where we have the most success is assisting Audio companies with their
speaker design. With the help of the Proteus Modeling software developed
for Goldmund, we may help making your next speaker one iof the top speakers
in the world with a very limited intervention of our Acoustic department.
contact us for any enquiry at
Since the beginning of the 20th century, Speakers have been designed
by assembling an enclosure, some drivers and a crossover.
For most manufacturers, even if we are now in 21st century and computer
technologies have been widely available, it remains a manual process
where the design of the bass loading of the enclosure, the selection
of the drivers and the schematics of the needed crossover are mostly
based on the designer?s experience.
In the best cases, measurements with a microphone and a recorder (or
a computer) are validating that the frequency response remains in the
acceptable tolerance on an as-wide-as-possible bandwidth and corrections
may be applied accordingly.
In most cases, however, the frequency response is only adjusted by listening,
adding errors of appreciation, unavoidable even for the best designers.
As for the phase response and its consequence, the linearity of the
group delay with frequency, it is practically impossible to correct
it in an analogue crossover design, and it usually remains ignored by
a vast majority of designers.
So the best speaker designers are very much remaining artist of their
specialty and completely ignore that today a scientific approach is
possible and that crossovers can be made perfect if properly designed
in the digital domain.
And crossover is not the only difficulty. Various diffractions created
by the shape of the enclosure, resonance created by its construction
etc? remain un-corrected in common designs.
It means that, even for the most talented manufacturers, creating a
new speaker is a long and painful process, usually taking long months
to be patiently refined.
Today, however, the solution to solve the complete design issue exists.
This solution is to create a ?Mathematical Model? of the finished speaker,
in which all the above parameters are precisely described. Only the
shape of the enclosure, its construction and the selection of the drivers
are left to the designer. The resulting performance of the crossover
in amplitude, phase and group delay are immediately known and corrections
can be made without listening.
Such mathematical model is extremely complex and requires years of development
for hyper-qualified scientists. So it remains extremely rare in the
audio industry, especially among small companies.
Goldmund decided to write its own Mathematical Model in 2004 and called
it Proteus. It took the company 7 years to get the first results and
the Proteus model has been constantly refined since. The hardest part
of the work has been to find a practical solution for correcting the
group delay distortion (or Time Distortion), a problem that Goldmund
has solved by the Leonardo process, now integral part of Proteus.
Today, when Goldmund creates a new speaker, the enclosure shape is designed
first then the drivers are selected. But the rest of the work is mostly
done by Proteus. By an iterative process, the designer can achieve the
best result in a couple of hours. This is why something like 10 pairs
of new speakers have been created by Goldmund in 2014, most close to
perfection, and very much sounding identical.