The Q113 Evolution is the best speaker of the project so far. That's what the results say after more than 100 people have given the sound scores. Read on to see what makes the winning sound and find out why the speaker got a button!

It's a well-known fact of engineering that if you can optimise technologies to make them work even better, there's a pot of gold at the end of the rainbow. Or something like that. If, for example, a car's engine can be made to use less fuel without losing performance, technology has made a step forward for us. But there are other examples. We can increase a computer's memory, make its work processes faster and make its components even smaller, so that the computer takes up almost nothing. A camera can be made to capture more shades of colour, the bicycle can be made lighter, the GPS can be made more accurate, and so on. It's all an expression of our aspirations as we use the methods of science to measure, calculate and engineer our way to better tools. But science is inherently man-made and probably still has some way to go before it will give us a full understanding of the world around us, because we cannot yet figure it all out.

Fuglerede. The listening team fine-tuned the speaker's sound by changing the components in the crossover.

The speaker is an excellent proof of that

At Loudspeaker Q113, we've built a small handful of speakers step by step, optimising primarily on the technical characteristics related to sound quality. We have not looked much at the economics of the designs, but have only gone after really good technical solutions. We have the best measuring equipment in the industry at our disposal and can measure with an accuracy that is almost unprecedented. In this way, we can develop better loudspeakers using good measuring equipment, even though the loudspeaker's own technology has been similar for many years. But the loudspeaker differs from the car engine, the computer, the camera, etc. in one crucial respect. It is not enough to optimise its technical design, because we obviously cannot measure or calculate the speaker's ability to deliver a sound experience. The speaker's most important task - to capture us in art - is a characteristic that cannot be easily translated into figures or curves. As humans, we can experience sound, nerve, energy, atmosphere and so on, but we cannot measure the experience. This does not mean that good speakers cannot be created on the basis of measurements. But it does mean that for the speaker to be able to convey a musical experience, its sound must be adjusted in place by humans, not by measuring devices. And that means that one of the big differences between speakers lies in this very process.

The Q113 Evolution loudspeaker went into production when 92 of The Engineer's readers chose it as the project's first winner.

Science first, then sound

The large number of volunteers is one of the great strengths of the Q113 project. A team of engineers took care of the measurements and calculations, while another team assessed the sound. No one has been involved in both departments, because there is an undeniable risk of having your opinions influenced if you have a foot in both camps. What should be "best" because the measuring equipment says so, has difficulty becoming anything else when you subsequently put your ears to it. That's why you've seen speakers first engineered to a level where they're as perfect as they can be according to the measurement equipment. And later, the speakers have been tested by the listening team, who have adjusted the sound without knowing the measurements. The measurement team then checked what the listening team had adjusted to make sure everything was technically sound.

The readers of the Engineer have made valuable contributions to the completion of the speaker's sound.

The big listening team gave the verdict

As many as 92 of Ingeniørens readers subsequently took part in a listening test, where the work of the measuring and listening teams was evaluated. This was done (again) without knowing the measurements. As you know, the big listening test resulted in the Q113 Evolution being declared the winner. It was a very valuable test, which I will get to in a moment, because new ideas came to the fore. Right now, the status is that DELTA in Hørsholm has compared the Q113 speakers in a specially designed listening room, where DELTA's expert panel (as well as 10 of Ingeniørens readers) have rated the speakers in a blind test. The results of the test will be, to say the least, very exciting to see. Now it's all about what change the winning speaker underwent when it was trimmed by the listening team. When the sound went from being "technically perfect" to being tuned to our ears.

The perfect starting point. The measurement team delivered an engineering gem to the listening team. Now it had to be evaluated with the ears.

Q113 Evolution begins as a legacy

When you look at the on-axis frequency response of the Q113 Evolution when it was reported finished by the measurement team, you see a very linear and smooth response. From 300 Hz, from where the measurement is usable and up to the hearing threshold at 20 kHz, the response is within 1½-2 dB, which is super nice. We're obviously benefiting from the custom-built drivers and the engineers' simulation work on cabinet geometry, crossovers, etc. And let's get this out of the way right away: If you want a speaker where you can hear even the smallest detail in the sound, or need a studio monitor, just buy the drivers for the Q113 Evolution, build a couple of cabinets, and assemble the crossover shown in the diagram below. You'll have some speakers that are truly "hole-in-the-wall" as they call it. Send me an e-mail if you can't really see the component values in the diagram. You can find me at ole@system-audio.com. The listening team was not completely crazy about the sound. It became tiring to have all the timbral details thrown in the head and the hunt went in to find an adjustment of the sound, which you can listen to for hours, and which still presents the music present and in a detailed way.

