the right range of frequencies to get the job done...

I have a similar problem in the audio reproduction system in my "portable amateur radio station"...
But I may make a much more radical change there after messing about with some compression horns... The ones normally employed for "handheld public address" or alarm sirens, etc.... 

Eric
KB7DQH

[revtonynewnham]

Hi

I'm not at all surprised about the bass problems - organ music is almost unique in requiring a substantial amount of continuous power in the extreme bass (down to 16Hz for a 32ft rank).  32ft Flue pipes are among the most difficult to produce electronically - 32ft reeds are easier as a substantial amount of the output is in the harmonics, which are not only easier to produce from a speaker, but also have the psycho-acoustic advantage of regenerating the fundamental pitch (c.f. Compton's "Harmonics of 32ft" stops that serve as a substitute for 32ft reeds).

Most proper studio monitor speakers are either 3 way, or attempt to shift the crossover away from the critical 2-3.5kHz region.  The acid test of a monitor speaker is reproduction of human speech (male & female) - a far more difficult task at mid-range frequencies than any musical genre - but in the semi-pro/hi-fi arena I only know of one reviewer who regularly auditions speakers on speech (Hugh Robjohns in "Sound on Sound" - a BBC trained recording engineer). 

Unfortunately, crossovers are a fact of life in most loudspeaker designs, especially if a large frequency range has to be covered - the art is designing the system so that the inevitable artifacts (phase distortions, comb filtering, etc) are minimized for the particular application (even using a single speaker plus sub-woofer needs a crossover).  I suspect that one of the problems with producing organ sound is that some part of the harmonic series will "spill over" into the tweeter and start to sound artificial - especially given the fact that most general purpose hi-fi and Sound Reinforcement speakers put the crossover smack bang in the middle of the speech band's harmonics (say 1kHz-3kHz) - just where the ear is most sensitive.

Every Blessing

Tony

[KB7DQH]

I'm not at all surprised about the bass problems - organ music is almost unique in requiring a substantial amount of continuous power in the extreme bass (down to 16Hz for a 32ft rank).

I have an old copy of a book entitled "Building Speaker Enclosures" that states that one should "reduce the power rating of a speaker by 50% when used for an organ" ;)   For exactly that reason.

especially given the fact that most general purpose hi-fi and Sound Reinforcement speakers put the crossover smack bang in the middle of the speech band's harmonics (say 1kHz-3kHz) - just where the ear is most sensitive.

Which is why 3-way systems with a proper speech-frequency driver I find "behave better"...

My favorite tweeters are of the Piezoelectric types as they electrically look like a capacitor and thus
require no crossover network...
Moving coil speakers will present an increasing load as the frequency they are presented with decreases, such is the nature of an inductor... Especially one surrounded by lots of Iron in a monstrous magnetic field.

Eric
KB7DQH

[David Pinnegar]

Hi!

Here's a practical demonstration of the tonal areas of organ stops which can be affected by speaker anomalies:
http://www.youtube.com/watch?v=vii521NGSos

Best wishes

David P

[David Pinnegar]

Hi!

From a comment left on YouTube today I rediscovered the video I did about the effect of speakers on electronic organ tone and realism:
http://www.youtube.com/watch?v=vii521NGSos

Whilst Manger speakers are very excellent, they are clinical . . . and for hi-fi I use units which give unparallelled dynamics the other day I heard some Ruark Crusaders which really brought the bottom end of organ music to life in a domestic sized room without the assistance of subwoofers.

For more than domestic use, then specific designs can benefit the total effect. It depends whether one is aiming for a domestic hi-fi reproduction in a living room, which I suspect most are, or a performance organ intended to _be_ the real thing rather than merely like the real thing. . .

Best wishes

David P

[dragonser]

Hi,
I recently was asked to repair a Mackie SRm350 powered monitor.  After fitting a replacement tweeter I tried the unit out ....
and yes when using a sine wave tone at medium volumes you could hear that as the tone was swept from about 500 hz to 3khz the tone did come from both the tweeter and the main bass speaker. on this model of Mackie there are two power amps, one for the bass speaker and then other for the tweeter.  the cross over is an electronic design.
I think that as has been mentioned there may be phase problems as the sound comes from more than one source ? it also makes sense that you might only notice this on some types of music / speech .
regards Peter B

P.S has anyone tried using a coaxial speaker [ that is one that has a bass unit with a tweeter built into the same chassis ? ]
I know this is a different application but one of these units was recommended to me to replace an old speaker in one of the Leslie speaker cabinets.

