[John Ashman]

No, John, the DACS are not good.  I know it, Overkill knows it  (and said so on the 6moons review) and many others know it.  I, unfortunately, had to end up evaluating it as a pure dac since I could not use the room or speaker correction properly with the problematic speakers, nor the crossover capabilties.  In bypass I was simply eavaluating the dacs, and compared to what I have they don't cut it.

No, Ted, you were evaluating the analog section in all reality.  There is *nothing* wrong with these DACs.  If there is, do you have a citation from somebody that has no financial interest in upgading who says that these DACs suck?  My point is that exceptional DACs are a dime a dozen these days.  They've largely been perfected.  The analog section isn't perfect and it does have a bit of noise.  Overkill has no credibility to say that DEQX DACs aren't up to par when they are ridiculous enough to use the Manger driver with it.  Talk about a *bad* match!  What you really mean, IMO, is that, as a stand alone DAC, it's not going to match a $2000+ component because that's not its purpose and the money isn't in the analog section, it's in the DSP.  But the DACs inside it are just fine and could easily be dropped into any high-end compnent and sound great.  Replacing the DACs won't do dick for the performance of the DEQX.  It may be possible to improve if you're clever enough to upgrade the power supply and analog section.  

I also wasn't specifically speaking of you, there are a lot of people who find that the DEQX is not probably worth $3500 to use only a room EQ and/or speaker EQ.  And I agree.  I wouldn't buy it for that.  Its real value is as an active digital crossover which is magical.  And more to the point, I don't find the VMPS speakers to be that bad of a match for DEQX, I was thinking more of horns, single driver systems, etc.

[John Ashman]

Ted is 100% correct -- look at these units as sophisticated DSP devices that offer convenience DACs.  To get the best possible performance from redbook CD, buy dedicated external units or use amps with digital inputs.  Right now I am only aware of TacT/BOZ/Lyngdorf and Behold offering direct PCM>PWM amplification.  But later in the year Spectron will join the club, ...

Ted's about 10% correct (and I *like* Ted!!!) - as I mentioned above, the 90+% of the performance increase you'd get would be because of the improved analog sections and power supplies of said outboard DACs.  If the budget is unlimited, you'd run digital into the DEQX, digital out to 3 high-end DACs and then into something like the McCormack 6-channel preamp, then into three amplifiers.  But you'd be adding $5000+ to the price tag.  If Overkill Audio was actually serious, that's what they'd do.

[ekovalsky]

90+% of the performance increase you'd get would be because of the improved analog sections and power supplies of said outboard DACs.  If the budget is unlimited, you'd run digital into the DEQX, digital out to 3 high-end DACs and then into something like the McCormack 6-channel preamp, then into three amplifiers.  But you'd be adding $5000+ to the price tag.  If Overkill Audio was actually serious, that's what they'd do.

Well I for one consider the DACs and analog output stage as one, being that together they take a digital signal and output an analog signal to be used by an amp, preamp, etc.

Yes the way to go is dig-in and dig-out.  Maybe you can convince DEQX to make a version lacking the analog circuitry at a lower price -- like the "DD" versions of the TacT gear.  A dig-only DEQX would match perfectly with a basic BOZ amp, which retails at $9k for the two channel version then $3k extra for each additional two channels.  For a tri-amp setup you would need six channels so retail is $15k.  Nice discounts are available if you know where to look



Not sure why you think the Manger is such a poor choice for the DEQX.  Overkill's implementation takes advantage of a 96dB/octave phase coherent crossover which lets the Manger play down to 200hz, where the woofer picks up.  Remember the best crossover is no crossover!

The pricing on the Overkill is another story, but the "no discount" policy is a myth.  I was offered a new system about 50% off MSRP by the manufacturer, once he realized I was considering the used/demo pair that was (and still is) on Audiogon.

[John Ashman]

Ted, the Manger driver is *very* high in distortion.  Unbelievably high.  But I guess it can go very high in frequency, but wow, I think it was as high as 5-10% distortion in places.  There is no need for a full range driver with DEQX - you do a proper 3-way crossover to the tweeter around 2khz -3kHz.  

Also, trimming out the analog section of the DEQX wouldn't save much.  If it did, maybe it wouldn't warrant upgrading at all.  They've considered making an "audiophile" version, but it's a chicken/egg thing.

