Why doesn't anyone use loudspeaker management instead of crossovers in high end desig

[bvbellomo]

Jon, I'd be interested knowing what your cousin thought of the difference between the Isis and Avalon was. Did the 6640 leave something to be desired vs the diamond tweeter? Was the c173 better than the c90? Was the bottom end of a huge floor stander just that much more powerful?

[wkhanna]

i (& all of my fellow audio friends, one of whom is an audio reviewer with access to a multitude of equipment) have & continue to hear differences in cables passing digital signals.
unfortunately, despite hearing many theories as to why, i could not tell anyone why these differences exist.

edit: one of the more reasonable explanations (to my feeble mind) that has been put forth concerns areas of time domain relative to signal transfer.......

[Evil Twin]

There is a difference between cables on an analog system. Digital cables either work or they don't work, all working digital cables deliver the exact same signal.

If this viewpoint helps you sleep peacefully at night, then read no further in this post.

Digital cables ARE analog cables, too- as Bruno Putsys notes, EVERYTHING is analog, and if you think otherwise, get out your 12 bit high resolution 1GHZ scope and look at the output of different S/PDIF cables or AES/EBU cables, differing in construction technology and and length, document the waveforms at the sending and receiving ends, then come back to the forum and make this claim again with a straight face, posting your documented results.

Pray you do not challenge me further on this point without documentation, unless you too wish to share Captain Needa's fate.

[Juhazi]

Thank you JM and 5th element for posts! I do understand the idea and goal of high-end, but I hate what all crap is thrown in that category.
Dsp integration is certainly a growing trend in hifi. Latest Stereophile mag has test of Beolab 90 and MartinLogan Masterpiece, in high esteem, I just read it.

http://www.stereophile.com/content/happy-new-year-little-early#7vMjTI7wS49a84bP.97

[TEK]

I think that there are points to be made on both points of view.
From my point of view a main issue is that a lot of high-end equipment is extremely hyped and not really worth what their cost is. However, what something is worth is of course a subjective matter.
If you pay 10k for something and that give you a subjective feeling that the sound is a lot better than something costing 100, then that feeling might be worth it for you.
An example, I was just in a store where they sold record players.
Dust covers, in plexi glass, costed around 250 usd.
If you go into a grossary store a plexi box with almost the same quality cost around 30 usd.
The cost difference is due to production, quantity and marketing issues, not actual quality issues. And absolutly not sound related issues.
I belive very mutch is the same with all audio gear, especially high end.
As it gets more exclusive, it gets more expensive. That does not nessesarly means that the quality is better. Bit it might, but how do you know???

[BobEllis]

If this viewpoint helps you sleep peacefully at night, then read no further in this post.

Digital cables ARE analog cables, too- as Bruno Putsys notes, EVERYTHING is analog, and if you think otherwise, get out your 12 bit high resolution 1GHZ scope and look at the output of different S/PDIF cables or AES/EBU cables, differing in construction technology and and length, document the waveforms at the sending and receiving ends, then come back to the forum and make this claim again with a straight face, posting your documented results.

Pray you do not challenge me further on this point without documentation, unless you too wish to share Captain Needa's fate.

Dark Lord, if you or your apprentice could grace those of us with minimal familiarity in the art of high frequency signal transmission and the workings of digital receivers with a brief explanation of the impact of such waveform distortion, it would be greatly appreciated. My working guess is that at lower system resolving powers digital receivers are rather tolerant of distortions but as system resolution and listener expectations increase waveform distortions cause the apparent timing of the signal to change. Is this jitter or at least a form of it? Understanding (to my feeble capability) of the mechanisms involved will help me justify expenditures to move me deeper into the dark side. I have yet to hear a difference in interconnects, but then again I have not tested since adding Ardents to my system and my DAC is not yet up to their level.

[wkhanna]

as has been explained to me by those in the industry (such as David Salz, the owner of Audioquest & most recently a VP from Nordost).......therefore, bias not withstanding.......:

the perfect cable is totally transparent.
no coloration or distortions from the original signal.

all other cable is subtractive only.

any improvement that a non-perfect cable provides to the audio quality is either one of improved transparency or a subtractive coloration that systemically compensates for other characteristics within the system.

I wonder if our wise but occasionally impatient lord might be able address & enlighten us, his loyal & most reverent padawans, on such heretic & blasphemous utterances, if in fact they are such?