[David Pinnegar]

Dear Peter

Your report is interesting as I'd have expected the electronic crossovers to have had third or fourth order filters with slopes of at least 18-24dB per octave. . . Not to do so really removes the theoretical raison d'etre for active speakers.

However, this is why I dared to suggest on the Hauptwerk forum that there might be better performance speakers to use for organs, if one is to create the reality of a performing instrument rather than a living-room hifi reproduction and those for whom Mackie speakers are a religion took this as an offensive insult. The modern software-writing generation simply have not had experience of the years when people concentrated on hi-fi as the latest technological fad . . . and these speakers are intended for the mass market serving the needs of 98% of amplified music which has no requirement for the midrange detail that the 2% of classical music demands.

Other solutions are not only possible but work very well. For conventional two way speakers with passive crossovers, I find Tannoys very convincing. As the treble horn is extended by the shape of the cone, one has integrated sound without a hint of division: the two units aim to provide a single wavefront. There are a pair of ex BBC 15 inch Tannoy units on ebay at the moment for around £600.

I'm working on some experimental coaxial units at the moment specifically intented for organ use. The difficulty is finding the exactly right electrocal and mechanical crossover point and characteristics.

I have found Manger transducers clinically accurate but arguably they don't quite give a performance - in contrast a super-accurate reproduction . . . and there are other units that I have demonstrated at EOCS meetings which are so supreme that I would not mention them in a public forum as their performance could make Hauptwerk based instruments be far too attractively competitive to pipe organs . . . A violin through Mangers sounds like a violin whereas the other units make it recognisably a Strad. . . .

There's a thread on the forum that I wrote about repairing Tannoy concentric units: if one is lucky enough to find a pair that needs reconing, often one can salvage the existing cones and voice coils and save a significant sum both on purchase and in repair.

Best wishes

David P

[Holditch]

Unfortunately these days Mackie SRM speakers have very low cost far eastern drive units, very cheap in comparison with the RCF units that used to populate the loudspeakers when the big American conglomerate tried to take over the very old and well established Italian loudspeaker manufacturer, so even if the crossover network was well designed, the audio quality and reproduction of the loudspeaker is let down by shoddy drive units with bad power compression characteristics, and cheap components. The comparison between a Tannoy dual concentrics and this cheap tat is vast!

If you ever take one apart you will know what I mean

[David Pinnegar]

Dear Marc

;) I was thrown off the Hauptwerk forum for mentioning such heresies!

The reality is that accessibility to organs via computer simulation, which is bringing in a lot of people to reignite interest in the King of Instruments, is only going to have lasting effect if the results of computer simulation are sufficiently stimulatingly inspirational. It's not going to happen if people don't pay proper acoustic attention to the final stage of the reproduction process.

From the prices that I heard people were spending on that forum on such active speakers, it's apparent that Tannoys on ebay are good value in comparison. However, such units aren't the only solution for all purposes :)

Best wishes

David P

[dragonser]

Hi,
well the unit I worked on was a few years old so may have been a different tweeter ?  I was quoted over £ 100-00 for a replacement drive unit from a Mackie agent !
regards Peter B

Unfortunately these days Mackie SRM speakers have very low cost far eastern drive units,
If you ever take one apart you will know what I mean

[David Pinnegar]

Hi!

Having been impressed by 12 inch units I have just tested a modern 8" Tannoy unit and I'm sorry to say that whilst good as a hi-fi speaker it fails my reality test . . .

Perhaps the acid test is a broadcast or recording of a live concert: if you can hear more than audience coughs and one is aware of the audience breathing around you, then such speakers are capable of not mere reproducing but producing reality . . . If anyone is in the UK, then also the broadcast quality contrast between BBC Radio 3 and Classic FM is telling - if your speakers sound great on Radio 3 but you can't listen for long to Classic FM on account of a grating or distortion, probably a digitisation noise, then such speakers can produce a reality . . . If the speakers are fine on both stations, then they are only capable of reproduction.

