comparison between digital x-over and passive filters

[ttan98]

They are many passive x-over designers and users of these designs here and some of them have also used digital x-over eg. DCX2496 as well.

The designer can design filter to shape the curve to fit the speaker curve in a very flexible way.,eg. asymmetrical, 1st order, 2nd, LR, Butterworth, etc. The digital x-over is less flexible but still it can be configure in variety of ways, eg.,asymmetrical, etc.to fit ones needs.

The digital x-over allows users to forces the audio chain to be in the digital domain throughout until it reaches the analogue amplifier(if you are using one).

I don't have that much experience with designing and experimenting with passive x-over, but I have started to use a digital x-over. I like the latter. The latter saves me money by not having cupboard full of expensive capacitors and inductors that I won't be using and I find hard to sell after using them.

my query is that those who has experience with using both methods please explain your experience eg which one do you like, etc. I am sure many here would find it useful, I for one would like to listen to those who have used passive components extensively.

cheers.

[ttan98]

The last time I posted this no one cared to response. There must be someone here who have experience with using both types of x-overs.

Those who are passive diehards will probably swear by it, that is using passive components. I suspect there are few passive users here have just bought DCX2496, to try and experience with using it. Maybe they are ready to comment now?

Thanks....

[cjd]

I've only heard one speaker with a back-to-back listen on one digital crossover (DEQX not Behringer) and one passive.

It was clear even then that there was nominally more low level detail in the passive crossover.

The Behringer crossed system I heard was so disappointing I don't like using it as a point of comparison - I don't know WHAT wasn't right, but again low level detail was missing and it simply made a mash of large orchestral works, comparatively. And with Seas Excel in there, very confusing. I don't know where the issues lay in that design - not enough time to be fair to it.

Expense to me is keeping 3 high quality amplifiers per speaker plus crossover: no way have I spent $1k on a passive crossover per speaker! Ok, so I could go cheaper amps, maybe only add $500 per speaker (I'm counting cost for 2 channels not 3 since I need at least one amp no matter how I look at it). I can see getting close to that cost passively, but no way will the noise floor on the active equipment be anywhere close to equivalent.

In a recording of orchestral works, based on a test a local audiophile did (he removed all signal that was -50dB and lower) room cues exist BELOW -50dB. THey're there, they're audible. The result of that test was a very dead and lifeless presentation - similar, in fact, to the Behringer system I heard (but not mushing the sections either).

Then again, I heard a high frequency issue over at Ryan's listening to his Khan's on his Behringer A500 - it wasn't there with his Pioneer amp section. So far I think I'm the only one that's heard it.

C

[CraigJ]

Ah CJD,
Next time you get to Mad town, please audition the Isiris Juniors that are utilizing the dcx, your mind "might" be changed. The "mash of large orchestral works" that I hear, is usually from poorly recorded orchestral works. ttan98, here is an excellent discussion regarding Victor's system; https://www.htguide.com/forum/showth...071#post331071

Cj

[RYDMOTO]

Has anybody looked at the JBL,EAW or Ashly digital crossovers?
I am wondering if you get what you pay for so to speak.

[cjd]

Ah CJD,
Next time you get to Mad town, please audition the Isiris Juniors that are utilizing the dcx, your mind "might" be changed. The "mash of large orchestral works" that I hear, is usually from poorly recorded orchestral works. ttan98, here is an excellent discussion regarding Victor's system; https://www.htguide.com/forum/showth...071#post331071

Cj

You have both a passive and active version? That is the question at hand, ultimately.

The album I used is one I'm VERY familiar with, from my own collection. I'll be the first to admit, recording quality comes second to performance quality, but in this case a poor recording isn't worth much for evaluating speakers.

On the general subject: I still don't understand why I should CARE whether digital can sound as good. It still costs a helluva lot more and introduces a lot more points of failure.

[servicetech]

Incorporating digital crossovers in a 5.1 system is difficult and expensive.

[ttan98]

You have both a passive and active version? That is the question at hand, ultimately.

It still costs a helluva lot more and introduces a lot more points of failure.

A new DCX2496 costs as little as $250(*) and cheaper at Ebay. You can get the audio signal digital all the way till it reaches the amplifier.