[bvbellomo]

Digital cables ARE analog cables, too- as Bruno Putsys notes, EVERYTHING is analog, and if you think otherwise, get out your 12 bit high resolution 1GHZ scope and look at the output of different S/PDIF cables or AES/EBU cables, differing in construction technology and and length, document the waveforms at the sending and receiving ends, then come back to the forum and make this claim again with a straight face, posting your documented results.

I shall assume full responsibility for this discussion...

Consider an analog recording. Even a short recording has an infinite amount of information - all frequencies to the infinitely small are captured, it may or may not allow an infinitely loud sound (infinitely high voltage) but has infinite dynamic range (no limit to the number of different volumes) and an infinite amount of time information (a signal can start at any time). If we were to write down any of these numbers, we'd have irrational numbers like Pi that repeat forever, and we'd need an infinite number of numbers them to represent our recording. When we transmit this signal, we lose some of it - wire has resistance, which causes voltage drop across all frequencies, impedance which filters out high frequencies, and several sources of noise (bad connections, radio waves) add noise.

Now consider a digital representation of that recording. We quantize this information, which results in information loss. We may sample 48,000 times per second, so any information between samples is discarded, making impossible to accurately store frequencies above 24,000Hz, and signal times are a fraction of a millisecond too soon or too late. We have a finite number of volumes, typically 65,536 different levels, which can give a large, but not infinite, dynamic range. We now have a finite amount of information, we can write this down as rational numbers with a finite number of digits after the decimal point, and we would need a finite amount of them. We could record the whole thing by hand on paper - we'd need a lot of paper, but it is possible.

Now, to transmit this digital signal, we have to modulate it (covert it to an analog representation). We represent all the numbers as binary and send them across the wire. People think of binary as either "on" or "off", but typically we alternate between 2 voltages - suppose 1 volt for 0 and 2 volts for 1. It is more complicated (for example Differential Manchester Encoding used with S/PDIF has clock information), but not much. Our analog signal starts as alternating between 1 volt and 2 volts, but gets sent along an analog cable and degrades the same as any other analog signal. The difference is, while we have signal degradation, as long as it is less than a certain amount, we don't have any information loss. On the other side, we demodulate it (re-encoding our analog signal to a digital representation), and anything greater than 1.5 volts is treated as 2 volts, and anything less is treated as 1 volt. So a high quality cable and signal might give us alternations between 0.9 - 1.1 volts and 1.9 - 2.1 volts, where an inferior cable gives us 0.7 - 1.2 volts and 1.6 - 2.2 volts. As long as we can resolve it back to the same binary numbers on the other end, we haven't lost information.

Of course, if our signal degrades too far, we get back different 1's and 0's. There is redundancy built into some encoding, such as Ethernet, so we can detect with a very high probability if any of our bits got flipped - and we discard the whole frame (a large amount of Ethernet data) and ask it to be resent. For data sources without redundancy, even a small amount of bit flipping gives us so much static and noise we can barely recognize the original signal. In practice, we usually end up with either perfect transmission or total garbage.

Think about it. If you download an mp3 file, you get the same (usually crappy) recording whether it went through very expensive equipment in Jacksonville or very cheap equipment in Japan.

[5th element]

Of course cables are different and as Jon says...

Digital cables ARE analog cables, too- as Bruno Putsys notes, EVERYTHING is analog, and if you think otherwise, get out your 12 bit high resolution 1GHZ scope and look at the output of different S/PDIF cables or AES/EBU cables, differing in construction technology and and length, document the waveforms at the sending and receiving ends, then come back to the forum and make this claim again with a straight face, posting your documented results.

No one is trying to say that the outputs are going to be different. Of course things are going to look different when enough bandwidth is tested but this is entirely the point. The cabling in any system is chosen specifically so that it's properties do not have any significant effect on the performance of the system.

Cables have parasitic inductances, capacitances, resistances, dielectric losses, varying degrees of shielding etc etc...this is not new and no one here is trying to sweep these under the mat, but cables are designed to perform a function. You can exceed this function, or try to misuse a cabling system, and when done...the cables will have a significant impact on the performance of the total system. But this is not a situation that you should ever encounter in your typical hifi system and nor is it anything you should ever encounter in your home network with cables sold that adhere to the various networking standards/specifications.

Exceeding a cables capabilities is easy to demonstrate. Pretty much everyone in a 1st world society is familiar with DSL internet, or rather what could be called 'high speed internet' transmitted via the old copper cables that were originally installed for phone communications. This is a prime example of using a cabling/transmission system that is NOT designed for the task.

We are all familiar with the way the adverts work, saying 'up to 24 Megabit' speed etc and then you never get any faster than 5. Ever wonder why this is? Cabling losses.