The measurement team's version of the Q113 Evolution. Here you can see the diagram of the crossover, which makes the frequency response of the speaker as linear as possible.

8 out of 13 components were changed

The measurement team explained to the listening team which components of the parts filter absolutely must not be tampered with. Some of the components are in a locked relationship with each other, so if you change one component, you must also change the others to ensure that the relationship still works. Each component has a function that affects the sound of the speaker, and the listening team was looking for a balance where the reproduction is perceived as clear and rich in detail, but without causing listening fatigue when listening to music for long hours. Unfortunately, it is not possible to describe all the experiments that were done. But I can tell you that a total of 21 versions of the crossover were built and tested, and the result was that 8 out of 13 components had to be changed before the listening team agreed that now it couldn't be done any better. That's still a radical change from the baseline!

The listening team's trimmed version. With a slight dip of 1-2 dB between 2-5 kHz, it makes the sound warmer and fuller than the starting point.

This is what the "humanised" sound looks like

When you subsequently look at the on-axis frequency response of the speaker that the listening team has aligned, you will see a small attenuation of 1-2 dB between 2 and 5 kHz. This is the metric used to describe the listener's preferred sound. The attenuation is in the so-called presense range, where our ears are particularly sensitive. The sound differences between, for example, different singing voices are registered precisely in this range, and speech intelligibility also lies very much in the speaker's ability to reproduce the present tense. The slight attenuation makes the sound more forgiving of poor recordings and gives it a touch of warmth. The attenuation is small enough that it doesn't make the sound either closed or recessed, yet large enough to give the sound character and fullness. I hasten to say that the recipe from the Q113 Evolution may not be transferable to all other speakers. Speakers have different characters and the grip that works in one design does not necessarily work in another. Then, of course, there is the question of personal taste. The essential message of today's post is that measurement equipment cannot help us all the way with a loudspeaker. Listening and experimentation are needed, even if the technical starting point is in a high class. I even believe that this is where the speaker really gets its qualities, and that in the end, it is the care in this process that creates a winner. Not least, there can be a benefit in separating the people who work on the technical side from the people who have to decide whether the technical side has been successful. The slight attenuation in the past tense is not a new discovery. It is part of a series of known methods of tonal adjustment. What is also called voicing.

Capacitors C2. Sits in parallel with the damping resistor of the tweeter along with a contact.

Capacitor corrects frequency response

When 92 of Ingeniørens readers rated the project's four speakers, the listening team got a surprise. In advance, the consensus was that the Q113 Evolution was the best speaker and the readers agreed with that conclusion. But the informal conversations with the test participants over coffee provided food for thought. Several felt that the treble reproduction of the old SA2K was in a class of its own. Warm and detailed at the same time. And on this point, some felt that the Q113 Evolution couldn't quite keep up. I took the impressions from the many conversations home, and reviewed the speaker again with the listening team. The fact is that there is a capacitor in parallel across the damping resistor of the tweeter. The capacitor's job is to cancel the effect of the damping resistor above 15 kHz and correct the frequency response. Could it be the effect of this capacitor that the critical readers would like to do without? We had to find out.

Damping. Without the C2 capacitor, the speaker will have a slightly warmer response. This is seen as a roll-off at the highest frequencies.

Listening again

The speakers were tested again, this time with and without the capacitor in question. As you can see from the measurement, the absence of the condenser means that the frequency response rolls off gently above 15 kHz, so that in total it is attenuated by 3 dB at 20 kHz. The experiment has been really interesting to carry out, and the readers' input has again pushed the project forward. Depending on the acoustics of the room, the sound balance of the music system and of course personal preference, the speaker will sound best either with or without this capacitor. The listening team believes that the sound is more nuanced with the condenser, but there is no doubt that the sound is more delicate when the condenser is disconnected. Like the SA2K. Therefore the solution was not either/or, but both/and

Knappen. Located on the back of the speaker. The user is supposed to experiment with the sound at home and then leave the button in the preferred position.

The speaker has a button

The button activates the parallel capacitor and corrects the frequency response when it is in the ON position and disconnects the capacitor in the OFF position, giving a slightly warmer tonal balance. The decision to select the warmer sound is thus left in the hands of the user, and in my (soon) 30 years as a speaker builder, I have not made such a solution before. A big thumbs up to the readers who thought we should take an extra look at that part of the sound. We've finished the crucial part of the project, but still need to pick up the wall speaker and the floor speaker. They are the most expensive speakers of the project, but so far they are not the best. This calls for a clean-up. And then there are the specialists at DELTA, who have been blind-testing the project's speakers. Which one did they like best? And can they help us get the failed speakers back on track?