[ted_b]

John,
I can't believe I have to defend myself against such comments as "Ted is 10% correct" or "you're evaluating the analog section, not the dac".?  When one buys an outboard dac, for example, they are listening to the chip (d/a) and the analog section.  I don't give a rats which one is at fault, I'm just saying that the sound of the Deqx's dac section (in common parlance) is not that great.  And so have others. It can be improved.  And one small step is to kick the chip up to 24/96, which Kim and Co. are planning.


I'm stepping out of this conversation.  It's going nowhere.

Mr. 73.4594% correct

[goskers]

A couple questions for the more experienced around here.  What are the major differences between a unit like the dcx2496 and DEQX units?

The pdc2.6's have up to 300db linear phase slopes.  The dcx2496 has a max of 48db slopes but it seems that if you still choose moderately well behaved drivers and match pass and stop bands well that anything above 48db slopes should not be 'needed'.

What are peoples view points on phase?  I know it would be great to have linear phase but how audible is this change?

Room correction is done on both with FIR filters if I am not mistaken.  In a treated room, I would think that room correction would be more of a non-issue.

Aside from these things I would love to hear about the internal differences between the two, specifically dac's and power supplies.

Thanks for the help

[JoshK]

Gosker,

I am no sage on the subject.  But with regards to DCX vs. DEQX there are a few points that I will make, whether that matters to you is your choice.

Let's say you want to push a dome tweeter down into the 1100-1400hz range because this allows you to have 11" of center to center spacing for say an MTM (most commercial MTMs don't adhere to the rule and thus have lobing problems).  There a number of robust dome tweeters these days that can do this with a steep enough slope, however LR4 is probably not steep enough.  Lets say in this design you want to use Dayton RS series 7" because they are inexpensive, high performance and have very pistonic behavior in their pass band.  But they have really nasty break up modes for which LR4 is certainly not enough even well below the mode.

The natural thought would be to use LR8 (48db/oct) on the DCX and you are good to go right?  Not really.  The problem is LR8 have been well known to have negative effects on sound because of excessive group delay and phase problems.  The DCX mimics analog crossovers and has the same phase and group delay as its analog peer because it uses IIR filters.   The DEQX uses an FIR filter, and although it isn't without its own problems it doesn't suffer from group delay and phase issues with such steep slopes, in fact it corrects for them.  

So besides FR, RC and phase correction, the DEQX also provides group delay correction (see the white paper from DEQX).  I recently came upon this fact when studying Jon Marsh's 8th order Cauer-Elliptic filter (passive) and the benefits it provides over LR4 & LR8.  The jury is still out whether the DCX can mimic the C-E filter so that it can provide the sufficiently steep slope for some really great designs but without the group delay and phase issues of the LR8.

I honestly don't really care too awful much that the DEQX isn't the last word in transparency.  I have heard this too.  Big deal.  There are already a number of reports on how to improve upon it relatively easily.  The TACT has been reported a few times to suffer the same problems.  You can mod the DEQX or get the digital outs and go ape crazy if you want.  Myself, I'll probably mod the DEQX at some point down the road.  Maybe I'll run it on batteries.

[jhenderson010759]

No, John, the DACS are not good.  I know it, Overkill knows it  (and said so on the 6moons review) and many others know it.  I, unfortunately, had to end up evaluating it as a pure dac since I could not use the room or speaker correction properly with the problematic speakers, nor the crossover capabilties.  In bypass I was simply eavaluating the dacs, and compared to what I have they don't cut it.

That being said, you don't buy the DEQX for its quality of DACS (Overkill uses it despite them; they're put ...

What a bunch of baloney.  Have you measured the D/A performance on the PDC?  I have - using the Stanford signal generator and Agilent spectrum analyzer at work, and the performance is excellent.  

Moreover, I have listened to this unit for several months now, and it is discernably quieter than was the DCX2496.  

The DEQ software is far easier to learn and use than SoundEasy, which I still use in conjunction with the DCX2496.  

The only reasonable criticism of the PDC is value-for-the-money.  The retail price is way too high.  Luckily, I bought an eval unit at 1/2 price, to help mitigate this problem.

[Davey]

What are the DAC's in the DEQX?  One of the AKM devices?

Davey.

