 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
129.33.19.254
I was reading about Emerald Physics speakers, and quietly wondering what kind of audiophile would willingly insert something like Behringer DCX2496 into his system, let alone buy a $5400 speaker system that relies on that device to perform crossover and DSP duties. Then I recalled that several inmates here use Thuneau Allocator to perform said functions in software, thereby eliminating not only need for the device itself, but also the need for sound-degrading AD/DA conversion in principle (DCX2496 is supposed to be inserted between preamp and amp). I believe at least one inmate (Presto) compared both solutions, and had no doubt that Allocator is superior in every way.
But, there are several questions/concerns I have in relation to that:
As I understand, to use Allocator, you need to have one of the following:
Option 1:
- sound card with multi-channel analog outs;
- either use software volume control, or multi-channel preamp;
- multiple (or multi-channel) amplifiers.
Option 2:
- sound card with multi-channel digital outs;
- multi-channel DAC (or multiple DACs);
- either use software volume control, or multi-channel preamp;
- multiple (or multi-channel) amplifiers.
I see many limitations to either approach, namely:
- scarsity of sound cards meeting requirements, especially in Option 2;
- inferior quality of breakout boxes/cables, that one may need to deal with in both cases;
- inferior DACs in Option 1;
- potentially sound-degrading software VC, unless DACs have adjustable gain (or preamp is used);
- potentially high cost, especially with option 2.
So far, that's in my opinion far too many trade-offs for just eliminating passive crossovers found in conventional speakers. Unless I just got it all wrong - in which case I'd appreciate being (politely if possible) corrected.
Thoughts?
Follow Ups:
Hi there.
I am looking for a feasible solution since years.
This was one reason why I started my Linux journey.
I still consider Brutefir&Acourate filters the top of the top in terms
of filter and sound quality.
My challenges:
1. I havn't seen a multichannel solution, which delivers better sound quality (in terms of microdetails and harmonics ) then my stereo setup. (my fullrange speaker runs without any passive filtering btw.!)
Multichannel soundcards usually come with poor outputstages (opamps)
and PS.
I am also not willing to spent +4k for a commercial multichannel DAC. ( assuming that these might be able to do a better job)
2. If you follow the "CICS logic", running heavy filter processing will cause sound-degradation on the PC. ( If decoding flacs makes a difference already - what do you think would large filter processing do?)
On the other hand I don't think that any HW based filter can achieve the result of the PC based filtering ( with all its flexibility)
3. And not only the processing might cause problems, it is also the filtering itself. No filter is lossless. You will gain time and phase linearity and also a better in-room response - on a first glance this will be breathtaking. Stagewidth & depth, in-room transient response, seperation - just great. You can't beat that with any 2 channel passive setup.
On the other side there is a big chance that you loose micro details/harmonics (the soul of music) - at least that's my experience.
What keeps me sitting in front of the speakers is the transported "soul"
of music. Not the sound effects.
Don't forget to think about what's required to process different samplerates.
4. You need to do some "measuring" and a pretty good understanding of what you're doing. Not many people are able to do that. (And don't have the equipment to do so.)
5. Multichannel amps, cabling, mains supply all this gets really expensive if you want to put it on to a highest-end stereo system level.
Without knowing what you'll end up with - this could mean to burn a lot of money.
Will it be worth all the effort and risk? I guess nobody can answer that.
Good luck.
Cheers
> > Will it be worth all the effort and risk? I guess nobody can answer that.
Actually, some of us can. From my POV there are those who, for whatever reason, refuse to listen.
> > Will it be worth all the effort and risk? I guess nobody can answer that.Actually, some of us can. From my POV there are those who, for whatever reason, refuse to listen.
I gather that you can answer the question, but I don't know what your specific answer is. Are you saying that you've tried similar stuff and found it to be worth the effort, or not worth the effort? I'd love to read about your trials and your impressions of the results.
What are people refusing to listen to?
Thanks,
Alan
Edits: 12/17/09
Hi Alan, My remarks are directed at the OP. He has a proven track record of trolling and causing trouble .
Yes, I have tried a lot of this stuff including the Thuneau Allocator, DEQX PDC, Behringer DCX2496 (modified and stock) and active analog crossovers. My 2c is that it can be worth the trouble but I certainly don't dismiss the notion that there is some risk involved, expecially if you have not traveled down this path before.
