I think it's a mistake to characterize digital xovers common discussion theme as band-aides for poorly implemented designs,
or being just a bunch of extra features, bells, whistles.
Quite to the contrary really.
Imo, a digital xover, a good digital dsp, is simply a tool for maximizing the sonic potential of even the most excellent acoustic designs.
I believe I can take any speaker with a passive xover and EQ implementation, and replicate it very closely, perhaps exactly, with multi-channel dsp.
And then improve it, most likely fairly substantially.
I think this is true for even such heralded acoustic designs with passive xovers, like Danley's SH-50.
The dsp tool set is just too deep...
- timing to the nearest sample 0.02ms (48kHz)
- levels to 0.1dB or better
- nearly unlimited minimum phase filters for in-band and out-of-band corrections
- extensive assortments of conventional xover types and orders
- linear phase xovers
Those tools aren't band-aids...they are tuning components in any well laid out acoustic design imho...
My point is that additional EQ possibilities can help the sound quality, making the comparison further in favor of DSP solutions.
Though I agree that the overall speaker design should be sound, and the amount of digital manipulations must be limited.
I think that part of negative experiences with DSP could come from bad implementations. For example, lack of internal precision, extra AD/DA conversions, low sample rate, jitter etc. And a too flat frequency response may not be preferred as well.
One of the most memorable systems I heard last year was the Steinway model B at Xfi. Lyngdorf digital amps with room correction, open baffle speakers with digital crossovers and filtering. Great great sound, though I had too little time to hear it properly regrettably...
Fedde
Scott L, I understand, and I empathize with your point of view. While I am certain that everything in an acoustic event can, in theory, be measured, I do not think we yet know exactly what to measure, and which of those things matter and which ones do not. We do not know enough about the ear-to-brain interface to fully understand all the aspects of sound which make a sound reproduction truly convincing. We are getting better every year.
Why does my equipment sound quality improve as it warms up... from turn-on to 20 minutes, the quality improves pretty dramatically... Why? what measurable parameter has changed? I doubt anyone knows, but something changes as the equipment warms up. Someday we may be able to point to some form of newly identified distortion as the culprit, but for now we simply accept it, and give our equipment time to warm up before serious listening...
The experiment you propose could be done. A high quality passive speaker could be made into an active one where the DSP crossover mimics the passive crossover. I tend to agree with Mark100 when he says:
A year ago I was very skeptical of Class D amplification. I had heard several mediocre examples, and I just thought that the whole Class D thing was "too good to be true". I was also very skeptical that an analog signal could be converted to digital, processed, and converted back to analog using an inexpensive "mini-DSP" and there would be no resulting degradation in sound quality.
Two things changed my mind... first I heard the NAD M22 class D amp, and I was pretty amazed. This thing is serious high end and to my ear it competes with the best amps out there. This is an N-core technology amp.... Second thing to change my mind: The Dutch&Dutch 8c speaker. This 3 way active speaker turned all my pre-concieved notions upside down. it uses DSP filtering and Class D amplification. Whatever quality is lost by the A/D and D/A process is made up for in the other advantages that come with an active speaker.
No - you cant measure a feeling - but we have good guidelines to know when we fool ourselves and when it might be true what we hear.
But you can measure a loudspeaker. And the point is not to fix a broken speaker with a DSP.... it is to make a better speaker with a DSP - more precise and more balanced.
No doubt, that when you do not build a good system by the basic rules of nature - you will not get a nice result - no matter the technology.
If you can get the best result out of your drivers, with an analog filter - for your taste. Then there is no need for digital filters - and no one can convince you to think or believe differently - and why should they? It's not a battle between technologies. But in many cases, digital filters are smarter and more efficient. So we have to look at the specific case and judge whether it is a good plan to do this or that. Cause there is not a perfect answer or one fits all.
The idea to build a perfect and neutral loudspeaker, is good, cause it leaves the "art" to the source - not the loudspeaker. And when the digital technology is thought into the design from the beginning - then a loudspeaker can be very good - potentially better than many analog designs.
Can you please list a few well known and generally highly praised active speakers with analog cross-over or passive cross-over that D&D 8c surpassed, and can you identify in what areas from the subjective point of view (not for example that 8c measured flat in comparison)?
Digital EQ also can improve speakers with analog crossover. One of my speakers with analog crossover is digitally EQed, and I can hear substantial improvement with digital EQ, because the room is not truly optimised even while the speaker is reasonably flat.
It's a respectable challenge to do everything in analog, but in reality, it's very hard to find an audiophile who applies high quality analog EQ to his speakers. Something like a Prism, Sontec, or Manley would be a serious contender for an audiophile EQ which is comparable to digital ones, but they are 50+ times more expensive.
