By the way, I must confess I considered buying a pair of Kii speakers after auditioning them. But I ended up too scared losing my DIY audio hobby...
...so now I am on a (long and windy) road trying to approach somewhat such quality with DIY...
I'm with you
Cant remember if I posted this link - but here goes again:
Heissmann Acoustics | Kits | Speaker developments
May I kindly ask what the Heismann Monitor has to to do with the Kii, other than that it is a nice design with the DXT tweeter? Anyone who is a bit versed in making Arta, REW, Clio etc. .frd and .zma measurements and simulators such as VCad can do the Heismann thing. A W300 waveguide with an XT25 family ringradiator will also nicely do the trick. The Kii on the other hand is very different, in that it is digitally delayed and tailored side- and backwoofer design, mimicking a cardiod digitally.
I think it's rather the other way around: the spec and design conscious people are often skeptical of anything written down by a manufacturer, while others seem to gobble the stuff up like bees fly to honey. I and others have pointed this out before, and next thing you know Scott L. makes a post titled:
Sure they did... They could not be bothered to make a decent (and expensive) 150Hz passive crossover, so they implemented an active one. This is probably a valid reason. It's even literally on their own website. And the reason for analog was already pointed out by Boden. And even this is probably a valid reason if you look at who buys this kind of stuff. Nothing has anything to do with analog sounding better than digital.
Scott L, don't get me wrong here. I'm not here to criticize your setup. I see many valid reason to keep a primary analog system analog. And as many have pointed out, there is no reason why only this fact would make is sound better or worse than any digital setup.Conceptually it really just makes a lot of sense. What I am critical of however, is the line of reasoning: Look here is speaker xyz, costing $shitload.. Must be a reason they chose Analogue-Active. That's not really an argument is it?
What actually bothers me a tiny bit about many of the digital active stuff, is the fact that when you have an analog source, it's often the ADC that is the most inferior part of the kit. It often has noise and THD figures that are quite a bit worse than the DAC's. You can get a nice 8 channel ESS DAC for not to much money, but the equivalent ADC.. it's quite rare.
Actually, for a ADC-DSP-DAC setup it makes sense to have a DAC with higher SNR than the ADC. Reason being, in the DSP you need to attenuate the signal with commonly 6-12dB in order to avoid clipping in the filter sections, both for IIR and FIR filters. So if the DAC had the same SNR as the ADC, the DAC would be the limiting factor of overall system SNR.
That said, it is not too hard to make an op-amp based xover with 130dB dynamic range. To get the same dynamic range in an ADC-DSP-DAC based crossover, you need the highest-SNR converters available, and that will not be cheap.
That is an interesting question. Here's a thought experiment:
Let's say you're making a studio monitor. So let's say the max level is set to 110dB SPL at 1m. And a good quality monitor should not have objectionable hiss when placing your ear close to the speaker. So let's say 20dB SPL at 0.1m. That gives a DR of 110-20+20=110dB. Doesn't sound too difficult. But to achieve a system with that DR in practice, the ADC, DAC and amp should each be spec'ed at 120dB DR imo.
And from this it's obvious that the commonly used ADAU1701 with it's builtin converters (min 95dB DR) is not sufficient for this kind of application.
Well.... yeah....
I like Heissmann's approach - cause he makes it neat and well described.If you take the DXT-Mon and add a - lets say - a 3 way fusion amp. Then you could do some of the same stuff as the Kii3. Maybe not exactly the same - but the same tweeter and the same overall design. And with a 3 way amp+DSP, you could also make some fiddleing - and do some cardioid-like stuff. I think Fusion amps, can be linked - so that many funny and interesting things can happen. Overall... still cheaper and more fun than just buying the Kii3
Last edited:
Buy some good stuff
I have this one:
Ground Sound
And it is absolutely transparent. Way more detail than many other designs I have heard. No "digital" sound or lack of dynamics or detail.
I have heard minidsp and fusion amps too.... works great. It's still the overall design that sets the limit - in my - and many others view. Even the biggest DEQX, uses the same DSP as the aged and cheap Behringer DEQ2496. Nothing overly fanzy or spectacular in these boxes. The first miniDSP's had and still sometimes have basic noise issues. But my groundsound is very quiet - even with hornsystems. To my knowledge - my groundsound DCN28 is also the only product ever, to have individual digtaly controlled analog volume for each of the 8 outputs.
And..... transparency..... what is it you want to hear?? The lack of good old workmanships in the mixing studio?? Distortion, channel crosstalk and high background noise from the mixing or recording equipment??
I can clearly hear all the faults in any recording and why would I want any more??
My experience is that most analog designs clearly hides and covers most of the problems and issues in the long path from the microphone to the listener in the living room at home.
People want turntables, valve amps and big sloppy and badly crossed drivers with a soft "natural" added distortion - that covers it all and makes it into a Citroen DS with hydropneumatic suspension - where you forget the road and just float over obstacles. But that has nothing to do with transparency - absolutely nothing
Yes it is. The night I returned from the auditioning I already made some EQ improvements to come closer in sound balance which improved things. And later on, I gradually made multiple improvements. Better bass system, better tweeter, active driving...
Two years later things sound much better. Didn't hear Kii3 since, so no idea how big the delta is nowadays...
Probably still significant, but I had a lot of fun!
Sure, live music should also be a reference. Still, a comparison of specific recordings on other speakers give an indication what is possible. Also, the trick is to find a balance in sound that makes all recordings sound good or at least listenable. As my main goal stays enjoying good music, not only for a few reference recordings...
I'd guess the signal is split into these frequency bands:
The crossover frequencies between these bands are a compromise between harmonic distortion (crossing higher results in less distortion) and dispersion pattern control (the dispersion of the midrange must match that of the waveguide tweeter; the cardioid of band 3 is composed of two monopole 'point' sources and therefore only works up to some frequency; frequency band 4 has no pattern control at all).
- High: 150? .. 90 deg conical: waveguide tweeter.
- High midrange: cardioid: midrange and side woofers. These drivers together act as a virtual driver with a cardioid dispersion pattern. In order to obtain the cardioid pattern, one copy of the signal is equalized and fed to the midrange, while another copy is equalized, delayed and inverted and fed to the side woofers.
- Low midrange: cardioid: side woofers and rear woofers. Similar story.
- Low: monopole: side woofers and rear woofers.
The crossover frequencies between these bands are a compromise between harmonic distortion (crossing higher results in less distortion) and dispersion pattern control (the dispersion of the midrange must match that of the waveguide tweeter; the cardioid of band 3 is composed of two monopole 'point' sources and therefore only works up to some frequency; frequency band 4 has no pattern control at all).
Last edited:
Its not the signal as such that makes the system cardiod (or not), it is the inter-driver delay and related inter-driver cancellation that creates the cardiod character over a certain frequency band to my best of knowledge. But I might be mistaken here.
So let us follow your suggestion: how could that be achieved practically? In other words: how should the drivers be fed?
So let us follow your suggestion: how could that be achieved practically? In other words: how should the drivers be fed?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.