Thread on DSP vs passive crossover, Linkwitz approach:
Alan Huth This is my third attempt to post. Sorry if it shows up multiple times. I assume the earlier versions were removed because I put a link in it. That link is to Linkwitz Labs. Home of the late audio legend Siegfried Linkwitz. Easy to find. His section on the Magic of Stereo is a good place to start. The purpose of this post is to point out the Linkwitz disagreed with your assertions about large drivers and DSP. He preferred smallest drivers possible for the target frequency range. And he preferred DSP to passive crossovers due to issues with unstable impedance and polarity. His X521.4 speakers were named, by some at least, as best of show at Axpona 2022. And they were designed over a decade ago, so he must have been onto something.
Snake Oil Audio so you want to say that there is a reason why Linkwitz never designed a legendary speaker? ;-) Yes there are many different opinions about HiFi no doubt and even people who design stuff that I don't like sometimes have good reasons why they build it the way they do.
Alan Huth I think it’s fair to say he is a legend, and arguable to say that some of his speakers are legendary. It might be worth while understanding why he came to the conclusions he did. Here he is on video: https://youtu.be/VC-sxvNzC8I
Rick G It’s a matter of taste and perspective.. I for one have no problem wifh DSP, but only for 80 Hz and below ;).. The late Linkwitz’s designs are well engineered and are very linear and coherent, but some people are sensitive to the sound of opamps in most Active XOs (with DSP or not), but some are not…
Snake Oil Audio Hey Alan, I watched the whole thing and Linkwitz does not disagree with Janos and me. He literally said that small drivers with large extensions are a bad idea and don't work well. The reason to pick a small driver if you want to build a dipole and you want to avoid beaming. Janos live cabinet is technically a dipole but there is no beaming at all since the VOL works completely differently and I am not using dipoles and never will so I can't see Linkwitz disagreeing with us. When it comes to room correction I am a fan of fully discreet equalizers like i.e. the Cello Audio Palette (shown in the video) but I don't like DSPs because of their integrated circuits.
Real World Audio Hi Alan, thank you for the Linkwitz talk recommendation! I have watched it (now I'm up halfway into part 6). I'm quoting Mr Linkwitz: "I take a very skeptical view of electronic equalization". He is skeptical of DSP, not passives... he has same exact view on the subject as I do. ;). Now as I'm 1hr into his lecture, I agree with him / he agrees with me on all points that he has touched so far. Quoting him verbatim on room EQ products: "they will all change what you hear" ... "we are not smart enough (for this technology) yet". In this talk so far (1hr mark) he has taken a firm stance that DSP is NOT fully developed yet, and that we know very little on how the brain works to have a definitive DSP solution to fix problems. He might have other videos on other aspects of DSP, yet in this lecture, up to the 1hr mark he has clearly made strong pointers against it. So, later on, he might have good things about it as well (there's no technology with ALL + or - ... every single one of them have compromises), and when he talks about DSP benefits (which I have not heard so far in his talk), that will not cancel the flaws of the technology that he has so eloquently described.
Real World Audio @Snake Oil Audio I agree, the ICs are a serious bottleneck to DSP. When they fail to provide a perfect solution in a preamplifier or power amplifier, they will have the same performance point in a DSP duty as well. It's not the technology, it's the quality of the components that it uses. Taking mediocre parts quality will not give heavenly results just because we are trying out a new recipe... it can drastically alter the end result, but does not support retaining all the source information, and will inject serious colorations of its own.
Alan Huth @Real World Audio Hi Janos. I guess it's a matter of which devil you sleep with. This is a quote from his webpage on crossovers:
"Crossovers may be implemented either as passive RLC networks, as active filters with operational amplifier circuits or with DSP engines and software. The only excuse for passive crossovers is their low cost. Their behavior changes with the signal level dependent dynamics of the drivers. They block the power amplifier from taking maximum control over the voice coil motion. They are a waste of time, if accuracy of reproduction is the goal."
And, his two legacy systems that people build today, the LXMini and the LX521.4 both use DSP, but they are available with active crossovers ($$$). This tone/naturalness/simplicity objective trades off, to some degree or other, with actual voice coil control. I would love for you to dig into that sometime. You know the guy has a point, if not in all circumstances.