[John Ashman]

Jim,
    My only quibble is "what price value"?  I mean, people buy $30K amplifiers that are 1% better than $10K amps which are 1% better than $2000 amps, if that.  You could argue that DEQX is a bargain at $10K because, IMO, it is the single biggest improvement you can make to your system and nothing else has the technology to do what it does.  If someone else had a unit that was even 90% of the PDC for $2000, then sure.  But the best units have *maybe* 25-50% of the capability, so, they can pretty much charge whatever they want and get it, just not from as many people.  I don't think DEQX is over charging for it, I mean, they're definitely not getting rich off of PDCs.  They could charge more, but they're not, they're charging what they need to, possibly less than that.  I doubt they're even profitable on it yet.

[jhenderson010759]

Jim,
    My only quibble is "what price value"?  I mean, people buy $30K amplifiers that are 1% better than $10K amps which are 1% better than $2000 amps, if that.  You could argue that DEQX is a bargain at $10K because, IMO, it is the single biggest improvement you can make to your system and nothing else has the technology to do what it does.  If someone else had a unit that was even 90% of the PDC for $2000, then sure.  But the best units have *maybe* 25-50% of the capability, so, they can pretty much  ...

In my opinion, the DCX2496 + DEQ2496 + SoundEasy is damn close to the PDC.  And that combination only costs about $700.  

But, as always, that last 10 or 15% of performance will cost you dearly.  Nevertheless, a PDC at $10,000 would not be a bargain.   Not when I have actually used (and still own!) the Behringer equipment myself.  

What the PDC and Behringer+SoundEasy combination have in common is measurement.  Measurement is the single most important aspect of these systems.  All of the hacking around that people do with their gear - exchanging this and upgrading that - is traceable to their unwillingness or inability to measure.  For $400 and a PC, you can equip yourself with world-class measurement tools to analyze your entire system.  When you do this, and when you take measurements, you'll readily identify the problems in your system.  Overwhelmingly, this will be your room acoustics.  But you can also readily quantify dispersion problems or even subtle distortion issues.  It just takes a little time and effort.  But a trained seal could do it.  

If you don't do this, you're a hack.  You are like a drunken hemophiliac in a razor-blade factory.  You are in a dangerous place and you won't exit unharmed.   At least, your wallet won't.  

People need to quit screwing around with the .000001% changes due to alternate digital transports and the .00000000001% changes due to cables and focus their efforts where they matter.  

Where is that?  Measure, and you'll see.

[ctviggen]

I think measurements are interesting, but what do you do with them?  My speakers are pretty much where they are going to stay; I don't plan on buying more gear; I will buy some additional acoustical treatments.  I do plan on using the measurements to better integrate my sub and mains.  Other than that, I don't plan on doing much with them.

[jhenderson010759]

I think measurements are interesting, but what do you do with them?  My speakers are pretty much where they are going to stay; I don't plan on buying more gear; I will buy some additional acoustical treatments.  I do plan on using the measurements to better integrate my sub and mains.  Other than that, I don't plan on doing much with them.

If you're simply enjoying music, and not actively/obsessively upgrading, then you don't need to measure.  You're in good shape.  

But if you are chronically dissatisfied with the sound of your system and habitually spend money or time altering it, then you must start with measurements and see where they take you.

For example, my living room was too reverberant.  So, I have purchased tapestries, lined them with Owens Corning SelectBoard (fiberboard) and mounted them at some strategic locations throughout the listening space.  I used SoundEasy to optimize and determine the effectiveness of their placement.  So, even if your gear remains constant, measurements can add value.

[ekovalsky]

Ted, the Manger driver is *very* high in distortion.  Unbelievably high.  But I guess it can go very high in frequency, but wow, I think it was as high as 5-10% distortion in places.  There is no need for a full range driver with DEQX - you do a proper 3-way crossover to the tweeter around 2khz -3kHz.  

Also, trimming out the analog section of the DEQX wouldn't save much.  If it did, maybe it wouldn't warrant upgrading at all.  They've considered making an "audiophile" version, but it's a chicken/egg thing.

Whoevoer tested the Manger and found 5-10% distortion probably tried using it over its entire rated range of 80-35khz.  It only operates as a bending wave driver above 150hz.  As it is not designed to move pistonically below that range, distortion rises dramatically below 150hz.  So the solution is to use a "brick wall" crossover to keep any frequency below 150hz out of the driver.  This is why Overkill chose to use the DEQX with a 96dB/octave filter at 200hz.  When used in this way distortion (within its dynamic envelope) is probably lower than ay other type of driver.  It also has incredibly flat amplitude and phase behavior and impedance in that range.  See graphs here.  The biggest limitation of the Manger driver is its dynamic capacity.  Since only one can really be used in a system, it cannot manage the SPLs that planars, cones, and horns can achieve.