One of the main attractive feature about active x-over is that you don't have to spend heaps of dollars buying expensive passive components. If I decide the speakers is worthy(I build more than one pair of speakers) then I can try to use passive. My point is that whether passive is actually superior to active or vice versa. I still have NOT compared them yet. Most probably with careful design and comparison both design may sound comparable that is my guess. I will try one day.

Point of failure I disagree, DCX is easy to use after you get use to it. You would be surprised a small amp can control the driver quite easily with active x-over. I use digital T-amp to drive a 87db driver quite easily obviously the bass is lacking but the dynamics is there.

Most people here own speakers with passive components, and some of you here also own digital x-over, a together of these 2 groups of people to conduct a comparison is possible. I can loan my digital x-over however I live too far away. Just a thought.

* a set of speakers whose high quality passive components can cost as much as $100-$200 and sometimes more.

[cjd]

A new DCX2496 costs as little as $250(*) and cheaper at Ebay. You can get the audio signal digital all the way till it reaches the amplifier.

And amplifiers for each stage cost $500+... well, depends what kind you use. But you get the picture. The A500 is NOT one I would consider for anything other than subwoofer duty after hearing one.

Point of failure I disagree, DCX is easy to use after you get use to it. You would be surprised a small amp can control the driver quite easily with active x-over.

Not an ease of use thing. Active electronics are all susceptable to things like power spikes, static discharge damage, heat fatigue, etc.

I would consider digital for experimentation absolutely. I might consider one beyond that as well, but flexibility would NOT be a reason - I just don't see it since passive is easy to mess with.

C

[IllNastyImpreza]

I have experience with the whole active vs passive comparison from my car audio days.

When it comes down to it, a fully active settup (once its been tuned) will always sound better than its passive counterpart. Now... whether this is due to the fact that each listener has a different ear, and the actives are adjustable I don't know.

But 99% of the true car audio enthusiasts will run an active settup over passive any day of the week.

now when it comes to HOME audio... you don't have the same size restrictions(when building passives), BUT you have to worry about 7 freakin channels !(vs only 2) even though it might sound better, its just not worth the money for most people...

[cjd]

When it comes down to it, a fully active settup (once its been tuned) will always sound better than its passive counterpart.

Always is a pretty strong suggestion. And not true.

Car audio guys run active because tuning passive is a pain in the tail, and there is SO much of it to do in a car setup.

Building an active that has the noise floor of a good passive is expensive.

*shrug*

C

[ttan98]

And amplifiers for each stage cost $500+... well, depends what kind you use. But you get the picture. The A500 is NOT one I would consider for anything other than subwoofer duty after hearing one.


C

I prefer mainly to 2 or 3 way speaker, 2 channels system.

For a 2-way speakers you don't need $500 amplifier to drive the tweeter! T-amp is enough.

Ideally for a 3way speaker, the tweeter and midwoofer interface can be passive and midwoofer and woofer are driven separately, then you need only 2 amps. High quality low to medium powered amps are plenty to choose from and relatively cheap($100-300) eg T-am, Chipamp, SKA, Aska, or pick your fancy, etc.

I also believe in efficient speakers, >=91-92 db speakers. Here you get a high quality(wide sound stage, dynamic and open) sound system at a relatively low cost. I am now experimenting with high efficient speakers systems, incl. Open baffle, horn, compression driver in 2 and 3 way configurations. With active x-over, I can achieve 95-96db using high quality drivers at an attractive price. Hence it becomes an affordable hobby!

There you are, others may prefer a different approach. There is no wrong way, ultimately if you like the sound, then it is the right way. It is not a competition here. I am just curious. Cheers.

[Dennis H]

I have experience with the whole active vs passive comparison from my car audio days.

When it comes down to it, a fully active settup (once its been tuned) will always sound better than its passive counterpart.

Apples, oranges.