My recommendation would be to listen to a properly designed active loudspeaker and speak to the owner about his first-hand experiences with his system. If you do that and like what you're hearing, the advise of the owner will likely cary a lot more weight that anything that I could possibly convey to you here. Seek out an audition of the Linkwitz Orion speakers or a system that uses a DEQX PDC and let us know what you think.
Hi Mac,
I haven't tried the Behringer nor the DEQX, but I have used the Thuneau Allocator / Arbitrator extensively back in the day. When I first was using the Allocator, I modelled the crossover of one of the pair of speakers I had built (Ellis Audio 1801B), as well as use it as the crossover to the subs I had built. Then I used it to try to develop better crossovers than the stock crossovers in the 1801B.
During my use of the program, I discovered a bug in the Arbitrator that caused audible distortion under certain crossover configurations. I'm not sure if that bug was ever fixed. Another issue I ran into once was a power failure while listening to the system. The Lynx 2B sends a spike through to the system upon power down, and my amps at the time stored energy and slowly dissipated when power was cut. There went my tweeters.
I have listened to the Orion's, but with the active crossover, not a digital crossover. I thought they sounded great with the active crossover.
I was going to build a pair of Orion's at one point, but the costs were prohibitive at the time. Of course, once I could afford to build them and buy a decent amp to drive them, I didn't have the time to put into the project :( I still think I might build them at some point, but now I am quite happy with my current speakers so I have little motivation beyond the fun of experimenting.
Two question: Did you find that one of the three crossover products you listed above outperformed the others? Have you had the opportunity to compare Orion's active crossover to a crossover implemented digitally?
Thanks,
Alan
Hi Alan, Of the digital crossovers I've used, I had the best results using the DEQX PDC 2.6. The Behrnger is a good tool for simple prototyping but not something that should be used if one is expecting good sound (my impression of the unit with an upgraded output board is the same).
AFAIK, there are a few people who have tried to replicate the transfer curves for the Orion in the digital domain, but the general consensus is that the results are not as good as using Linkwitz' analog crossover. You might try searching the OUG for "digital crossover".
A huge advantage to this digital crossover concept is the potential for audiophiles to either build for ones self or have built speakers with higher quality cabinet construction and drivers for far less money than current commercially available passive crossover speakers.
If you look at the pricing of hi quality speakers that use off the shelf drivers; a good portion of the price is in the R&D and typical mark up.
Typical passive speakers:
$2000.00 cabinet + $2000.00 worth of drivers + $200.00 worth of passive crossover parts = a $16,000.00 or more set of speakers.
Compared to:
Computer + State of the Art firewire 8 channel dac $4500.00 to $6000.00 + $3500.00 for 6 channels of very good Class D amps + $4k to $6k for DIY or commissioned built speakers using similar cabinet construction and drivers as $15K to $25K commercial speakers.
Add in room correction and you have a approximately $16K total system that would be hard to beat at any price.
Upgrade path; absolutely! Start with the Lynx card and lesser quality amps, maybe $1500.00 DIY or commissioned built speakers equal to a $6K retail speaker. Easy, fun path to the above type of state of the art system.
The computations or permutations are endless; end result better sound for less money.
The piece of the equation that you (or anyone in this thread - aside from Aljordan) fail to mention is what's involved in *properly* designing a crossover network. This is *not* something that can realistically be done by ear. You need to invest in measurement equipment and really learn how to use it. Even then, it can be quite a challenge to come up with the optimum transfer curves for a multi-way speaker. The task is somewhat simplified if one is just trying to replicate a "stock" crossover in the digital domain, but even then it's not for the uninitiated.
When "audiophiles" start ruminating about "how expensive" it is to buy six "audiophile-grade" amplifiers or DACs or interconnect cables it becomes obvious to me that they are missing the big picture... and the boat. Just my 2¢.
I think the missing piece of the equation is room correction.
As far as optimum transfer curves and measurements; the required data for most drivers is available from several sources including the driver manufacturers. Most crossover software has the facility of importing this data and making use of it. I don't think it is as hard as some people think.
I have been for about a year listening to a pair of custom 2-ways with subs using some very well know drivers; used in the same combination as several highly regarded commercial designs. The crossover and room correction is provided by a complete Lyngdorf system. This combination best by a far margin a pair of speakers that I also own that are worth as much as this entire system.
The set up of the crossovers and measuring the room took 1 1/2 hours.
The difference between room correction and no room correction is immediate and profound, a major improvement.