Back to the DEQX.  If there is not much to be saved by eliminating the DACs and analog output stage, that implies they are cheaply implemented.  Meanwhile, digital outputs (which are cheap to implement, since the signal is already digital) are a $500 option.  What DEQX should do is offer two base units, one with analog outs and one with digital outs.  That way the customer can choose which path he/she wants to take and is not forced to pay $500 extra for six digital output jacks on the rear panel.

[Aether Audio]

Gentlemen,

I haven't posted in this forum until now but I have been watching it from time-to-time.  For whatever reason, I felt motivated to chime in now though.

jhenderson010759 wrote,

If you don't do this, you're a hack.

...

People need to quit screwing around with the .000001% changes due to alternate digital transports and the .00000000001% changes due to cables and focus their efforts where they matter.

Where is that? Measure, and you'll see.

AMEN!!!  I really can't ad much to this other than this one point.  As an example, I will refer to amplifier design to make that point.

As many of you know, some of the best sounding amplifiers aften use little or no negative feedback to achieve stability and/or linearity.  It is true that the use of larger amounts of feedback to achieve an excellent sounding design is possible, but it is no trivial matter and the designer had really better know his stuff.  Any attempt to do so without a PHD level of knowledge will most often end up with a rather poor sounding device.

In the early days of audio it was perceived that a large open-loop gain combined with equally large amounts of feedback was an automatic formula for producing a quality design.  We know better now.  Large amounts of negative feedback or “correction” are not a "cure-all" by any means.

Just the same, it is definitely no small matter to produce a zero or low negative feedback design that works well either.  That same PHD level of knowledge is almost a prerequisite as well.  The one relaxed requirement for a designer of this type amplifier is that his design is far less likely to oscillate at rf frequencies and self-destruct if everything isn't "perfect."

The major drawback is that both tubes and transistors are inherently non-linear devices, so without the advantage of much, if any negative feedback, the odds are the design will suffer from higher levels of conventional distortion.  Such distortion products are clearly audible and really do need to be eliminated as much as possible to achieve a pleasant sounding amplifier.

If one were to take a survey of the best sounding and most reliable amplifiers that have been produced to date, one would likely find that the majority of these designs would have a common design approach.  More often then not the would have the following:

(1)  They would have each stage designed to be as linear as possible WITHOUT the use of any feedback.

(2)  They would use localized or "nested" loops around each stage before global feedback was applied.

(3)  They would then only require small amounts of global feedback - if any.

The upshot is that the best designs do not require large amounts of global feedback to produce a linear response with low amounts of distortion.  It only makes sense to start out with the most inherently linear design possible - then tweak it with global feedback to clean things up a little.

So what's my point?  The same holds true for speakers.  The DEQX or any other digital gizzmo is no fast track or guarantee of superior performance.  To achieve “world class” performance, you need to start with as linear of a design as possible and then use such devices to fully optimize the design.

It his been stated in this thread that the DEQX and it’s ilk will put a lot of speaker companies out of business as now any “do-it-yourselfer” will be able to produce as good of results as the “pros.”  Well,  that statement may not be popular with some manufacturers as it is likely the truth of the matter – at least in many cases of inherently inferior design.

But as far as the few companies that are producing product based on real and superior technology goes, I don’t think they’re loosing any sleep over these devices.  Those companies offering truly superior product will simply implement the DEQX et al into their systems and produce even more superior products.

Let’s put it this way, not only do we at SP Technology not fear the DEQX…we embrace it whole-heartedly.  We have one in our facility now and if all expectations are realized, we will be offering it as an upgrade package for our systems.  We like to think of it as the ultimate tweak on an already inherently linear and superior design.

Any would-be “hackers” out there that think they will be reproducing our waveguide technology with minimal effort and in short order may find themselves a little “in over their heads.”  Yet, without such a device or something similar, they are juggling drivers and parameters that are inherently less linear – as most other manufacturers have already been doing for years.  Simply adding the DEQX will not get you there.  But then again, you’re all better off with it than without.  The big question is:  Is the financial savings of building your own system reduced by the added cost of the DEQX, such that it is still cost competitive with available designs being offered by manufacturers of truly quality products?  Maybe…if you really do your homework.  Study hard, have fun and good luck!!!