People with no design experience, no measurement gear and no XO design software, working in a car where none of the drivers are placed for the best sound, like down by your feet or bouncing off the windshield, and you're way off center for the stereo image -- sure, active is easier to tune by ear by adjusting the relative levels of the drivers. If you're in a more optimum environment, like a living room with the speakers at ear level, and you sit midway between the speakers, and you have the gear and the skills to do it right, there's no clear advantage one way or the other. Some things are better about active and some are better about passive. What's considered good in the home is a whole higher level than what's considered good in a car. The car environment is so inherently bad that the best you can do is kinda okay.

[servicetech]

1: With a car environment you're only dealing with 2 channels.
2: Multi channel amps are no more expensive then 2 channel amps of the same total power
3: Basic electronic crossovers are built into the amps, even external electric crossovers are inexpensive.

In a home Hi-fi they have already figured out separating the deep bass from the rest of the sound is a good idea. That's why we have a subwoofer output.

[cjd]

I prefer mainly to 2 or 3 way speaker, 2 channels system.

I find my big 3-ways need a little help <30Hz. 4-way I guess. A sub IS generally best crossed in actively though, where you're tuning the room as part of the system, and every new room is a different tuning. Also, low level detail is generally not in this range. Not to mention that the component size and cost to cross at 40 or 50Hz passively is... silly.

For a 2-way speakers you don't need $500 amplifier to drive the tweeter! T-amp is enough.

Or, not. T-amp is OK for nearfield mini-monitors, but large systems that it can power comfortably are both large and expensive in their component cost up front, in my experience. But you're right, I'd probably put in ~$150 monoblocks. Actually, in this case I'd probably build a single power supply dual channel amp to handle both mid and tweeter so that'd be ~$250ish. But that's still per channel, so $500 for a pair of speakers...

Ideally for a 3way speaker, the tweeter and midwoofer interface can be passive and midwoofer and woofer are driven separately, then you need only 2 amps. High quality low to medium powered amps are plenty to choose from and relatively cheap($100-300) eg T-am, Chipamp, SKA, Aska, or pick your fancy, etc.

Ok, so... wait. what? This is a different discussion, you know? :P And I think especially when we look back and see you may instead be talking about 2-way + sub - can't tell. But, the issues I've found in digital (so far) or cheaper active are far less an issue as the frequency at which they're used goes down. And mid-tweeter is far more critical of low level detail, which so far is where I've found passive to most clearly stand out ahead. We're easily talking -40 or -50dB for some of this.

I also believe in efficient speakers, >=91-92 db speakers.

So do I. Mine are >90dB (probably just shy of 91). Well, my main music rig. Nor do I consider the ~350W on tap per speaker to be anything less than a minimum. Though were I to bi-amp I would probably run less on the mid/tweet.

Here you get a high quality(wide sound stage, dynamic and open) sound system at a relatively low cost.

What? Low cost? Quality low distortion high-effeciency drivers are nowhere NEAR low cost, on relative terms. They're not necessarily silly expensive, but they're half again the cost of some of the quality low distortion stuff like the Dayton RS line or Peerless Exclusive line. That may be, in part, my particular taste in drivers with minimal sonic signature.

It is not a competition here. I am just curious. Cheers.

I'd agree with that. I've simply tried to highlight where I felt the systems I *have* heard comparisons of digital vs passive fell short.

One other thing I suppose I should note: I'm probably more critical of low level detail changes than most (though those "veils lifted" are sometimes really just low level detail changes, whether accentuation of a frequency to make them move obvious or actual noise-floor reduction). And they're not things I'll necessarily be aware are *missing* (the example of the orchestra being mangled and mushed goes way beyond this as a problem - I have NO idea what caused that at this point) unless I'm dealing with a side by side comparison.

Sometimes we get so carried away saying "I can hear a difference in an A/B test" that we forget that we really probably can't remember well enough to note a difference in anything other than a near direct switch. Even if people claim to be able to.

C

[EspenE]

The last time I posted this no one cared to response. There must be someone here who have experience with using both types of x-overs.

Those who are passive diehards will probably swear by it, that is using passive components. I suspect there are few passive users here have just bought DCX2496, to try and experience with using it. Maybe they are ready to comment now?

Thanks....

