Is there a sticky or something listing the most popular DSP gear? I’ve found a few PA amps with simple crossover, and pro crossovers with eq costing tons with AES/EBU etc, and of course miniDSP stuff but what else is there really for the regular consumer? I’d like to try but don’t want to spend a ton. I have two identical integrated amps which could be used for two way active. I have a NUC as source, it has mini jack analog, USB or SPDIF optical out. What to put between? Oh and please no gear with external PSU. I can’t have more of those. Give me something standard 45cm width to fit my rotel amps.
Used pro gear is what i recommend.
The list of processors Fulcrum Acoustics develops presets for all their speakers makes for a good selection base. Processor Configurations | Fulcrum Acoustic
They will all be standard rack mount.
In new gear, Motu and other similar studio processor brands gather good opinions.
Channel I/O count, and whether you want FIR capability or not, are probably the starting considerations.
I use a NUC as source, and REALLY like a QSC Core110f processor connected via USB.
It's an open architecture processor, which at first may seem daunting, but compared to passive xovers it's a walk in the park 😉
Good prosound dsp is an eye and ear opener, imho.
The list of processors Fulcrum Acoustics develops presets for all their speakers makes for a good selection base. Processor Configurations | Fulcrum Acoustic
They will all be standard rack mount.
In new gear, Motu and other similar studio processor brands gather good opinions.
Channel I/O count, and whether you want FIR capability or not, are probably the starting considerations.
I use a NUC as source, and REALLY like a QSC Core110f processor connected via USB.
It's an open architecture processor, which at first may seem daunting, but compared to passive xovers it's a walk in the park 😉
Good prosound dsp is an eye and ear opener, imho.
The thing about amplifiers is that one size does not fit all frequency bands. Assuming by passive crossover you mean at the speakers, you have to pick an amplifier with enough power for bass and low enough distortion for the high end. There is also amplifier-driver coupling issues that a passive crossover interferes with. Exposing your tweeters to the power levels needed for the bass is a big risk. Bi-amping or tri-amping means you can safely and cleanly operate at significantly higher levels. The clean power per frequency is drastically improved by separating frequencies into separate amplifiers. An amplifier that clips at 100W at one frequency, clips at 50W total with two frequencies, and clips at 10W total with ten frequencies.
And passive crossovers are limited to analog methods. This is not different from analog electronic crossovers but digital signal processes can use memory that is impossible in analog circuits. This means that certain phase response problems can be handled much better.
And passive crossovers are limited to analog methods. This is not different from analog electronic crossovers but digital signal processes can use memory that is impossible in analog circuits. This means that certain phase response problems can be handled much better.
I think the general problem with dsp for many is that it's to complex to use in daily life and that the sound quality of the ad and da are not up to par with what you can do with analog, even with a digital source (when using a good dac).
For diy nerds, it may be easy to setup up, but not for most users, and even a lot of diy nerds like me don't want to fuss about technical stuff each time we listen to music or move the system. I follow what happens in dsp land with great interest, but the main issues remain, to complex and to expensive for real high quality.
That is something dsp builders need to adress to make dsp mainstream (which it isn't att all yet for hifi). For pro sound or studio monitors it is a no brainer for me, but not for hifi. Passive is so much easier to do, and in general cheaper than high quality dsp for low power systems. And once done right, you don't have to think about it anymore
So DSP builders, this are the points to work on:
- top levle ad/da convertors for reasonable prices (It can be done, look at toppings DAC's that are top level and cheap)
- intuitive setup and adjustment, also fit for less technical people, preferable with a good working audo-calibrate (with mic and clear instructions) system that at least get near where it has to be.
- and the most important: make it a "set and forget"-proof system. That is a very important thing for most music lovers who are not so techincal minded (+95% of the people).
And high quality amps in any style should be cheaper also. For class D that is the case, but many still don't like class D for it's typical sound (which came a long way, i know). I have class D amps, which are reasonable good, but always loose to my class AB amps (tube and transistor) on sound when i compare.
Minidsp is close on that last one, just like some of the very expensive pro audio systems, but still hard to calibrate. And the prices are still to high for those high quality things (which minidsp is not on ad/da convertors, the dsp itself is good).
But on the meantime, keep experimenting with it down here, it's interesting to read and study. And after all each should use what fit his needs, be it dsp, analog active or analog passive.