Basically speaking you have a main junction box for your housing area that supplies that area with the telephone/DSL internet. The closer you are to this junction box the shorter the copper cables to your house. When you are within 100 meters or so you can sync in and connect with the advertised 24Mb. But as you get further away, cable losses come into play and the speed that the system is capable of transmitting starts to decrease.

For what its worth the data rate of an uncompressed CD is 1.4Mbps. Even on the old telephone cabling system you can transmit bit perfect data over several miles, via modern DSL protocols, and this is in a severely compromised system.

At a basic level pretty much all modern audio equipment is constructed and soldered to copper clad PCBs. Now is there anything particularly special about this copper? Absolutely not. A quick google returns figures of 99.8% purity, certainly nothing like the 99.9999....% etc that cable companies try and push. But you don't see any of the well designed audio products failing to meet astonishing levels of signal integrity because of it. If you need less trace resistance you can use a PCB with 2oz or 4oz copper. It's a little bit more expensive, but not by that much. I use 2oz board for all my PCBs.

If the copper on the PCB is absolutely fine as a conductor inside the box of my hifi equipment, why oh why do I need something ridiculous as a means of carrying the signal 30cm from my CD player to my preamp? The truth? I do not! I could probably use a section of steel coat hanger and find no degradation in signal integrity within the bounds of what the system is called on to reproduce. Of course the coat hanger wouldn't stand up to scrutiny when compared with something like Belden's 1694a. The 1694a has been specifically designed to transmit high speed video signals over long distances of and the coat hanger has not, but you'd mostly likely have to push the bandwidth orders of magnitude above the audible bandwidth to find anything of significance.

Of course we're not talking about a rusty coat hanger here. We're talking about a properly terminated and soldered coat hanger, using suitable connectors, which brings me onto something else. Connectors are an area where things can fall down because these are usually held in place by material tension. With the signals being transmitted between two conductors that have been pushed within close proximity of one another.

All you need though is a connector of solid construction and you'll be fine. A professional connector from a reputable brand, such as Neutrik, will do just fine. These are designed and marketed specifically at industrial professionals who need reliability and excellent signal integrity. They don't cost a tremendous amount either.

So the next time you need some interconnects (or a new SPDIF cable) do yourself a favour. Buy some Neutrik connectors (or something comparable), some 1694a and go crazy. If 1694a is good enough for high speed video over large distances, it's more than good enough for your analogue audio signals and low speed digital. Not to mention that characteristic impedances are largely ignored in S/PDIF systems too. To this end some companies do offer BNC connectors for their digital connections but most do not. It simply isn't required at the frequencies S/PDIF works at and the cable lengths that are typically used.

Much the same are network cables. If standard network cables are good enough for data transmission between super computers in places like CERN and for Google data centres. Then it's a pretty safe bet they are absolutely fine for sending your audio from one place to another in your home network.

If a DIY cable (we are in the DIY section are we not?), constructed from suitably chosen, industry standard connectors and cabling, sounds different to audiophile boutique cables then you can pretty much bet it's because the audiophile cables are junk. It's well known that some audiophile cables have had extremely high internal capacitances that have pushed, some poorly designed (only conditionally stable) amplifiers into oscillation, where they quickly then became acquainted with silicon heaven.

Thank you JM and 5th element for posts! I do understand the idea and goal of high-end, but I hate what all crap is thrown in that category.
Dsp integration is certainly a growing trend in hifi. Latest Stereophile mag has test of Beolab 90 and MartinLogan Masterpiece, in high esteem, I just read it.

You are of course correct that it is unfair to tarnish all with the same brush. There are lots of companies and individuals within the audiophile community that do not adhere to the ridiculousness that others do.

Of course the whole industry, as a general rule, is about pretty boxes and certain levels of dogma, but it's easy to see why. Heck I like pretty boxes and shiny lights too! I appreciate the aesthetics, just not the ludicrous price tags. But the dogma goes hand in hand with making money. Subjective aural impressions are intangible, unquantifiable and so easily manipulated, by anything that isn't actually related to 'the sound', that it's easy to pray upon people's engineering naivete, gullibility, sense of aesthetic, sense of comradery (yay I now too have a Valhalla cable!), sense of participation (I chose my cables all by myself!)...you name it, it can impact on the way you think it sounds because our aural memory is so absolutely terrible.

In this day and age all boutique loudspeaker manufacturers should be producing loudspeakers like B&O. Active, digital input, DSP based and with internal amplification matched to the rest of the system. Then again all boutique loudspeaker manufacturers should be producing loudspeakers that measure superbly, but that's not happened either.