Alan Huth I'm really going out on a limb here (sorry Mr. Linkwitz, wherever you are). His design goal for his magnum opus LX521 was to create a speaker system where the room reflections were an almost exact copy of the direct sound: same frequencies, just quieter and delayed by at least 5ms. To do that he made dipoles with the smallest driver diameters possible, to maximize polar qualities. Since excursion was a problem with small diameter speakers, he went with a 4-way design to cover all frequencies without distortion. Now, requiring a 4-way, he decided that passive crossovers introduce too many alterations in the signal due to uneven attenuation and frequency-dependent changes in amplitude and polarity. Hence, the evil DSP. Costs outweighed by benefits in this situation. I hope I did him justice with that summary. Someone please correct anything I got wrong.
Real World Audio @Alan Huth Interesting... as in the talk he did not speak kindly of DSP either. (Maybe later on he does, but that will not make his initial negative statements on DSP magically disappear.) After all, no technology has only bright side.... it is exclusively up to the implementation which one will be subjectively preferred to a listener. Either DSP or passive can be done in a very good, or a deplorable fashion. ; 👍✨
Real World Audio @Alan Huth Hello Alan, thank you for the superb summary, you did very well! Indeed, for a 4 way system DSP makes very good sense. If I were to design a 4 way I would use DSP myself. However, the 4 way sound is not at all my cup of tea, it takes one on a very different path. Because of personal preferences, nothing else. (Technical justification can be packed to BOTH sides - so no winner based on technological justification.) My subjective impression of seriously multi-way VS as simple as it gets should not affect anyone though. Simplicity = soul. Multi way = show. The road is different to each, and that's the choice it boils down to. We can call either as the best, or as a waste of time, the bottom line is we have to make a choice to get the technology that suits our needs. I hope that clears waters a little, Cheers; Janos
A link to the Linkwitzlab website, which continues the legacy of the late Mr Linkwitz:
On Room DSP Pro and Con:
Optimize A Janos subscriber here ;)💕 But if we have your system Janos. All that negative aspects Janos talk about is not in his system! In that system we insert a DSP. So not any downside in the electronics that Janos talk about in this video. Then why insert a DSP into Janos system.. Why go from analog to digital then apply adjustments and then back to analog... That drawback is far smaller than the benefits! When acusticians say 30% to 50% is YOUR room acustics that you hear. That is a much higher and bigger problem than if we do ADC and DAC conversation. (The room is the BIGGEST COMPONENT, more then in one way.)😉 You use the DSP to correct your ROOM! (Those 30%-50% is a GIANT proportion of what you hear). Of course you fix in the physical world with room treatment as much as possible first. Then implement the DSP and correct and prolonging it into DSP and fix your room and the bad idea to put a altec in a box and putting that box in another box that we call a room. And we listening to reflections from the floor, ceiling and all walls that end up into our ears plus nulls and peaks from our room dimensions. So taking the help of DSP to correct the room that is the majority of what you experience is a missed opertunity. Even in Janos system. It is a shame in my opinion that people don't see the room correction capabilities that DSP bring to the table. (Another trick DSP enable you to get a personal tailored to YOUR individual taste! What more do you need.. and so much more.) Then there is many other benefits that we can do and is offered with DSP and is topics for another days. Great interview! 💕
Real World Audio Thank you Optimize, superb discussion! I see the good that DSP can achieve, please do not get me wrong. However there are shortcomings to it that have stayed my hand. These shortcomings are: 1. Implementation. DSP requires an implementation, and its parts quality is going to have an impact on sound. Opamps and ICs affect every device they are used in, even DSP. As Rick said some people can't hear their negative impact, while others can. I have not experienced a DSP yet that did not make the sound less mechanical than with DSP off. 2. Rooms ultimately cannot be fixed. Rooms produce reflected/delayed sound. While the total energy level of speaker output + room is a net sum, our ears experience them as initial wave versus reflection. Altering the speakers output to match a total energy level seemingly fixes the room, - however it doesnot fix the room, just the total energy. At the price of substantially altering the energy content of the initial wave coming from the speaker. This gives very different subliminal cues to the brain on how to process sound. So while the total energy content becomes similar to an optimal room situation, but the tonality/perception is altered. With all that being said, I have not experienced a DSP that I preferred versus it being taken away from the system. So, my choices boil down to empirical not philosophical aspects... I am not against the technology itself. Just do not like what it does to the music. (Makes me focus on the sound instead of the music - that's how I hear it.) However, I'm quite sure if I strated playing around with it in my system, I could take it further. However, the limit is time and effort. It is not an area where I am expert, others playing around with it who are more experienced with DSP will contribute much more than I could.... Cheers, Janos