Speaker Design

[Jasper]

I was thinking about different speaker sounds and wondered what a designer has to consider when voicing a speaker.

For example is there always a trade off between different speaker qualities? In other words do you, say, have to sacrifice frequency range for soundstage (unlikely I know) but it's just a "for example".

I'm guessing the answer must be yes, but does anyone have an opinion on what those trade-offs are?

 

[mosfet]

The single biggest compromise of recent times (I’m talking over the last tens of years) has probably been sacrificing efficiency, scale and low-end extension for greater domestic acceptability. Smaller boxes and smaller drivers in ported enclosures. Large efficient loudspeakers in infinite baffle enclosures have a smooth and extended low frequency roll off rather than the sometimes boomy hump of ported designs. But they have to be big to achieve good low frequency extension and hence not in keeping with modern domestic living.

 

[Alex A]

Would love to read a review of your Gamuts Jasper!

 

[Jasper]

Alex_A wrote:

Would love to read a review of your Gamuts Jasper!

Well I'm a bit crap at reviews so I'll give you my impressions and hope that that will suffice.First of all, build quality: they look beautiful. They are the now fashionable "boat" shape design similar to the Cremona Auditors. There is apparently eleven coats of lacquer on the birds-eye maple veneer. The finish is excellent and they give the impression of being carefully built (as you'd expect at this price level).

There are four terminals on the back and the speakers are supplied without jumpers. I have not bi-wired them so can't tell you whether it makes a difference. They are WBT terminals. The main driver is a 7" Vifa or Scanspeak design I believe. Lars Goller who designed these speakers used to work for Scanspeak so I guess he knows his way around these drivers. Talking of Lars, I was in email contact with GamuT prior to getting these speakers and Lars answered emails personally and very promptly. It was a joy to deal with GamuT. Despite, there being quite a delay in supply, they kept me updated and couldn't have been more helpful.

All this would count for nothing if the speakers were a disappointment. Happily, they are not. The speakers are not fully run in yet with a recommended 200 hours needed before they are fully on song. My first impression, during the first few hours was of a slightly thin sound with missing treble and a slightly "overblown" bass. Having had some bass problems in recent times, I entered upon a "one step at at a time" process of tweaks. First I filled the Custom Design RS300 Reference stands (supplied by Shadow - thanks Paul) with the filler, which did tighten things up somewhat but there were still some of the same symptoms. I have a suspended wooden floor so this wasn't a big surprise. Next I attached Cerapucs to the stands. This made an immediate and substantial difference by decoupling the stands from the floor. The bass tightened and the treble and mids became clearer and more prominent. In the last week or so I've placed the speakers on slate slabs and this has made a further, small improvement.

Now to the difficult part - describing the sound. First and most impressive of all is the ability to create a soundstage of depth and width. These speakers just seem to know how to do it intrinsically. Despite a relatively small distance between the speakers of 7 or 8 ft, the imaging is stunning. What you might expect from this ability to image and create a soundstage would be a bright transparent sound but on the contrary, these speakers are incredibly musical (if you know what I mean). They just make you want to close your eyes and get into the music. Fantastic.

However, if I were being picky I'd say that I'd like the bass to be a little tighter, and the trebles just a little "brighter". I suspect that this may come as the speakers run in and it may be that a little more amplifier wattage might help.... All in all, though I have to say that I am very, very happy with these speakers, they just get to the heart of the music whilst being able to dig up stacks and detail and create a wonderful soundstage. Anyone considering speakers in this price range should definitely audition them. They do nearly everything right and I suspect there is quite a bit more to come.

 

[technobear]

Jasper wrote:

...is there always a trade off between different speaker qualities?...

Yes.

Otherwise all speakers would be perfect and sound the same.

There is I believe a saying:

output level - efficiency - bass extension - pick any two

Small boxes restrict bass but you can overcome this with good design. However, the resulting speaker will have very low efficiency. The upside is low box colouration (assuming it's a good box).

Bigger boxes yield deeper bass at higher efficiencies but have large panels which resonate more easily giving box colouration which is harder (and more expensive) to control.

Imaging is often better from small and/or narrow cabinets.

Small and/or narrow cabinets necessitate small drivers but small drivers need large cone excursion to produce deep bass and distortion increases with cone excursion so a big driver that is not moving much will produce lower distortion in the bass so we want a nice big box to put it in but big boxes are not so good at imaging and around we go.

Some other factors to consider:how many drivers, where to set the crossover frequencies (if any), what crossover slopes, what type of loading to use (closed box, reflex, dipole, horn, panel, etc.). All these choices come with a matching set of advantages and disadvantages.

Are you starting to get the picture? Speaker design is part art,part science.

 

[mosfet]

Are you starting to get the picture? Speaker design is part art,part science.

Up to this sentence that’s a pretty fine and well-understood summary technobear. Although I don’t think a ported design can successfully overcome the limitations of a small loudspeaker at low frequencies - the bump at resonance followed by a sudden truncation of low frequency output is a bit of a fix. Never a match for a large driver in a generously sized enclosure. It is however all about compromise.

It’s the “art†part I want to know about. Apart from the aesthetic, what does this mean? I’ve seen hi-fi design described as an art a few times now, all I can think is it refers to the subjective opinion of the designer. Is that art? All the really great loudspeaker designers I can think of (without dropping names) were engineers not artists.

 

[murray johnson]

An engineer designs something to perform a particular job. Efficiently and ideally at the lowest cost.