Take care,
-Bob

[John Ashman]

Two points:

1.  Notice there are no distortion figures in there.  The guy who told me this is a reputable guy, others agreed and it apparently is that bad throughout most of its range.  Plus, it doesn't matter that it's so good at phase and accuracy, these are *easily* fixed by DEQX.  The fundamental things a driver needs to do for DEQX is be very low in inband distortion and have good dispersion.  The Manger appears to be interesting, but not appropriate.  I'd like to hear NHT's Xd (at <10% of the price) next to the "ultimate audio" setup and I bet you the the  more expensive system gets bitchslapped.  

2.  You could save a few $hundred by not including an analog section, but then the unit would be nearly unsellable to most people because it would require $thousands in ancillary gear.  So having that *relatively* cheap analog section is a life saver when you want to sell it.

PS - I'll apologize for the "bitchslapped" comment, but won't retract it!  It's true  And I just am sick of overdone, overpriced pretentious audio (and often, the people that love it - nobody here, trust me).  I thought the whole point of this was music!?!

[goskers]

I completely agree with Bob in the fact that just because you have a unit like the PDC does not mean that a lot of basic elements for good speaker design do not apply.  

Many people have said that the best systems assembled for PDC usage are extremely pistonic in their respective passband.  This of course means that mag cone, ceramic, ribbon and other speakers would come into play.

My goal is to use either the RS180's or the upcoming peerless hds exclusive in an open baffle dipole mtm arrangement with either a small AC ribbon or a proper OS waveguide.  The lower portion are dual TC2+'s per side in a u-frame.

[skrivis]

Actually, its the the other way around.  Fabric would ring more in the bandwidth but would be fairly well self-damped, aluminum is more pistonic, but rings like a bell, typicall around 22kHz where it shouldn't be perceptable.  Some poorer aluminum tweeters range down in the audible range causing fatigue.  I've alwas found fabric tweeters to be "pleasant", but not realistic on cymbals (I'm a drummer).  I'd rather have realism than this current audiophile trend towards rolled off, "pleasant" tweeters.  I do l ...

Metal drivers ring like a bell.... yes, they do. Unfortunately, the effect isn't limited to high frequencies. Resonances at high frequencies will also cause problems at lower frequencies.

My take on this is that I'd rather have realism than this current audiophile trend toward bright, metal "detailed" tweeters.

Metal drivers may be pistonic over a certain range, but once they misbehave, they really misbehave.

Ted Jordan has worked around some of this by only using cone drivers. The part of the driver farther away from the voice coil (and it's control of motion) is then controlled by the surround.

Domes want to become ring radiators at high frequencies (just like cones want to become smaller cones). With a doped fabric dome, the impulse from the voice coil dies out toward the center of the dome.

With a metal dome, there's not much damping as the impulse propagates through the dome. As long as the stiffness of the material controls things, its fine. Once that's no longer adequate, the whole thing breaks up, and does so rather violently.

[skrivis]

Could you actually even EXPLAIN what you're talking about with this?  Like in two or three sentences.  You're speaking in riddles, alluding to problems that aren't even the subject of this thread.  Or maybe you can start a thread and tell everyone about it, but at this point, you're just hijacking the thread for no apparent reason.

To get at the basic issue, stereo has major flaws. They're built-in.

Let's say we're recording a concert with 2 spaced mics.

A musician on the left side of the stage stands up and says, "Now Hear This."

Our left mic picks it up first, because it's closer. Our right mic then picks it up, somewhat later and at lower amplitude.

When we play this back over a typical stereo loudspeaker system, the left channel produces "Now Hear This" first. Then the right speaker. Just like what the mics caught.

Ok so far?

But we hear the sound that came from the left speaker first in our left ear and then our right. The right channel is the reverse.

A single event ("Now Hear This") has been transformed into two events, one per channel, and then four events, two per ear.

I believe that csero was pointing out that there are huge problems with most speaker systems that aren't being addressed.

DEQX and the Xd may be part of the solution, but there's more to be done, and it's useful to have that point brought to our attention. Not only that, but some people may have differing opinions on what the worst problems are.

Differing opinions? About audio gear?


 

[JoshK]

skrivis is right.  Once I heard Csero's system it really helped me see that the standard audiophile route of upgrades wasn't addressing the key bottlenecks and that a further leap forward could be done addressing the issue of stereo reproduction even using "lesser" equipment to yield better overall results.