An organ adviser who sometimes writes on this forum told me that when asked for advice on electronic organs he specifies speakers with a cast chassis . . . I'm not convinced that that is the telling criterion - perhaps for a start I would look for paper cones - straightforward paper, no plastic, no composite materials . . .

For some time I have been working on transducers that are both capable of realism and ultra thin that can be installed in a pipe-organ being mothballed behind the display pipes and not involving nor endangering any pipework behind . . .

Best wishes

David P

[David Pinnegar]

Having been impressed by 12 inch units I have just tested a modern 8" Tannoy unit and I'm sorry to say that whilst good as a hi-fi speaker it fails my reality test . . .

Hi!

I find it so very sad that reproductions of the pipe organ that people are liable to confuse with the real organ can mislead "the man in the street" into not appreciating the King of Instruments as it should be appreciated.

It was for this reason that on the Hauptwerk Forum I took issue with those promoting the us of general purpose PA and similar "HiFi" speakers for the reproduction of organs with Hauptwerk software representing the "organ" as a real organ. Having disagreed with the software writers, I refused to be bullied into revealing quite how one much achieve better for the reason that I did not wish to encourage electronic reproductions generally to have the ability to displace the grandeur of real pipe organs . . . and paid the price of disagreeing with the software writers by being thrown off the forum. This, however, is hopefully a forum where people are allowed to disagree and thereby to enable better results and better appreciation of the Organ to be achieved. . . .

Eminent audio engineers on the Hauptwerk forum posted justification of conventional speakers especially with a paper relating to the design of BBC monitors . . . It was very learned and appeared the very pinnacle of perfection . . .

However, of conventional woofer - tweeter designs I have found the Coaxial designs to be far supreme although when I played a 12 inch Tannoy to a friend he remarked that it was the sound of his local cinema . . . and Tannoy 8 inch units can be exciting but vary between the very good and the really good without achieving the total excellence of recreating reality. There is a difference between the models with plastic cones which are very good and the ones with paper cones which are really good but with a stiffer suspension have less good bass.

However, whilst looking at forum stats, mention of Member Colin Pykett here led to
http://www.colinpykett.org.uk/reedpipetones.htm
and this explains the justification for my apparent prejudice against two way speakers for some organ stops. This has to do with formants.

There are many organ stops that can be reproduced with a fundamental note and a few harmonics and a fizz of upper harmonics and white noise on top.

However, organ sound is more complex than this, and in its totality much more complex in its multiplicity of pure tones requiring zero distortion and zero intermodulation than most other programme material.

Colin Pykett's analysis of Reed pipes pointing for instance to the predominant formant of the Vox Humana at 1850Hz suggests exactly why two way speakers, even top ones such as http://media.musicalplanet.com/pdf/TNY310.PDF plonk the crossover and its related phase and dynamic disturbances just exactly in the sort of regions that are critical to such ranks of organ pipes.

I am not going to discuss exactly what I do recommend as a design philosophy here for "pipeless" organ speaker systems for the same reason that I did not on the Hauptwerk forum but urge anyone commissioning such systems not to be satisfied with the second rate use of conventional speakers. When such instruments pose as "The Organ", for them not to be anything less than outstanding is damaging to the reputation of the instrument.

Best wishes

David P

[David Pinnegar]

Before closing my browser windows on Colin Pykett's research articles, a little browsing further brought http://www.colinpykett.org.uk/principals1.htm in which

appeared. It's interesting how the tone depends particularly on the Seventh harmonic, near to two octaves above the note.

For the middle octave of our organs, between C=256 and C=512 Hz (approx - the mathermatical pitch is an easily remembered guideline) this means that any speakers used to reproduce diapasons should not have any dips in their amplitude response nor phase anomolies between 1kHz and 2kHz for the middle octave of our instrument nor between 2kHz and 4kHz for the treble octave if the bright and distinctive tone is not to be lost and muddied.

This immediately rules out many brands of conventional speakers.

Best wishes

David P