With all the buzz over the DCX going on right now, and the subsequent difficulties that have been brought to light concerning its integration with consumer gear I can't help but wondering if there is any real benifit over an analog XO such as a Rane AC-23? I mean, if all you need the unit to do is provide XO duty, where is the benifit of an extra DAC stage?
1. Is there any sonic benifit to the operation being done in the digital domain?
2. Do the analog units have the same level problems associated with them as the DCX does as far as input and output voltage issues?
3. Has anyone out there ever used an analog XO in their systems? Even better, has anyone ever AB'ed them with a digital unit?
Any answers or even comments would be greatly appreciated.
1. Is there any sonic benifit to the operation being done in the digital domain?
2. Do the analog units have the same level problems associated with them as the DCX does as far as input and output voltage issues?
3. Has anyone out there ever used an analog XO in their systems? Even better, has anyone ever AB'ed them with a digital unit?
Any answers or even comments would be greatly appreciated.
Seth,
I have both an analog x-over and the DCX2496-DEQ2496 combination that I use with my dipole system. See pictures in this thread.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=34140&perpage=15&pagenumber=3
My analog x-over is a custom job specifically designed for my system. It is based on BB2406 op-amps, - 15 chips in total. The digital x-over requires, as you noted, a more complex associated electronics, which complicates things. You need a pre-amp before it, unless you use the digital input, and you need a volume control after it. Naturally you do not have any of those issues with an analog x-over.
As for the 'sonics', I cannot tell the difference at all between the digital and analog boxes. In fact, I cannot even hear an appreciable difference regardless, if I use the digital input or the analog input with the DCX2496. It seems that the A/D converters in DCX2496 are quite good.
What you get with the DCX2496 is a tremendous adjustment capability. I am still tweaking the numbers on the filters I use with it. The problem that I encountered with the DCX-DEQ combination is that if the digital input is used, I often see that the input levels go into red, - i.e. show an overload. This is strange. In the end I will most likely stay with the digital solution.
Vadim
I have both an analog x-over and the DCX2496-DEQ2496 combination that I use with my dipole system. See pictures in this thread.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=34140&perpage=15&pagenumber=3
My analog x-over is a custom job specifically designed for my system. It is based on BB2406 op-amps, - 15 chips in total. The digital x-over requires, as you noted, a more complex associated electronics, which complicates things. You need a pre-amp before it, unless you use the digital input, and you need a volume control after it. Naturally you do not have any of those issues with an analog x-over.
As for the 'sonics', I cannot tell the difference at all between the digital and analog boxes. In fact, I cannot even hear an appreciable difference regardless, if I use the digital input or the analog input with the DCX2496. It seems that the A/D converters in DCX2496 are quite good.
What you get with the DCX2496 is a tremendous adjustment capability. I am still tweaking the numbers on the filters I use with it. The problem that I encountered with the DCX-DEQ combination is that if the digital input is used, I often see that the input levels go into red, - i.e. show an overload. This is strange. In the end I will most likely stay with the digital solution.
Vadim
Interesting
The reason I ask is that I do use a pre-amp and it only has unbalanced connections. I have just now seriously begun to look at building a bi-amped active speaker for my mains. I had originally intended to buy very large and expensive amplification for the task, but as I read more, it became apparent that that route was unnecessary and likely overkill, I could relax the amp criteria a bit and put some more money into drivers and construction/finish. As such, these amps are not balanced and so I am stuck with unbalanced pre-out and unbalanced amp in. In that situation, it seemed to me that the analog units would be easier to work with. I really like the idea of being able to adjust slope and Q, but I figure the LR4 should be fairly usefull for just about any situation.
I want to keep the pre in the chain for volume control and the simple is that these speakers must also do my HT bidding. I don't have the real estate for a pure 2 channel system yet.
The reason I ask is that I do use a pre-amp and it only has unbalanced connections. I have just now seriously begun to look at building a bi-amped active speaker for my mains. I had originally intended to buy very large and expensive amplification for the task, but as I read more, it became apparent that that route was unnecessary and likely overkill, I could relax the amp criteria a bit and put some more money into drivers and construction/finish. As such, these amps are not balanced and so I am stuck with unbalanced pre-out and unbalanced amp in. In that situation, it seemed to me that the analog units would be easier to work with. I really like the idea of being able to adjust slope and Q, but I figure the LR4 should be fairly usefull for just about any situation.
I want to keep the pre in the chain for volume control and the simple is that these speakers must also do my HT bidding. I don't have the real estate for a pure 2 channel system yet.
Seth,
You can use the unbalanced connections with DCX2496, so you are not stuck at all. Simply make your own cables, as I did, that have the XLR connections on one end and the RCA connectors on the other.
Look in the DCX manual for connection diagram and only use the (+) pin and the ground pin, living the (-) hanging. You can also short the (-) pin to ground with most likely the same results. Definitely, for the output there is no penalty in going single-ended.