Here is my take on the subject:

I have designed and used passive, analog electronic and digital (DCX2496)crossovers in many projects. For me, crossover network design always means carefully optimized filter transfer characteristics, matched to the actual frequency response, rolloff and phase characteristics etc of the actual driver units in the actual box. Measuring equipment and some means of crossover optimization is a must. This holds true for passive, analog and digital networks.

That being said, for me analog electronic networks is by far the best type of crossover. But this involves building and modifying your own electronics. Admittedly, this is not always feasible.

The Behringer DCX presents an attractive way for obtaining an electronic crossover optimized to your actual speakers without having to construct and build your own electronics. However, this convinience comes at a price. IMO the sound of properly built analog electronic crossovers is superior to the Behringer - even if the transfer characteristics are quite identical. Even well-designed passive crossovers usually sound better than the Behringer (maybe except for very low crossover frequencies with correspondingly large passive inductors).

The reasons for this:

1. The Behringer is noisy. There is no master volume control. To minimize noise, you will want to keep the signal levels on the Behringer's inputs/outputs as close to clipping as possible. That means a volume control after the Behringer - between the Behringer and your power amps - is a must.
With power amps of normal sensitivity/gain and efficient speakers, the hiss level will IMO be objectionable without a volume control between the Behringer and your power amps. Keep in mind that for a 3-way stereo system, this has to be a 6-channel close tracking volume control - not easy to implement. There exists some 6-channel electronic volume controls as DIY kits, but they easily cost almost the same as the Behringer unit.

2. The Behringer DCX is a pro device, and is easier to implement with pro equipment. The levels, gains and interconnects match. With consumer equipment, it is another matter.

The ideal config would be a digital signal into the DCX with the DCX feeding its outputs into a 6 channel volume control before each amp.

The less ideal config would be use a high-gain preamp feeding the DCX' analog inputs. (You need a fairly large signal the analog inputs in order to maximize signal-to-noise-ratio. Most likely, your consumer preamp is a –10 dB device with RCA outputs. The Behringer is a pro +4 dB device That meants your signal is about 12 dB low. You need to amplify the signal 12 dB - if not your signal to noise ratio will be worse by 12 db, and you also loose those 12 dbs of digital resolution). You then feed the DCX signal through a 6-channel attenuation network before the amps.

3. The Behringer can correct for large differences in sensitivity between units - for example 90 dB/W for a bass unit compared to a 112 db/W horn driver. But it uses some of DSP processing power to do that, and the sound quality suffers. Sensitivity corrections can just as easily be accomplished with passive resistors (L-pads) directly on the speakers - with audibly better results.

4. The Behringer is very flexible, and has a multitude of adjustments. With it, you can accomplish just about any transfer function. But for good results, it is totally necessary to measure the speaker individually - both without and with filtering. Measuring equipment is a must. I also doubt that it is really possible to get consistent, good results without a crossover optimization package. I use lspCAD.

The Behringer is a complex tool in a cheap package. IMO, it can certainly be a fun and useful tool for fast and easy prototyping. The low price can fool us into believing that it is a unit that is simple to implement and use. For good results, it is not. But used with the above caveats in mind, I would have no reservation to recommend it.

[cjd]

Hey, some of us run true balanced signal out of our pre-amp stage. I've got 25 foot balanced cable runs to the amps, and 18 inches of 12ga from the amp to the speakers, in my main rig. Much easier to run the cable that way.

C

[fjhuerta]

But 99% of the true car audio enthusiasts will run an active settup over passive any day of the week.

Isn't that because of the size of the passive components required for subwoofers in car, losses associated with said components, and the difficulty of trying to tune a passive system in a car?

[servicetech]

Isn't that because of the size of the passive components required for subwoofers in car, losses associated with said components, and the difficulty of trying to tune a passive system in a car?

1: It's because going active is actually LESS expensive than a passive setup when it comes to car audio. Mass production has driven active crossovers down in price. In HT active crossovers are considered ultra high end and have a price tag to match. Multi channel automotive amps are relatively inexpensive, no more than a 2ch of the same total power.

2: Since every car is different having the flexibility to easily make system changes onces a system is in installed is essential. Acoustics are almost impossible to predict in most vehicles.

3: You're only dealing with 2 channels instead of 5-8 channels in HT.