Last edited:
I have never been able to do a head-to-head comparison of really great speakers. I hear a pair of really special loudspeakers, and then a couple of years later, I hear another pair. Was the second better than the first? I don't know. But I remember which speakers really impressed me. What I really look for in speakers is this: If I close my eyes, how easy is it to forget that I am in a room listening to equipment, and "suspend disbelief" and imagine myself listening to a live performance. If this transformation comes easilly, I consider those speakers to be quite special.
The Revel Performa F228 BE impressed me quite a bit when I heard it a couple of years ago. Five years ago I listened to a pair of large Martin Logan's, which were more than 20 years old when I heard them... they impressed me. Thiel CS3.6 captivated me 10 years ago. Before that Proac Response 2. 30 years ago I was quite blown away by the B&W 801 Matrix, as well as the Klipsch K-horns (which I heard at the same shop as the 801's).
A friend has a pair of Magnepan MG-III, and I have always been impressed by those...
All I can say about the Dutch&Dutch 8C is that it impressed me in the same way as the others i mentioned above. I have always wanted to hear a pair of Wilsons, but I never had the chance. Same with Magico. I have not listened to the Kii. I would like to hear the Barefoot MiniMain12 someday.
Even if one were to do this, what makes you think any subjective findings are due to analog/digital or active va passive? All your comparing is two different speakers(system)s comprising of many parts and design criteria.
What would that get you? Unless they were all in the exact same room what would you win from such a subjective comparison.
Both the Kii and the D&D try to work with the room they are in. Those are some of the strong points they have going for them. But put a speaker with worse off axis performance in a space that doesn't highlight that negative, could it still sound beautiful? Of course it can.
I do not know. Are the screw holes for the WG in the datasheets? SEAS isn't very helpfull with providing additional mechanical data in my experience. The datasheet is the data you get and the rest you have to measure yourself.
You shouldn't see the DSP as a magic box that solves your problems. It just doesn't. The loudspeaker with DSP have the same acoustic problems as passive loudspeakers.
There's a whole world of crappy passive speakers and there's also a whole world of crappy DSP speakers.
The DSP just provides a different toolset with different advantages and dissadvangages. As someone else already mentioned, it's not a battle of technologies. When designing a speaker you have certain problems, you use certain tools to solve those prolems. Based on requirements (performance/cost/time/flexibility/aestethics you name it) you pick solution A or solution B. DSP isn't always better, analog isn't always better. They are toolsets to solve certain problems.
Hi Mark. Nice to hear from you. My follow up would be a question: What does the process of A-to-D, and then after the processing, D-to-A, sound like ?
Yes, of course, I do realize that if one starts with a digital source, then, the first conversion process has already been handled earlier in the chain of events; but some of us still use pure analogue music sources.
My Magnums seem to have the mounting holes further out... and it's a 25mm - and the DXT looks like a 26-27mm.
Could 3D scan the DXT and remould it to fit the Magnum..... scanners and printers are getting cheaper by the day
My Magnums seem to have the mounting holes further out... and it's a 25mm - and the DXT looks like a 26-27mm.
Could 3D scan the DXT and remould it to fit the Magnum..... scanners and printers are getting cheaper by the day
If you are going to make your own. Simulate it in axidriver (or similar software) first.
Gonna play with the DXT as it is, to begin with.... I have much to learn before I try and design a waveguide
I have a set of Scanspeak R2904/700005 lying around.... maybe it could be fun to mount them in a WG300 - they're only 10 Euro a piece... the waveguides that is
Earlier I did a level matched blind-test with analogue and digital input to my DSP controlled active speakers. One thing was needed is to select between two presets. The analogue source was a high quality DAC. The source was high quality blu-ray player in both cases.
In short, I was not able to detect any differences and struggled to decide by ear which input was the active. So the result was a random guessing.
The test was in my home in calm conditions and the preset changer was my brother who is not so much interested in audio reproduction, and he was laughed at me quietly with my results.
Question was not directed at me, but I can answer it, as I heard it at the same time in the same system, analog signal vs. the same signal digitized than converted back to analog, send to the same pre/amp/speakers.
The room had very serious treatment installed through proper process involving measurements, with traps and reflectors on all walls and ceiling. Speakers were Revel Salon 2.
The unit doing AD was well respected among professionals, I do not recall model, but the price was in range of $3000 - 4000 to give you an idea, so I am certain it was not worse than AD used in DSP speakers being discussed in this thread.
I heard significant loss of transparency, loss of details, openness and dynamics, so much so that I would never accept such detrimental impact on signal to gain controlled directivity and more on and off-axis flat response that DSP would give. And this was only with one additional AD step, I am wondering how massacred signal gets when processed through DSP. That is why I started this thread, but so far it does not seem anyone gained experience that would allow objective subjective assessment for the lack of better words.
Last edited:
The test, although not much more detail, but I don't feel the need for any further describing, the result was so obvious.
https://www.diyaudio.com/forums/class-d/306815-hypex-fusion-plate-amps-151.html#post5908644
https://www.diyaudio.com/forums/class-d/306815-hypex-fusion-plate-amps-151.html#post5908644
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.