For diy nerds, it may be easy to setup up, but not for most users, and even a lot of diy nerds like me don't want to fuss about technical stuff each time we listen to music or move the system. I follow what happens in dsp land with great interest, but the main issues remain, to complex and to expensive for real high quality.
That is something dsp builders need to adress to make dsp mainstream (which it isn't att all yet for hifi). For pro sound or studio monitors it is a no brainer for me, but not for hifi. Passive is so much easier to do, and in general cheaper than high quality dsp for low power systems. And once done right, you don't have to think about it anymore
So DSP builders, this are the points to work on:
- top levle ad/da convertors for reasonable prices (It can be done, look at toppings DAC's that are top level and cheap)
- intuitive setup and adjustment, also fit for less technical people, preferable with a good working audo-calibrate (with mic and clear instructions) system that at least get near where it has to be.
- and the most important: make it a "set and forget"-proof system. That is a very important thing for most music lovers who are not so techincal minded (+95% of the people).
And high quality amps in any style should be cheaper also. For class D that is the case, but many still don't like class D for it's typical sound (which came a long way, i know). I have class D amps, which are reasonable good, but always loose to my class AB amps (tube and transistor) on sound when i compare.
Minidsp is close on that last one, just like some of the very expensive pro audio systems, but still hard to calibrate. And the prices are still to high for those high quality things (which minidsp is not on ad/da convertors, the dsp itself is good).
But on the meantime, keep experimenting with it down here, it's interesting to read and study. And after all each should use what fit his needs, be it dsp, analog active or analog passive.
Well, I agree to a good extent, say 90%. There are people with varying degrees of spirit, skills and technical abilities (system design, programming etc.), and most people tend to know where they stand when compared to the professionals in each field. For example, a parametric equaliser is scarier to many, when compared to the intuitive graphic equalisers, isnt it? Even crossovers, I believe, are more or less similar.
I think it's best not to use A/D conversion if the source is analogue. However, if the source is digital (mostly the case these days), a D/A conversion is inevitable and any related processing (EQ, crossover, limiting etc.) maybe easily carried out before the DAC and power stages.
On the whole, the size and seriousness of a system also decides the crossover methodology. Digital crossing and time-alignment are fully adjustable and that flexibility matters a lot in experiments and professional applications, unless the target application involves low power like 200W etc. Besides, a lot of studio, cinema applications (like atmos etc.) often combine digital sources (satellite link, hard-drive etc.) with multi-way and multi-channel, leaving digital processing the only practical way to go with. In such cases, the digital methods may even result in reduced overall cost when compared to the analogue/passive method. The digital methods are also less prone to issues caused by ageing, drift and component tolerances.
Last edited:
With a passive you just have the cabinet, drivers and the passive crossover that make the loudspeaker.
With an active you have the cabinet, drivers, AD-DA converters, DSP and amplifiers that make the loudspeaker.
Easy interchangeable of amps and DACs: the fact that I can plug and play whatever single stereo DAC and amplifier with a passive is nice.
Great during listening sessions when friends bring their latest DIY amplifier or DAC.
With 3-way active you need three stereo DACs and amplifiers and they are not so easy interchangeable because they are part of the active setup.
The amplifiers that I like are Pass based class-A, I can handle the 150w dissipation of one class-A, but not from two or three class-A amplifiers.
One thing that I really like about active is the easy tuning of the bass level:
just change the spl-output of the bass amplifier a little.
Especially handy since there is so much difference in the amount of bass in different recordings.
With an active you have the cabinet, drivers, AD-DA converters, DSP and amplifiers that make the loudspeaker.
Easy interchangeable of amps and DACs: the fact that I can plug and play whatever single stereo DAC and amplifier with a passive is nice.
Great during listening sessions when friends bring their latest DIY amplifier or DAC.
With 3-way active you need three stereo DACs and amplifiers and they are not so easy interchangeable because they are part of the active setup.
The amplifiers that I like are Pass based class-A, I can handle the 150w dissipation of one class-A, but not from two or three class-A amplifiers.
One thing that I really like about active is the easy tuning of the bass level:
just change the spl-output of the bass amplifier a little.
Especially handy since there is so much difference in the amount of bass in different recordings.
Regarding complexity: It is possible to build a 3-way active system that does not involve multiple cables, lots of boxes, lots of complexity. But it takes planning and thoughtful design.