4: Noise floor isn't an issue in an automotive system.

[ttan98]

Here is my take on the subject:

I have designed and used passive, analog electronic and digital (DCX2496)crossovers in many projects. For me, crossover network design always means carefully optimized filter transfer characteristics, matched to the actual frequency response, rolloff and phase characteristics etc of the actual driver units in the actual box. Measuring equipment and some means of crossover optimization is a must. This holds true for passive, analog and digital networks.

That being said, for me analog electronic networks is by far the best type of crossover. But this involves building and modifying your own electronics. Admittedly, this is not always feasible.

The Behringer DCX presents an attractive way for obtaining an electronic crossover optimized to your actual speakers without having to construct and build your own electronics. However, this convinience comes at a price. IMO the sound of properly built analog electronic crossovers is superior to the Behringer - even if the transfer characteristics are quite identical. Even well-designed passive crossovers usually sound better than the Behringer (maybe except for very low crossover frequencies with correspondingly large passive inductors).

The reasons for this:

1. The Behringer is noisy. There is no master volume control. To minimize noise, you will want to keep the signal levels on the Behringer's inputs/outputs as close to clipping as possible. That means a volume control after the Behringer - between the Behringer and your power amps - is a must.
With power amps of normal sensitivity/gain and efficient speakers, the hiss level will IMO be objectionable without a volume control between the Behringer and your power amps. Keep in mind that for a 3-way stereo system, this has to be a 6-channel close tracking volume control - not easy to implement. There exists some 6-channel electronic volume controls as DIY kits, but they easily cost almost the same as the Behringer unit.

2. The Behringer DCX is a pro device, and is easier to implement with pro equipment. The levels, gains and interconnects match. With consumer equipment, it is another matter.

The ideal config would be a digital signal into the DCX with the DCX feeding its outputs into a 6 channel volume control before each amp.

The less ideal config would be use a high-gain preamp feeding the DCX' analog inputs. (You need a fairly large signal the analog inputs in order to maximize signal-to-noise-ratio. Most likely, your consumer preamp is a –10 dB device with RCA outputs. The Behringer is a pro +4 dB device That meants your signal is about 12 dB low. You need to amplify the signal 12 dB - if not your signal to noise ratio will be worse by 12 db, and you also loose those 12 dbs of digital resolution). You then feed the DCX signal through a 6-channel attenuation network before the amps.

3. The Behringer can correct for large differences in sensitivity between units - for example 90 dB/W for a bass unit compared to a 112 db/W horn driver. But it uses some of DSP processing power to do that, and the sound quality suffers. Sensitivity corrections can just as easily be accomplished with passive resistors (L-pads) directly on the speakers - with audibly better results.

4. The Behringer is very flexible, and has a multitude of adjustments. With it, you can accomplish just about any transfer function. But for good results, it is totally necessary to measure the speaker individually - both without and with filtering. Measuring equipment is a must. I also doubt that it is really possible to get consistent, good results without a crossover optimization package. I use lspCAD.

The Behringer is a complex tool in a cheap package. IMO, it can certainly be a fun and useful tool for fast and easy prototyping. The low price can fool us into believing that it is a unit that is simple to implement and use. For good results, it is not. But used with the above caveats in mind, I would have no reservation to recommend it.

Thanks for your insight, I read through them slowly and will be re-reading it.

I have noted your experience and surely take that into considerations.

[JRT]

Thanks for your insight, I read through them slowly and will be re-reading it.

I have noted your experience and surely take that into considerations.

If you want to try a rapid prototyping development cycle with your loudspeaker design, then you might consider using SoundEasy on your PC with a multichannel sound card (one that works well with SoundEasy) for the early portion of that development phase. When you get close, you can translate that to passive filters, or translate it to active filters in the analog or digital domain, or use a combination. You can also use SoundEasy to develop filters specifically for the DCX2496 if you don't want to use a PC sound card for longer duration evaluation of your design efforts.

Here is the link to SoundEasy.

http://www.interdomain.net.au/~bodzio/

Here is a link to a guide.

http://www.musicanddesign.com/Guide.html

While you are there, take a look at John Kreskovsky's NaO and NaO-Mini desings that use a combination of passive and active filters. The design economics of that combination is worth serious consideration in my opinion, especially for one-off designs or low volume production.