The emergence of computer audio that allows for this type of technology at reasonable cost can be a game changer in audio if people are open minded enough to explore it.
I applaud Cics for adding VST support to his already outstanding software. I fully plan on exploring this feature in the near future with a number of different options.
You probably mean the boat you and few others are in... and I agree completely. Thanks but no thanks, as they say.
There was a time where I was in your boat. Fortunately I wasn't as close-minded as you and have since moved on. Perhaps attempting to learn or glean from others' who have experience greater than your own will allow you to do the same. In any case, good luck with all of your audiophyle pursuits. And, please excuse my ignorance in attempting to enlighten you while responding to your question.
As I stated from the onset, I don't think this technology is suited for you.
"ignorance is bliss" - you could say that one more time. Or make it a prayer - even better.
You mean, it didn't help your condition? Apparently not.
. . . *properly* designing a crossover network.I agree completely.
Obviously, I accept that active crossovers offer several significant advantages over passive ones for solid engineering reasons and that there are many excellent active speakers on the market (at a price).
What I'm less willing to accept is that a well-designed passive crossover (which takes a deal of expertise and effort to realise) is inherently inferior to a hit-and-miss effort cobbled together by an amateur like me using consumer-grade software.
I've looked at the literature on several DIY and after-market active crossovers and have yet to see a clear explanation of how best to set crossover frequencies to match the parameters of the target passive device (i.e. optimally for the given drivers in the given cabinet).
The "the devil is in the detail" of crossovers if it's anywhere. Perhaps the DIY-active approach is valid for well-known designs where the relevant data are available but, for extant DIY or less well-known models, I'm happy to trust the original designer.
If I'm talking nonsense and the problem is in fact relatively easy to solve, I'd be delighted to know.
Dave
Edits: 12/17/09
hey carcass93
I got inspired last year by dawnrazor to pursue an idea I´ve got for a longer time.
-I wanted a 3 way active setup for my magnepan MG3.3R´s (magnepan being famed for brilliant driver technology but rather poor XO´ers).
I also wanted to dig into hard-disk digital transport, so the the challenge was to implement a harddisk music server coupled with a digital XO (thuneau).
the (in my humble opinion) most audiophile hard-disk system is the cMP2 project. this advanced configuration of a windows based computer, optimizes a very long string of issues with windows and computers in general. it actually upgrades a PC to a dedicated runtime audio server and does address very serious limitation to PC audio.
combining this brilliant effort from cics (the main designer of cMP2) with the digital VST crossover from thuneau has cost me a serious amount of time and correct configuration trouble in the last year. but now cMP2 can be run directly to allocator and is a much easier setup.
I´m quite convinced that allocator is best served with the lynx TWO-B soundcard (-or a few other multiple-output cards) for best results. fiddling with an external word-clock to synchronize different DACs is costly and may not yield anything than added jitter....
I have setup a dedicated windows XP computer with all the recommended hardware from the cMP2 site and combined it with the allocator and a lynx TWO-B card. the computer runs on a dual-core E7400 CPU down-clocked to 900MHz and under-volted. totally stripped from unnecessary components and processes.
the Wav music files are read directly from RAM and processed directly from the allocator XO to the sound-card which feeds 6 channels to the speakers.
there is actually a direct link from the synchronized DACs to each power-amp.
surely I have great respect for an accidental digital malfunction, as there are absolutely NO protection circuit if anything goes tits-up ....
in this regard its a dangerous system I listen to. and EVERY configuration change is tested with the outmost of care.
nevertheless it´s the cost of driving i the formula 1......
my magnepans are controlled by digital bits in the most direct way, but I can assure you I have not regretted any minute of it. allocator and cMP2 direct to the power-amps is just ......sensational !
I was comparing a stock Behringer DCX2496 with STOCK outputs to a M-audio Revolition 7.1 24/192 capable PCI card.
I used identical transfer functions in each crossover for it to be a fair comparison (i.e. I did not use phase correction or any other features that the Behringer could not do).