As I said at the very beginning of this thread, I am very enthusiastic about the Hypex Fusion amps. They are fully self contained plate amps with built-in DSP filters and N-core amps. Once set up, they are "fire and forget", with auto on/off capability. The only cables running to the speaker is a signal cable (RCA, XLR, or digital s/pdif) and a power cable. All other wiring is internal to the speaker cabinet.
If the Hypex product had not been available when I did my project, I would have been more hesitant to go active. But with Hypex, it was a very easy design integration.
The N-core amps have excellent sound quality. The built-in A/D and D/A processors use highly regarded microprocessors and software.
Designing the DSP filters for a 3-way system is no more difficult than designing a passive network for a 3-way... it is probably a little less complicated. However, Hypex may have set unreasonably high expectations by claiming the process is easier than it actually is. It is not a process for inexperienced newbe's to quickly and simply achieve high performance results. But an experienced designer CAN achieve a high performance result rather quickly.
j.
As I said at the very beginning of this thread, I am very enthusiastic about the Hypex Fusion amps. They are fully self contained plate amps with built-in DSP filters and N-core amps. Once set up, they are "fire and forget", with auto on/off capability. The only cables running to the speaker is a signal cable (RCA, XLR, or digital s/pdif) and a power cable. All other wiring is internal to the speaker cabinet.
If the Hypex product had not been available when I did my project, I would have been more hesitant to go active. But with Hypex, it was a very easy design integration.
The N-core amps have excellent sound quality. The built-in A/D and D/A processors use highly regarded microprocessors and software.
Designing the DSP filters for a 3-way system is no more difficult than designing a passive network for a 3-way... it is probably a little less complicated. However, Hypex may have set unreasonably high expectations by claiming the process is easier than it actually is. It is not a process for inexperienced newbe's to quickly and simply achieve high performance results. But an experienced designer CAN achieve a high performance result rather quickly.
j.
Offtopic, but would you care to expand upon the audible effects of the bypass caps which look like Soviet era silver micas?
That was a passive/active joke.
I can not know as I have not listen to your system which I surmise to be much more better than mine 🙂
There are some things you can do with passive components that you cannot do with active; such a zobel network to flatten impedance, notch filters, shunts, etc... These are things that affect the impedance and electro-mechanical response and damping of the driver, and can be combined with active or passive networks to get the result you want.
I forgot that I later changed to Russian MBGP military surplus which I found to be considerably better than the Clarity and Mundorf caps.. There used to be a lot of them on eBay: Hermetic Military PiO Capacitor 10uF 630V Russian USSR MBGP МБГП Paper Oil ~600v | eBay as an example - I found these to be quite good and much less expensive than some considerably inferior sounding metallized film. I've attached a very early picture of that XO, the sole plastic capacitor in the picture is the woofer zobel. A slightly improved version of this XO lives on in another friend's system. One of the big mistakes I made here and subsequently partially remedied was putting several of the coils too close together - I fixed later by re-orienting a couple of them. No evidence of coupling - not really surprising given the very low power involved.
The line level crossovers are 3 way 24dB/octave LR type of my own design. The capacitors in those are REL RTE polystyrene foil, 2% tolerance and selected where possible to 1%, and matched in the same location on the other channel. I kind of like SMD so the resistors are Panasonic 0.1% 1206 thin films, op-amp based, and there I selected OPA1641.. I/O is all balanced.
There is a lot of tube hardware in the system which demands periodic maintenance. I waited too long, having completed some repairs and whole tube replacements in the amplifiers.
It's a bit of a challenge when sometimes all you want to do is listen to music. I have a lot of sources as well - some like the TD-124s need attention soon. The Studer A810 is the only thing really at the moment doesn't need my ministrations.
The line level crossovers are 3 way 24dB/octave LR type of my own design. The capacitors in those are REL RTE polystyrene foil, 2% tolerance and selected where possible to 1%, and matched in the same location on the other channel. I kind of like SMD so the resistors are Panasonic 0.1% 1206 thin films, op-amp based, and there I selected OPA1641.. I/O is all balanced.
There is a lot of tube hardware in the system which demands periodic maintenance. I waited too long, having completed some repairs and whole tube replacements in the amplifiers.
It's a bit of a challenge when sometimes all you want to do is listen to music. I have a lot of sources as well - some like the TD-124s need attention soon. The Studer A810 is the only thing really at the moment doesn't need my ministrations.
