By skyforbes 
This is a very special entry and a marks a point in the evolution of the system where i shall say i managed to reach a breakthrough in the overall quality of the system.
I am not sure where the benchmark is (and maybe one never knows for sure), but i always knew the bass representation of the SVS SB-3000, as impressive as it may be as a 13’’ subwoofer, was not formed a seamless, invisible unit with the Genelec 8030c.
Now i knew that there is this table here:
which ranks i think after Sean Olive preference rating model the combination of a sub and a speaker. And in the end this was also the reason for selecting the 8030c in first instance. This is a decision that is hard to beat.
But what the table does not show is what subwoofer plays best with which speaker. And that may be also something combinatorically quite costly if not done automated in some way.
So how did i now manage to make the recent breakthrough in sound quality? What happened? Well first let us take a quick step back and look at the 2 fundamental feedback types for subwoofer construction:
- Feed forward systeme
These rely on the fact that one can model the impulse response in total. Examples would be SVS, but also Perlisten, and in a different way, also Abacus (Dolifet) systems. And also REL for example. The REL subs (the good ones) seem to be amongst the fastest when it comes to group delay. Perlisten has quite an advanced FIR based prediction model, while the Dolifet system is elegantly modelling on the pure electrial lane, somewhat eliminating deviations already on the level of measuring voltage and current at the output of the amplifier, deriving the predicted motion of the coil (and indirectly so the membrae, to some extent). Everything is done in the voltage and current domain! This is a very elegant and tempting concept: the correction is more or less happening in realtime because it is not entailing any matter objects.
In general, those system tend to have a low group delay in between 30 and 50ms @40hz, in room. - Feed back systeme
There are also some systems that are using feedback cycles to eliminate distortions in the system. The point with that approach is while one can model things on the electrical (current, voltage) lane like the Dolifet system, or on the to be expected or predicted room (or membrane) response, using FIR DSP presets, all those cannot exactly model the entropy in the system in total. When i realized this, i knew why so many Audiophiles consider the Velodyne DD10+ the best subwoofer (in the non-6-digit-domain). This model is the only model using a piezo sensor to measure the actual movements of the membrane. So this means leaving some ‘physical artifacts between the membrane and the ear’ aside (Heat, micro-air motion in front of the membrane), that way a very high level of precision display is archieved using this model.
PSI Audio subs do it in a different, indirect way, but are also considered quite a robust model with excellent results, and the technical setup lies somewhat in between Abacus and Velodyne Servo technology:
In contrast to the Dolifet techology, the real delta is measured on the current and voltage level, but not on the membrane. The advantage is the correction of thermal (entropy) and magentic deviations that are hard to model, and this happens in < 10 µs. In some ways, all concepts are interesting.
So which one is the best? 😀 There is no way to not use a smiley at that point. It is probably hard to say. To me, the assessment that really checking the impulse response on the driver level seemed like the most logical thing to do (since the membrane is what moves the air — not the amplifier or the current), and the human feedback from Audiophiles seem to support this notion.
One can find beauty in concepts, and logically, measuring and correcting at the coil is probably the best compromise conceptually. The PSI subs are considered amongst the best worldwide. But there lies a certain beauty in correcting the deviations on the ‘last physical domain where something can go wrong’, if you leave air properties aside as mentioned. This seems like the most robust system to me.
So i knew: this was what i had to go for. These days (10/2025), the newest server model from velodyne is shipped for the first time: the SPL-X series. So since my room is not too large (about 30m2), i also knew that i would rather go for 8’’ or 10’’ instead of 13’’, also with respect to future upgrades using another subwoofer to compensate for room geometry. Fortunately, i managed to get a new SPL-X 10 yesterday, after listening to it in the Audio Reference HQ in Hamburg, Germany.
Now it is time to quickly introduce this amazing company: Audio Reference. They bought Velodyne Audio, which used to be an american company until about 2017, before being sold to Audio Reference, a Hamburg based family business, which in turn is ‘reference’ in terms of absolute high end audio equipment like d’agostino and stuff like that. And they sort of saved Velodyne Audio from being another company being sold while loosing its identity; it is still called Velodyne Audio and they are engineering and developing the devices in Germany. That is cool huh?
So i was really lucky to get one directly from the manufacturer, mostly they are already sold out due to the high quality of their products and the low batches provided. So the americans prefer the 18’’ models, for me, a 10’’ is more that enough (actually 2, but now it have 1). They are not the cheapest of all subwoofers, but oh boy, i could not have invested my money in a better way!
If you look at the pure numbers or measurements, one may not even find so much of a difference, with the exception of room modes being triggered slightly less, at least the non-super deep ones. And a slightly lower group delay at around 40hz, and then of course further down below, which seems to be correlated. It does not have the weight of the SVS in terms of total energy somehow produced. But there are 2 very fundamental improvements:
- As mentioned by audiophiles in reviews, the subwoofer basically blends in perfectly with the monitors. It becomes invisible, and that is perfect.
- Somewhat related to that, this also means that the space control and definition is improved substantially, in a way that i did not expect that upfront. I knew as soon as i listened to it in the test studio and i also read that in some reviews of the DD10+. But experiencing it yourself, and how much more brilliant and smooth the system sounds overall, how much the staging and topology of instruments is improved is really impressive.
Now the old cable discussion kicks in again — after time aligning the new subwoofer using the invers-null-method, which means inverting the correct phase and see where the deletion around the XO frequency is adding up mostly without broadening, i realized at wide, incredibly dense and precise stage. And then i thought: what it i reinstantiated the Mogami 2534s again? Would that not be more alive and ‘the right amount of dynamics’ to be reintroduced into the system? Now i did that, and the result is that i wish for a cable with both properties .. The stage becomes more deep and the signature more alive, musical and dynamic, but i loose some of the precision in terms of stage width in contrast to the Supra EFF ISL.
This is probably something to optimize for. The current working hypothesis is something like the Sommer Epilogue II, which seems somehow like an interesting compromise. Let’s see. Here is an image of the new Velodyne SPL-X 10’’:
If i would have to put a number on the improvements made, it is easily by 100%. It is hard to quantify: how would you evaluate ‘realistic space bottom instantiation’? It is somehow absolutely fundamental, and yet an acoustically subtle aspect.
If you reason buying new speakers, think about buying this subwoofer first. I would be not surprised if you see that the improvements are more fundamental than lets say exchanging a Dynaudio for a ADAM speaker. Even though now find it very tempting to know how EFF ISL cables would sound with an ART tweeter 😀