I have since modded my Behringer to have passive single ended outputs which sound a whole hell of a lot better than the stock balanced outputs littered by scads of cheap op amps fed with god-knows-what quality DC. I used clarity caps for coupling caps. I also modded the digital input quite a bit. I removed the 110ohm AES/EBU transformer (which is "sorta" SPDIF compatable) and put in a proper SPDIF 75ohm terminating resistor network according to the receiver literature. I then upgraded my mod and replaced the purchased 75ohm SPDIF cable with RCA ends and installed an isolated RCA jack on the back and ran quality 75ohm coax directly to the board and terminated with a 75ohm terminating resistor. I think just routing the SPDIF input direct with a proper 75 ohm termination is better than the "through board traces and ribbon cables" foot-long path to an incorrectly sized transformer for SPDIF. Even in the Behringer literature it says that SPDIF could 'very well work' but they do not guarantee it. The Behringer digital input was designed for AES/EBU 110ohm and is NOT optimized for SPDIF. And making a "AES to SPDIF adapter" with a RCA to XLR cheater cord does not cut the mustard here. Sure it "works".... but it's not the best. I am not getting the "frying egg" problem with the CS8416 digital receiver chip like some people are getting a LOT. I think it happened ONCE and I reseated the ribbon connectors and it worked again. It was probably cycling the power that did the trick! ;)
I just might pull out the old Behringer again since I have one design which does not require variable Q crossovers and uses "textbook" filter Q's and next to no equalisation. Then I could compare the two quite easily.
There is nothing stopping Behringer users from using the Arbitrator in the signal path. In fact, you would then only need the standard 2-channel VST host plugin for Foobar / Winamp. The Arbitrator does not apply the phase correction to specific channels - it applies it to the input signal AFAIK, so this should work.
Foobar -> VST plug -> Arbitrator -> SPDIF OUT -> Behringer
In fact... Arbitrator will work with any DAC and any set of speakers. If you know the crossover points and ACOUSTIC slopes (Not electrical slopes by counting passive components) you can dial in the phase correction and make ANY SPEAKER on the planet truly transient perfect.
Cheers,
Presto
Hi,I am not familiar with that particular pair of speakers except that I've heard good things about them. So, as the speaker system was reviewed with the Behringer in the mix, the benefits of the digital crossover in that system must have overcome the drawbacks of the cheap Behringer device, otherwise the system wouldn't garner such good reviews.
I think the advantages of a digital crossover can work well in some systems, but I don't think that they can be considered a universal means of eliminating passive crossovers. When I speak of advantages, I am speaking about those systems where the digital crossover is designed into the system. There are still problems to overcome when trying to roll your own with something like the Allocator.
As I see it, the main advantage of a digital crossover is the ability to (somewhat easily and quickly compared to traditional crossover development) implement a transient perfect crossover, without the phase problems and losses of passive crossovers, as well as the ability to customise the resulting system to the room the speakers actually live in.
The disadvantages are:
- Price of good quality multi-channel amplifier system. Although, digital crossovers don't have the power losses of passive crossovers, so much less powerful amps should be able to be used.
- Having to digitise analog signals. Not a problem if you only listen to digital, but a big problem if you have a lot invested in a turntable setup.
- Sound of the various digital filters that are used for digital crossovers - preringing, postringing, linear phase, minimum phase etcetera. Some people feel the filters are too invasive.To your points specifically:
"I see many limitations to either approach, namely:
- scarsity of sound cards meeting requirements, especially in Option 2;"The Lynx 2B has some pretty good sounding on-board DACs for a reasonable price. Still, quality is expensive.
"- inferior quality of breakout boxes/cables, that one may need to deal with in both cases;"
Not all pro audio devices use breakout cables. The Prism Sound Orpheus does not use break out cables at all, but retains eight analog output channels.
"- inferior DACs in Option 1;"
Inferior to what DACs, I guess is the question. The Lynx 2B can sound quite good in the right computer. It is inferior to some DACs and superior to others. The Prism Orpheus sounds amazing, as does the Metric Halo ULN-8. I don't believe these latter two DACs are inferior to anything."- potentially sound-degrading software VC, unless DACs have adjustable gain (or preamp is used);"
Yes, there is this potential, but in my experience it depends on the software and how much you have to attenuate the signal. A good 32 or 64 bit digital volume control can be far more transparent than some of the analog preamps I've heard."- potentially high cost, especially with option 2."
Yes, with quality DACs and amps the cost is high. Do the sound of less expensive DACs and digital amps outweigh the drawbacks of a passive crossover? I suppose it depends on the abilities of the designer.
Other issues to rolling your own crossover with something like the Allocator are:- Experience needed to do it right. You just can't slap a fourth order crossover at 80 Hz and 2200 Hz and expect it to perform well. Measuring the drivers correctly is a total pain in the keister.