Attachments
You can use those techniques in active systems as well. I do.. 🙂
Notches and other filters can all be implemented in the active line level stuff or in DSP. You can do quite sophisticated stuff without the electrical losses and non-linearities of passive components operating at significant power levels.
There is little or nothing really in passive implementations that cannot be implemented in active, and driving drivers directly as I do generally eliminates the need for zobels if the amplifier is a voltage source - it won't care at all about variations in load impedance. Since I sort of like 0 fdbk SE tube amplification I have interactions that can be addressed through the addition of zobels or I can just equalize it out - I currently do both. Zobel on LF driver and direct drive to all.
The only argument I would make against active is cost and complexity, and a comparative lack of easy tweakability unless you have a lot of experience/knowledge. My system continues to evolve, at times it is a headache. Based on what I hear though when everything is right I will never go back to a passive implementation for my main listening system. I would however not attempt to replicate it anywhere else! 😀
I evolved a 5 way system.
X/O points 90Hz, 350Hz, 1200Hz 12000Hz.
Chosen to suit the drivers and the horns I could make.
Just the inductors and caps for the bass tapped horns and mid bass straight horns cost more than the excellent Najda DSP X/O solution that was developed on DIYAudio!
Going active mean I could remove Zobels, caps, inductors and L-Pads (as each driver had to be equalized volume wise due to varying efficiencies).
Those passives sap dynamics and clarity, I don't care how much $$$ you throw at them.
They get really big and heavy if you want 4th, 6th or 8th order.
The infinite variations you want to try when designing and testing passives make it a very costly enterprise.
You can't time align the drivers.
The tapped horns are over 9ms behind the rest due to horn length.
Distances get very long at low frequencies. You can move stuff that far away to align.
Even sitting on them and the rest away from you, you can't do it!
Then there is room mode control.
If you want deep bass, church organ, electronica deep to be done well, most rooms have modes. Could be 29Hz like my room and/or other frequencies.
Major manufacturers roll off their speakers to prevent you bring unhappy with them booming..
You see specs like 32Hz -3dB.
So at 29Hz and lower there's less energy.
So you might introduce something like Antimode in line level that calibrates itself to reduce.
Done well the ADC and DAC conversion is very transparent.
When I first got Najda I put it into my analogue / passive system on pass through.
I could not hear ANY sound deg.
Upgraded with Burson OP amps is a peach.
As Najda is 4 way only (unless you have a rare expansion board), I do still have a passive crossover on the Raal tweeters. 3rd order Bessel.
From there, removing the passives, adding carefully chosen amps (SET amps for the ribbon tweeters and compression drivers), and bigger high damping factor silicon amps to control the 12 and 15" bass drivers made a revelation in my system.
I can introduce PEQ filters to tame my down to 20Hz flat tapped horns.
Done carefully this is very effective.
You can have room shaking bass if the performer intented it, or fast tuneful bass as intended, all the way down.
I only have PEQ on the tapped horns. At those frequencies it doesn't mess with the sound like full room correction seems to[emoji2961].
The rest is as is and they are close enough and sound better left alone.
Different gains of the drivers taken care of in Najda.
You do need quiet amps! Connecting SET amps directly to 200Hz tractrix and 550Hz Le Clèac'h horns, with the horn gain and no sapping passives, compression drivers that are 115dB/W efficient, you need good amps.
But that's all part of the hobby and journey.
So for me, it's horses for courses.
For 2 / 3 way conventional box speakers, passives are fine.
For more complex multi driver systems, active is a total no brainer - for me at least, and I've done both.
Getting X/O points and slopes right too suit the drivers means X/Os are transparent.
Time aligning properly is not to be underestimated.
Also you could spend $20,000 on one masterpiece full range amp, and I've borrowed a few to try and they get whipped on bass duty or on dynamics a little well built EL84 SET amp on the compression drivers can achieve.
Also I agree the DSP X/O system needs to be good quality.
I tested Behringer DCX and even modded it was lacking in SQ.
X/O points 90Hz, 350Hz, 1200Hz 12000Hz.
Chosen to suit the drivers and the horns I could make.
Just the inductors and caps for the bass tapped horns and mid bass straight horns cost more than the excellent Najda DSP X/O solution that was developed on DIYAudio!
Going active mean I could remove Zobels, caps, inductors and L-Pads (as each driver had to be equalized volume wise due to varying efficiencies).
Those passives sap dynamics and clarity, I don't care how much $$$ you throw at them.
They get really big and heavy if you want 4th, 6th or 8th order.