The Millau viaduct is an interesting example.



Any number of civil engineers could have tendered for that contract. Their solutions would all have been different. Some would undoubtedly have been cheaper to implement but few would have had the elegance of the design chosen. I doubt very much whether the same design would have been picked in the UK. Is this just engineering or could it in some respects, be called art?

An audio designer faces the same problem. There possibly is no 'best' way of making a loudspeaker, amplifier or turntable. The designer will have a subjective opinion on how the problem might be solved and then use suitable materials and techniques to express this opinion in the form of the product. Is it art? Who cares?

But that is what is meant by the part science/part art description of design.

KEF Egg.

[align=right]

[/align]

A cost effective, aesthetically pleasing, decent sounding, mini home cinema system. Being egg shaped is no bad thing for a speaker enclosure but KEF don't make other speakers this way. There's some science in this but a degree of creativity too.

 

[The Hifi Gallery]

very very,

 

[mosfet]

There’s certainly art in the aesthetic, yes murray that’s what I said.

[Norway is the place to go for really stunning looking suspension bridges, one round every corner almost]

It’s the idea that art (whatever this means) plays a part in the nuts and bolts design of loudspeakers I’m less sure about. What does this mean? A little pretentious perhaps.

 

[technobear]

I think it stops being engineering and starts being art when you run out of formulae to tell you what to do and have to go by experience, instinct, trial and error, etc.

 

[murray johnson]

Hi Mosfet,

At the recent London show I had a chat with Brian Taylor (formerly of Impulse loudspeakers) but now of Aspara Acoustics. He was telling me about visits he used to make to see the late Spen Hughes. (founder of Spendor and one of the originators of the BBC monitor sound) While undoubtedly a very good engineer there were many tales of quite Heath Robinson solutions to certain aspects of design simply in order to achieve a desired acoustic/subjective result. Its somewhat fanciful to think that really successful speaker design can be achieved simply by using a computer & a microphone. Adequate design perhaps. I suppose it really depends on your perception of what a successful design is.

You should try it sometime. Take it as far as the conventional engineering will allow you to and then start to realise what choices have to be made re materials (cabinets/cones) and what compromises have to be made re size & cost. Making something exceptional requires many such subjective choices to be made. Ultimately though, it comes down to what you would want the speaker to be able to do. It is an expression of your taste. There's nothing pretentious about it.

 

[mosfet]

I’m a pressure vessel engineer by vocation murray (well I was until a couple of years back following a sideways move for more money but less job satisfaction). There’s no subjective choices here, so I’m coming from a background of pure engineering if you like. Might put some of my comments into better perspective.

Not the same as loudspeakers of course, I appreciate the design of loudspeakers must involve a subjective element and would be poor design if it was straight from AutoCAD to production line without any listening. I still think it’s a little pretentious to call this an art, but if it’s the best word that fits, I suppose it will have to do.

 

[Uncle Ants]

mosfet wrote:

I still think it’s a little pretentious to call this an art, but if it’s the best word that fits, I suppose it will have to do.

Would craft be a better word?

 

[murray johnson]

Some might say it was pretentious to describe the design & manufacture of pressure vessels (boilers?) as a 'vocation'. Were you really 'called' to do it?

Speaker design certainly involves an understanding of engineering principles, the properties of materials, acoustics, electromagnetism. It requires a good deal of experience in understanding how these elements might interreact and how such measurements that are made might be interpreted. It then also may involve some creative implementation of new ideas to achieve a better result & to express what the designer feels he wants to achieve. Is there art involved? What is art? Is the Angel of the North art or a large example of metal casting & fabrication. Is a Ferrari a work of art or a piece of engineering.?

 

[i_should_coco]

One of the fundamental compromises is size vs LF extension vs sensitivity. You can have any two.

If you download a program such as WinISD (freeware) you can play around and it will calculate a box volume based on the Thiele-Small parameters of the drivers you choose. That will get you something which works. I would imagine a lot of budget speakers are designed something like that - just plug the figures into your chosen design program and get the box made - there's no budget at the low end for lots of R&D and tinkering.

Detailed voicing goes much further than that, you can spend a very, very long time twiddling with the basic box design to get something sounding good - the box may have a certian volume that physics and the driver characteristics dictate, but that can be implemented in various proportions.

Crossover topology and component choices can make a very big difference as can box material and construction.

Things like radiusing, quality of glue, even veneering can change the stiffness characteristics and so the sound.

And all that is only the start...

 

[Purite Audio]

What is the least compromised design possible do you think, something like Romy's macondo?

 

[i_should_coco]

To be honest, I think *all* speakers are a compromise (size, cost, LF, smooth response) in some way. I don't believe that it's possible to build one without some tradeoffs.

I've not heard Romy's but I know someone who has and says it is very good. Of course, I would imagine it's not to everyone's taste. Macondo is of course a compromise in that it's bloody enormous and expensive.

 

[enjoy_the_music]

There are many analogies one can draw to this subject!

I think some of the best speakers just come from sound principles (engineering truth and fact) and artistic process (what comes from within the designer with a view to stirring the emotion of the end user).

The latter is more empirical/suck it and see...and often provides the most musical results!

Subjectivity is what makes it all so fun..and at the same time so much of a pain in the ass.

 

[Purite Audio]

Most uncompromised designs are expensive, in whatever field, but in terms of an uncompromised transducer where does the state of the art lie?