[Juhazi]

"analog audio = infinite, digital = finite"
Sorry, I don't buy that. Analog audio suffers from many many bandwith, dynamics and resolution-reducing problems - noise, distortion, etc. nonlinearities etc. that appear at every step of production and reproduction path, including cables!
It is just romanticism to call it infinite.

[sfdoddsy]

Ive been using digital speaker management for many years, ever since I built my first linkwitz Phoenix doppelgangers, which became Orionish simulacra, then lX521 inspired cousins, and now a Seas CA18 that is vaguely reminiscent of the Small Lyngdorf system.

I started with the Driverack mentioned by the OP, but it wasn't flexible enough. I tried various other off the shelf solutions and combos of DSP and active crossovers. For a long time I used the Behringer DCX, but it is flaky. Then I switched to various configurations of Mini DSP.

Unless you build your own, or go full out for a DeQX or similar it is probAbly the best choice.

The sound quality is good enough for Linkwitz.

That said, whilst I still use the MiniDSP i combine it with a passive crossover for tweeters and mids. It is not really for sonic reasons, but more for aesthetic and practical reasons involving multiple amps and cables.

I am also a long time user of DSP for EQ.

[AJINFLA]

I'm going here this weekend.
My local audio club is meeting there and I'm friends with the owner, who's part of the club. There are over 100 members. Several >$100k systems (some much much higher), I've heard them all. I've also demo'd at over a dozen high fashion audio jewelry shows, as a reference. I've had countless bling pieces through my own system, sent by manufacturers, several club members are owners of high end companies (no names to protect the innocent ). So I consider myself fairly immersed in the audio bling world.
I've also been to a few recording studios, like the ones that make most all recordings. Seen the "quality" of the cables, electronics, ADC/DACs etc. that make the music embedded in the media being used by audiophiles to judge their mega $$ cables and electronics.
Out of the hundred members, there is one system that stands head and shoulders above the rest, by consensus. A gent who has some huge DIY OSWG waveguide/dipoles + 8 subwoofers. 15 yr old Pass and Cary amps. Home Depot burial cables. A Minidsp 2x10 is the heart of the system.
It's with some amusement that I read comments about "quality" issues with "cheap" digital/active based on sighted evaluations.

cheers,

AJ

[BobEllis]

I think we've been trolled. The op made one post and hasn't been on this board since 8 hours after starting this thread. Interesting discussion anyway.

[IslandHydro]

. . . .

[JonMarsh]

Probably got a bit too hot in the kitchen

Nah, we're not an opinionated bunch or anything... and we don't bite. And if we do, most of us have had our shots... :W

[Alaric]

most of us have had our shots...

I know I had a shot before diving in this thread. LOL

[Zvu]

OT

Was MiniDSP 2x8 on the bench ? I can't find it.

I think nanoDIGI 2x8 is a good solution. It does not have ADC (that i don't want or need) and has 4 SPDIF outputs so i can choose which DACs to use.

[Squidspeak]

If this viewpoint helps you sleep peacefully at night, then read no further in this post.

Digital cables ARE analog cables, too- as Bruno Putsys notes, EVERYTHING is analog, and if you think otherwise, get out your 12 bit high resolution 1GHZ scope and look at the output of different S/PDIF cables or AES/EBU cables, differing in construction technology and and length, document the waveforms at the sending and receiving ends, then come back to the forum and make this claim again with a straight face, posting your documented results.

Pray you do not challenge me further on this point without documentation, unless you too wish to share Captain Needa's fate.

OK Evil one, I have to say this statement clarified some of my preconceived notions about digital and it makes sense to someone who is far from a "Tech" person. But are you saying different digital cables can
transmit altered information, cause sound/music is info in the digital realm. Not doubting just trying to digest the conversation.
Also what I am concluding from the discussion about analog or digital cables is the level of investment in the chain is more
relevant as quality goes up (high end). Is your opinion on the DBX or MiniDSP or Berh. units that at their price points they
cannot be using quality parts? I currently use the Behringer DCX but was considering the newer DBX as an upgrade, do you
or any other forum members have an opinion between the two? These units are in the budget/level of the rest of the system.

[JonMarsh]

You know, quality is a matter of perspective.

My own experience as an EE is that often other engineers in our organization overlook issues with cables and probes when doing data collection.