If you intend to use the analog inputs with the DCX, going single ended will force the DCX to compensate internally. This compensation introduces distortion due to the fact that the op-amp generating the (-) portion of the signal inside the DCX is not the same as the op-amp driving the (+) pin from the outside the DCX. This distortion might be rather negligible and perhaps we are splitting hairs here, but theoretically it will be there. So it is better to use true balanced connection on the input of the DCX2496.
As for the power amplifiers, I know from experience that they do not matter as much as many believe. These days you can get a very clean power amp for $2-3 per Watt. With this in mind, I would suggest putting most of your resources into speakers.
Vadim
You can use the unbalanced connections with DCX2496, so you are not stuck at all. Simply make your own cables, as I did, that have the XLR connections on one end and the RCA connectors on the other.
Look in the DCX manual for connection diagram and only use the (+) pin and the ground pin, living the (-) hanging. You can also short the (-) pin to ground with most likely the same results. Definitely, for the output there is no penalty in going single-ended.
If you intend to use the analog inputs with the DCX, going single ended will force the DCX to compensate internally. This compensation introduces distortion due to the fact that the op-amp generating the (-) portion of the signal inside the DCX is not the same as the op-amp driving the (+) pin from the outside the DCX. This distortion might be rather negligible and perhaps we are splitting hairs here, but theoretically it will be there. So it is better to use true balanced connection on the input of the DCX2496.
As for the power amplifiers, I know from experience that they do not matter as much as many believe. These days you can get a very clean power amp for $2-3 per Watt. With this in mind, I would suggest putting most of your resources into speakers.
Vadim
My main concern with the DCX is the extra A/D and D/A conversion. I have read the excellent posts some have put up concerning the quality of the converters in the Behringer and it seems that my concern is seemingly unfounded, but its hard to imagine another conversion after my pre could possibly be ideal. Thus my interest in the analog units. If it really makes no difference, then its not a problem, but this report of certain software making the Behringer sound harsh really has me worried. I think we have all heard dramatic differences when different DAC containg devices are compared. Admittedly, I don't know if that is a product of the actual DAC or the output stage, but in any case, a Denon CD player doesn't sound anything like a Primare when you use the analog outs. I was worried the same will happen with this Behringer.
Seth
I just have substituted the passive XOs between my dipole subwoofers and my Fostex FF85K fullrangers with an active analog x-over, the Behringer CX2310. It´s almost the same as your Rane AC-23, less the time delays.
If you calculate the cost for two steep passive x-overs at 120 Hz, the CX2310 in comparison was almost a bargain. The CX2310 fits seamlessly between the pre- and poweramp stages of my NAD 312. Changing from balanced to unbalanced connections was no problem and the input and output attenuators are still at their 0 dB position.
The FF85K now is effectively relieved from low frequencies and I could adjust the x-over exactly to the point, where the box resonance of the w-dipoles can no longer be heard.
The Behringer CX2310 very efficiently solved my x-over requirements. But: You can´t define different slopes or frequencies for the HP and LP parts of the x-over. So your x-over points have to be well in the linear range of both drivers.
Only digital x-overs will give you full freedom and control in shaping the HP and LP contours of the filter.
I just have substituted the passive XOs between my dipole subwoofers and my Fostex FF85K fullrangers with an active analog x-over, the Behringer CX2310. It´s almost the same as your Rane AC-23, less the time delays.
If you calculate the cost for two steep passive x-overs at 120 Hz, the CX2310 in comparison was almost a bargain. The CX2310 fits seamlessly between the pre- and poweramp stages of my NAD 312. Changing from balanced to unbalanced connections was no problem and the input and output attenuators are still at their 0 dB position.
The FF85K now is effectively relieved from low frequencies and I could adjust the x-over exactly to the point, where the box resonance of the w-dipoles can no longer be heard.
The Behringer CX2310 very efficiently solved my x-over requirements. But: You can´t define different slopes or frequencies for the HP and LP parts of the x-over. So your x-over points have to be well in the linear range of both drivers.
Only digital x-overs will give you full freedom and control in shaping the HP and LP contours of the filter.
Seth,
In principle, the best A/D conversion is no conversion at all, but if you look at the specs of the A/D stage, it is evident that this box is good to 16 bits. This is an excellent result and it tells me, and my ears confirm, that the DCX2496 is quite transparent. So, I am not worried about the extra A/D stage. Naturally, I would still like to avoid it if I can get around it.
As I stated before, the use of the Digital Input often results in overloading, as per input indicators. So, I often chose to use the analog inputs and, again, I do not perceive this to be a problem.
Having said all that I still would like to modify the output structure of the DCX. I would like to do my own D/A conversion and my own filtering. However, I want to do this not because I would necessarily get a better sound, although it is certainly desirable, but because an engineer in me wants a state of the art electronics and I know how to make this DCX better.
I also cannot confirm the problem with the software Ver.1.16. I do not hear any differences vis-à-vis the Ver.1.15. I tried both and the 1.16 results in more processing power. Perhaps I should revisit Ver1.15 just to be sure.