The infinite variations you want to try when designing and testing passives make it a very costly enterprise.
You can't time align the drivers.
The tapped horns are over 9ms behind the rest due to horn length.
Distances get very long at low frequencies. You can move stuff that far away to align.
Even sitting on them and the rest away from you, you can't do it!
Then there is room mode control.
If you want deep bass, church organ, electronica deep to be done well, most rooms have modes. Could be 29Hz like my room and/or other frequencies.
Major manufacturers roll off their speakers to prevent you bring unhappy with them booming..
You see specs like 32Hz -3dB.
So at 29Hz and lower there's less energy.
So you might introduce something like Antimode in line level that calibrates itself to reduce.
Done well the ADC and DAC conversion is very transparent.
When I first got Najda I put it into my analogue / passive system on pass through.
I could not hear ANY sound deg.
Upgraded with Burson OP amps is a peach.
As Najda is 4 way only (unless you have a rare expansion board), I do still have a passive crossover on the Raal tweeters. 3rd order Bessel.
From there, removing the passives, adding carefully chosen amps (SET amps for the ribbon tweeters and compression drivers), and bigger high damping factor silicon amps to control the 12 and 15" bass drivers made a revelation in my system.
I can introduce PEQ filters to tame my down to 20Hz flat tapped horns.
Done carefully this is very effective.
You can have room shaking bass if the performer intented it, or fast tuneful bass as intended, all the way down.
I only have PEQ on the tapped horns. At those frequencies it doesn't mess with the sound like full room correction seems to[emoji2961].
The rest is as is and they are close enough and sound better left alone.
Different gains of the drivers taken care of in Najda.
You do need quiet amps! Connecting SET amps directly to 200Hz tractrix and 550Hz Le Clèac'h horns, with the horn gain and no sapping passives, compression drivers that are 115dB/W efficient, you need good amps.
But that's all part of the hobby and journey.
So for me, it's horses for courses.
For 2 / 3 way conventional box speakers, passives are fine.
For more complex multi driver systems, active is a total no brainer - for me at least, and I've done both.
Getting X/O points and slopes right too suit the drivers means X/Os are transparent.
Time aligning properly is not to be underestimated.
Also you could spend $20,000 on one masterpiece full range amp, and I've borrowed a few to try and they get whipped on bass duty or on dynamics a little well built EL84 SET amp on the compression drivers can achieve.
Also I agree the DSP X/O system needs to be good quality.
I tested Behringer DCX and even modded it was lacking in SQ.
Last edited:
Тhere are probably a hundred different Mundorfs. There is an abyss between Clarity MR and ESA.
And still, those dull, ill defined in the bass, Soviet ex-submarine horrors kill them all? 🙂
Not IME. Have you tried any Siemens/ Bosch MP of young Berlin Wall vintage?
I have tried some Siemens caps they were good. I was actually surprised by the performance of the MBGP caps, my expectations were quite low, my experience with Soviet surplus caps excluding the FT-3 which was OK IMO was not good. And what is up with those micas - awful.
At this point it'll remain irrelevant as I unfortunately have neither the time or space for another system and will continue on the active path with this one.
At this point it'll remain irrelevant as I unfortunately have neither the time or space for another system and will continue on the active path with this one.
No argument about the micas 🙂
And yes, your active path seems very impressive but hard to turn into an easily repeatable, universal recipe.
Any in particular? I developed a distrust of old used silvered micas for use in valve amps, in the lower nF range and measured for capacity before use.. but I never stopped to figure out why.
There was a fad around Russian micas for use in phono stage EQ and a few other places. I bought a lot of them based on raves here, and quickly decided they were terrible sounding - much prefer polystyrene or polypropylene foil types. Tight tolerances at low cost, but they sounded like... mica. I don't remember if I investigated any further, they were unlistenable.
People outline active as being more expensive to get going. I'd wager a properly stocked inventory of passive crossover parts is not cheap either! Especially if you live in places where shipping is a major cost and self manufacture (e.g. inductors) is difficult.
PS: I am a passive guy about to venture into active.
I'd also recommend limiting your self to a single digital to analogue conversion (if like me you are a digital source person) - if that helps you decide which side your put your crossover in.
PS: I am a passive guy about to venture into active.
I'd also recommend limiting your self to a single digital to analogue conversion (if like me you are a digital source person) - if that helps you decide which side your put your crossover in.