- Potential to very easily shatter your beautiful expensive tweeters if you let a full range signal get through to them, or if your computer sends audible spikes on bootup and shutdown. In this manner an external digital crossover is a better solution.
- Difficulty in routing signals correctly. Some cards like the Lynx 2B have good software mixer and hardware loopback capabilities to deal with this problem, but the routing is not so easy nor flexible with all sound cards / external DACs.Still, its fun to experiment with all of this stuff, particularly if you like to build your own speakers.
Alan
Edits: 12/15/09
There is option 0
Use another active crossover then the Behringer.
Bryston, DESQ make them.
What ever their qualities are, they don’t come cheap:
http://thewelltemperedcomputer.com/HW/ActiveCrossover.htm
Price.
If I look at the Orions by Linkwitz lab, it struck me that the electronics are not really expensive.
ASP/crossover assembled and tested: $1030
12 channel ATI amp: $1800
Sometimes I wonder, if a PC sets you back $500, why should a good amp using comparatively very simple tech, cost a multi fold.
http://thewelltemperedcomputer.com/HW/ActiveSpeakers.htm
The Well Tempered Computer
CMP WITH THUNEAU ALLOCATOR INTERACTION
Cics, I am a newbee,have cmp2-cplay from the very start and want to use Thuneau allocator as final touch.
What causes the problem when using CMP with ALLOCATOR (must to restart), is is Cplay or CMP.
I am asking this because if it is Cplay than maybe using FOOBAR as a player
will solve the problem
Thank you in advance
Zoran
Really like the idea of no passive x-overs. Another option not listed is using active speakers with digital input. Computer does 6 outs using Allocator and each active speaker takes 3 digital inputs (low, mid, high). No analogue I/Cs and no external amps. Speaker does D/A conversion.
Here's JMlab's SM11 L-R:
It's rear-end:
Either Digital (to 96k) or Analogue (does A/D conversion) input is used. Speaker does x-over processing in digital domain. One could bypass this by feeding each leg directly (digital or analogue). Not quite there yet.
That takes main disadvantages of active speakers - inability to invest over time and upgrade gradually - one step further. I, for one, would not be interested, especially considering that such system of relatively high quality would be simply cost-prohibitive.
That's not even touching on the subject whether speaker enclosure is appropriate place for a high-quality amp (it is not, in my and many others' opinion).
Adam makes a whole line of Ultra High-End self powered speakers.
Music is the Bridge between Heaven and Earth - 音楽は天国と地球のかけ橋
Class D amps are definitely suitable for putting into speaker enclosure. But, opinions on whether they are suitable for high-quality audio reproduction, let alone for $30,000 (or $50,000, as in ADAM's flagship model) speaker system, differ shall we say drastically.
But, opinions on whether they are suitable for high-quality audio reproduction, let alone for $30,000 (or $50,000, as in ADAM's flagship model) speaker system, differ shall we say drastically.There customers are not likely "Audiophiles" then...I would venture to say that Audiophiles actually don't enjoy music. But instead obsess about non-issues to the point of making themselves hear things that no other human can hear. In fact Audiophiles hearing is so acute that they can HEAR what they READ! Meaning if they read something has .000001 of some negative affect and another product has .00001 of that affect they can actually HEAR the difference. But they have to read it first...:-)
Audiophiles hear with their eyes and imagination. There is a reason why Expensive "LOOKING" Products sound so good.
Music is the Bridge between Heaven and Earth - 音楽は天国と地球のかけ橋
Edits: 12/17/09 12/17/09
The latest issue of VoiceCoil magazine has an interesting article on a new product from a new company that allows savvy DIYers/Audiophiles to effectively multi-amp their existing speakers or new designs. The product is not unlike the offering from Hypex in that it's a "plate amplifier" with digital crossover and amplifiers built-in. It comes with software (that presumably supports measurement capability) and allows you to emulate any number of conventional analog filter topologies (LR, Bessel, etc). The Hypex product uses their own Class-D amplifiers (which are excellent) and this new offering uses B&O Class-D amps.Edit: The name of the company is Digmoda and the product is called "D-Pro".
Edits: 12/17/09
Very nice information....Added it to my bookmarks for future reference.
Music is the Bridge between Heaven and Earth - 音楽は天国と地球のかけ橋
I do agree that a (stock or "modified") DCX2496 does not belong in any high quality audio system.
> > Thoughts?
What I will state is the obvious. Multi-amplification is not a viable solition for you.
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