Vadim
In principle, the best A/D conversion is no conversion at all, but if you look at the specs of the A/D stage, it is evident that this box is good to 16 bits. This is an excellent result and it tells me, and my ears confirm, that the DCX2496 is quite transparent. So, I am not worried about the extra A/D stage. Naturally, I would still like to avoid it if I can get around it.
As I stated before, the use of the Digital Input often results in overloading, as per input indicators. So, I often chose to use the analog inputs and, again, I do not perceive this to be a problem.
Having said all that I still would like to modify the output structure of the DCX. I would like to do my own D/A conversion and my own filtering. However, I want to do this not because I would necessarily get a better sound, although it is certainly desirable, but because an engineer in me wants a state of the art electronics and I know how to make this DCX better.
I also cannot confirm the problem with the software Ver.1.16. I do not hear any differences vis-à-vis the Ver.1.15. I tried both and the 1.16 results in more processing power. Perhaps I should revisit Ver1.15 just to be sure.
Vadim
Seth Smith said:
The difficulties essentially arise mostly because of consumers not being aware of level differences, etc, and a basic unfamiliarity with pro gear- more knowledgeable posters work solutions out without much in the way of problems _ grin_
I am a believer that _everything_ you put in the signal path has some effect -
Passive components have effect also. Those who say digital units add an extra ADC/DAC step if you do volume control with a preamp before the Digital crossover are correct - however, they forget or overlook how many passive components they are putting in the path with analog units.
IMHO, the tremedous flexibility of the digital crossovers tip the scales firmly in the direction of digital -
Several posters in this thread say that sonically they hear no difference between passive and Digital - I can't really speak to that because I have done no direct comparison.
It may seem like it, but it is no small matter to be able to shift your crossover point 2 or 3 hz "specifically" or to be able to add a couple of EQ points of 1.4 DB here or 2db there or whatever - nor is it a small matter to be able to select 24DB slopes or 48 DB slopes -
Along with the ability to set delay (if you need it), these and other features can combine to give you truly superb integration and flatter response.
I was one of the early adopters of a DCX2496. I did so because of the delay - It turned out a good bit better than I had hoped.
YMMV - all standard disclaimers apply _grin_
Regards
Ken L
Onto input voltage again, my chain looks like this:
Transport - Pre - DCX - Amp
Now, my volume control will be done in the pre, this will create variable voltage out to the XO. From what I understand, this could be problematic as the unit really needs the most voltage you can feed it to achieve the cleanest performance. Is this going to be a problem in my situation? I use a Rotel RSP-1066 and I'm not sure on its voltage output level. Any insight?
You guys have been great so far, please continue to pour on the info!
Transport - Pre - DCX - Amp
Now, my volume control will be done in the pre, this will create variable voltage out to the XO. From what I understand, this could be problematic as the unit really needs the most voltage you can feed it to achieve the cleanest performance. Is this going to be a problem in my situation? I use a Rotel RSP-1066 and I'm not sure on its voltage output level. Any insight?
You guys have been great so far, please continue to pour on the info!
Seth,
Considering your signal chain, you should not do the volume control as you intend in the pre-amp. Reason being is that you must ensure that the DCX gets at least 6-10 Volts peaks at the input. Adjusting this number down, as you would with a pre-amp volume control in order to maintain the desired volume level, would degrade the S/N capability of the A/D converter in the DCX.
Instead, you should use your pre-amp to set the input level into the DCX as high as it is possible without overdriving the DCX and leave the pre-amp’s volume control along at that point. Also, I do not think that the Rotel RSP-1066 is capable of delivering such a high voltage as the DCX requires for the best operation.
Certainly you can do a DIY op-amp based signal booster and not use the Rotel at all. I have an older Bryston BP-25 that I heavily modified for that task.
The volume control must come after the DCX. It would have to be a 6-channel deal. There are number of DIY possibilities, with only a few commercial offerings. Look here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=34606&highlight=
and here
http://www.decibelhifi.com.au/prod69.htm
for options.
I use motorized Alps 4-gang potentiometers for the task.
Vadim
Considering your signal chain, you should not do the volume control as you intend in the pre-amp. Reason being is that you must ensure that the DCX gets at least 6-10 Volts peaks at the input. Adjusting this number down, as you would with a pre-amp volume control in order to maintain the desired volume level, would degrade the S/N capability of the A/D converter in the DCX.
Instead, you should use your pre-amp to set the input level into the DCX as high as it is possible without overdriving the DCX and leave the pre-amp’s volume control along at that point. Also, I do not think that the Rotel RSP-1066 is capable of delivering such a high voltage as the DCX requires for the best operation.
Certainly you can do a DIY op-amp based signal booster and not use the Rotel at all. I have an older Bryston BP-25 that I heavily modified for that task.
The volume control must come after the DCX. It would have to be a 6-channel deal. There are number of DIY possibilities, with only a few commercial offerings. Look here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=34606&highlight=
and here
http://www.decibelhifi.com.au/prod69.htm
for options.
I use motorized Alps 4-gang potentiometers for the task.
Vadim
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.