[ttan98]

If you want to try a rapid prototyping development cycle with your loudspeaker design, then you might consider using SoundEasy on your PC with a multichannel sound card (one that works well with SoundEasy) for the early portion of that development phase. When you get close, you can translate that to passive filters, or translate it to active filters in the analog or digital domain, or use a combination. You can also use SoundEasy to develop filters specifically for the DCX2496 if you don't want to use a PC sound card for longer duration evaluation of your design efforts.

Here is the link to SoundEasy.

http://www.interdomain.net.au/~bodzio/

Here is a link to a guide.

http://www.musicanddesign.com/Guide.html

While you are there, take a look at John Kreskovsky's NaO and NaO-Mini desings that use a combination of passive and active filters. The design economics of that combination is worth serious consideration in my opinion, especially for one-off designs or low volume production.

Now all those with good knowledge are coming out to contribute that includes you.

I know the feature you stated here, btw I have Soundeasy as well.

[pixelpusher]

Hi,
I haven't posted much but I must relate my experience on this.

I am nearing the end of a long development process of a 3 way dipole speaker system in which I have used both Soundeasy and LspCad. Upper crossover is passive and lower is active employing a Marchand tube crossover LR24 at 400hz. Recently I tried a Behringer DCX2496 in place of the Marchand. I was quite surprised at the improvement particularly in the lower midrange and bass. I then installed Jan Didden's active output mod for the DCX2496. The improvement is profound in every way. I am extremely pleased!
Mike

[ttan98]

Hi,
I haven't posted much but I must relate my experience on this.

I am nearing the end of a long development process of a 3 way dipole speaker system in which I have used both Soundeasy and LspCad. Upper crossover is passive and lower is active employing a Marchand tube crossover LR24 at 400hz. Recently I tried a Behringer DCX2496 in place of the Marchand. I was quite surprised at the improvement particularly in the lower midrange and bass. I then installed Jan Didden's active output mod for the DCX2496. The improvement is profound in every way. I am extremely pleased!
Mike

J. Didden's mod is not cheap about the same price as the unit if I am not mistaken...I will do the mod one day. Still developing my speakers.

[Ludvig]

Here is my take on the subject:

1. The Behringer is noisy.
2. The Behringer DCX is a pro device, and is easier to implement with pro equipment. The levels, gains and interconnects match. With consumer equipment, it is another matter.
3. The Behringer can correct for large differences in sensitivity between units - for example 90 dB/W for a bass unit compared to a 112 db/W horn driver.

I've been using the DCX2496 for a couple of years, both for prototyping and for final crossover work in my Hifi/HT-system.

Your points are absolutely correct. To get rid of much of the problems I have modified the gain stages on all inputs and outputs of the unit. This way I have normal sensivity (about 1.00 Vrms full scale) on both inputs and outputs which gives about 10-12 dB better S/N. The results of this modification was substantial. A also use a passive LPAD on my waveguides to cut about 20 dB in sensivity. These two techniques has given me a fully satisfactory results far better then standard out of the box use.

While I still think that an analog active crossover would give a few sonical benefits, the flexibility of the DCX is more important for me. A crossover need to be touched and tweaked every now and then. That's why they exist :lol:

[augerpro]

Can you guys document or link to these DCX mods?

[ttan98]

Can you guys document or link to these DCX mods?

this link is very lengthy, it contains lots of useful info and many mods but you have search within the threads to find what you want.

the site is:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=15943

[pixelpusher]

Can you guys document or link to these DCX mods?

Here's a link to Jan Didden's DCX2496 mod

http://www.linearaudio.nl/6chan-1.htm

[AJINFLA]

this link is very lengthy, it contains lots of useful info and many mods but you have search within the threads to find what you want.

the site is:

http://www.diyaudio.com/forums/showt...threadid=15943

The only thing missing is a minor inconvenience called DBT's, aka Audiophile Kryptonite. The bane of the I can hear this and that Golden Ears.

cheers,

AJ