There have been so, SO many posts on various forums about the issues with any digital medium that transmits data and clocks simultaneously on the same signal lines, that I'd be surprised if we have to talk about ringing and termination issues with connectors and cable lengths, (due to reflections ringing back and forth in a certain time interval creating "ambiguity" in the exact arrival time. Most folks, of course, don't have a 350MHz or 500MHz scope at home to check out what's happening at the digital receiver. And let's not even start talking about optical S/PDIF and the intrinsic jitter in all of the receiver chips I'm aware of...

At a certain level of performance, nearly any cable with electrical continuity and no gross issues may do. At other levels of performance, that may not be the case without audible degradation. Getting really transparent sounding digital seems to take more effort than most realize- I didn't set out in life wanting to buy a customized TotalDAC but after comparing a heck of a lot of different pieces of equipment since 2006, upgrading when something sonically better came along, that's where I wound up. In that process there were a lot of units that were good, but didn't make the cut at a particular step as being better enough to justify upgrades. That includes the PS Audio PerfectWaceDac, a couple of Meitner models, and the MSB Platinum DAC III. For years now, I've also been using DAC's with rechecking units such as the Brainstorm DCD, or more recently, the Mutec MC-3+; both with rubidium reference standards.

I do often "mess around" with lower cost gear to see what is good these days that I can recommend- in that category, the AURALiC Vega comes to mind (measures very, VERY well, used to be it took $10K or more to match it) and a possible new cheapie that's sort of a chip off the shoulder of a TotalDAC. It's good to have more reasonable priced things to be able to recommend.

On a speaker like the Ardents or better, any of these will reveal their capabilities and strengths, give good associated electronics.

I will say that regardless of what one says about 24 bit volume controls using high precision math, I've so far never encounter a case where having the DAC set to reference output and going through a reasonably top line preamp didn't sound better than using it direct with 20dB of attenuation, particularly with CD material, which is 16 bits and STAYS 16 bits at the most whether you're up sampling to a higher frequency sample rate or not. Since I do listen to a lot of CD level material, that is important.

YMMV, what I experience and decide should in no way limit your enjoyment of your system as currently configured. Just be aware that there may be vistas ahead of you, which you are not aware of yet.

[joec]

actually, I don't think the BeoLab 90 is high end, mostly because of ICEpower amplifiers, which I have evaluated.

The thing is, especially in DIY, this is about what makes you happy- if they're making you happy at a price and system complexity you're good with, then just sit back and enjoy.


BTW, how about a survey- what are you miniDSP guys using for power amplifiers for your multi-way systems?

This thread is very interesting to me! I built the Modula MT years ago, and then this past summer I built an LX521 with the MiniDSP 4x10, and I use a home-built UCD180 eight channel with SMPS1200 power supplies. I have been trying to get good measurements of the amplifier before I clean up the (admittedly bad) internal wiring so that I can do a comparison, but it's been harder than I expected to get good measurements of the amp with my sound card as an interface.

Here's a blog post that I made showing my amp: http://totallyshould.com/?p=926

I did go to Music Lovers in Berkeley, and I spent about three hours with the Wilson Alexx driven by the big Boulder monoblock amps. Overall my impression was that the sound of my system in my room is harsher and less pleasant than what I heard at Music Lovers. However, the bass articulation, neutral timbre, and overall imaging was at least as good when I heard the LX521 in Siegfried's home. My own listening room is too live and needs treatment, and I know that it isn't living up to its potential right now, and I'm careful not to extend my conclusions to the MiniDSP or the amplifier before I sort that out. My friend has the Orions with the ASP and a B&K amplifier, and his imaging and tone is better than mine... at least in his room that is more dead and has more soft surfaces in it.

I'm using the unbalanced connections on the MiniDSP right now based on the measurements posted by Tom on the Neurochrome site. Hopefully in the next couple of months I'll have an adapter for the eight-pole neutrik connectors so that I can easily try some different amps in my room.

[ModMark]

But are you saying different digital cables can transmit altered information, cause sound/music is info in the digital realm.

If you can hear the difference between a low and high end digital cable. both working, then blame the protocol instead of the cable. Mr. Marsh is correct, analog signals are used to transmit digital data and the quality can vary greatly. That is why engineers developed error detection/correction protocols to deal with this problem. My 20 year old Cat5 cable running from the attic office to the basement router may be corrupting 50% of the packets but the Ethernet protocol just keeps re-sending the packets. In a office environment, this is a problem but not at home.

I assume high end audio equipment uses the latest HDMI protocol but this may be a wrong assumption. Seems silly to sell a high end $20,000 system only to have the data corrupted by using old technology such as S/PDIF. The latest HDMI standard now includes an HDMI Ethernet for a secondary data transmission connection.