Yes, I have Quad ESL57 which I intend to cut off at about 200Hz. I tried full range with the subs filling the bottom and XO at 200Hz sounds better. Subs are based on 15" vintage drivers in open baffle setup. Using DCX allows me to set the timing right and achieve very good coherence. The bass is really awesome. But DCX is extremely picky about the gain, accept SPDIF only, and has limited EQ options. As you can imagine in open baffle I need some serious compensation.
The timing issue (caused by different buffer handling mechanisms if I understand correctly) in DACs worry me a lot. If I cannot measure it then playing up with the phase may not be good enough to get great results.
I downloaded JRiver trial, but I can't find how to route the L/R channels to surround channels? Can you help?
Try this:
Tools/Options/DSPstudio& output format/Parametric equalizer, here push "add" button and select "mix channels". Use "copy and replace destination with source"
Output format is a bit strange to understand... basically it must match the number of phisical channels of your output device(sound card)
And I'm surprised that you are surprised that I am surprised - I think.
I've run the DEQ into the DCX from a digital signal for years without clipping problems. As long as you understand gain structure and the nature of the digital signal, all is good. Most people do NOT understand those things, so they run into trouble. it isn't common knowledge, it has to be learned.
Running the DEQ and DCX in series with a digital input and never clipping, in no way gave me less dynamic sound in the end. Sure, S/N ratio goes down, but the devices have far more dynamic range than I'll ever need. And it's even better with modified outputs.
If you want to do the same thing in software, you will have the exact same problems, unless you understand gain structure. There are plenty of guys here who can help with that.
Yes, you do. But there are some non-intuitive ways to do it. In the beginning, most people want to boost all the places that the bass is lacking in OB. From there you might try shelf filters and broad parametric. I've been down that road with the DCX and Open Baffle.
But you can try a more "passive" approach with good results. Think backwards. Does open baffle have a falling bass response or a rising mid response? Depends on your point of view. If you leave the low end were it is and fight the rise, you can usually achieve better results.
Do you need to cross your OB subs at 200Hz? Try a 1st order LP at about 50Hz. Listen to what that gives you. Bring up the subwoofer amp gain so that the low end of the subs matches the the ESL panels. How good is the blend? Roll the 1st order filter up and down the frequencies a bit and listen. Don't think about numbers, just listen. You will probably find a good blend. Once there, listen for a frequency that might stick out, and use the EQ to tame it. You shouldn't need more than about 1 or 2. After you've fixed those, you may need to tweak the crossover point a bit.
I am using 2 channel inside 7.1 container.
I asked the Q on JRiver board and am already doing the mix and copy thing, and still not getting the same output from surround and rear channels.
Strange!?!?
zmyrna,
Could you go into a little more detail on the hook up for a two way active connection plz?
Doug
Could you go into a little more detail on the hook up for a two way active connection plz?
Doug
the set up I am working on aims at reducing the number of cables/connections and box units.
I am trying to connect a source through HDMI to an AV receiver which then powers active stereo multiway speakers.
7.1 AVR can do a 3 way system,
5.1 AVR can do a 2 way, with or without subs.
I was going to use the nanoAVR box, but now I found out that I can do it on a computer and avoid the extra box btw the source and the AVR. This would work great as long as you are okay to use your computer as your sole source.
On the computer DSP (JRiver), you route the front L/R signal to the surround and rear channels too so that all channels on one side plays the same signal. Then you connect your AVR such that for example left front goes to the left tweeter and left surround goes to the left midwoofer. Then you simply apply your crossover filters and EQ to each channel on your computer DSP.
One caveat I just found is that your AVR should have a processing free surround mode which is called sth like direct / pure audio passthrough.
I am trying to connect a source through HDMI to an AV receiver which then powers active stereo multiway speakers.
7.1 AVR can do a 3 way system,
5.1 AVR can do a 2 way, with or without subs.
I was going to use the nanoAVR box, but now I found out that I can do it on a computer and avoid the extra box btw the source and the AVR. This would work great as long as you are okay to use your computer as your sole source.
On the computer DSP (JRiver), you route the front L/R signal to the surround and rear channels too so that all channels on one side plays the same signal. Then you connect your AVR such that for example left front goes to the left tweeter and left surround goes to the left midwoofer. Then you simply apply your crossover filters and EQ to each channel on your computer DSP.
One caveat I just found is that your AVR should have a processing free surround mode which is called sth like direct / pure audio passthrough.
You are implying that I don't understand the gain structure where I think the fundamental issue is that we do not agree on the fact that sound quality goes down when you lower the gain in Behringer tools. I hear it and I am not happy, you don't hear it and you are fine.
My setup is a bit tricky because I do not use any opamps or transformers in output stage of DCX so my output signal is very low. On top of that I use class D amps which require signal at certain level which additionally complicates things. My DCX is fed from Squeezebox which must run at 100% volume for quality reasons. It gives me digital signal on the input of DCX at about -10dB which is high. Without any extra gain or EQ applied to output channels some albums reach clipping level. My Quads to play relatively loud need my amp's pot to be on 90%. That is why I do not want this setup to stay even though I like the sound. The easiest way would be to add trafos or tubes to the output stage of DCX but I would prefer to avoid this cost.
With regards to your suggestion to attenuate instead of boost, I had pretty good results doing that when I used OB with vintage Saba drivers. They were very easy to drive. Quad ESL is a totally different beast. They are very quiet and lowering the gain is unfortunately not an option here.
Last but not least I have a great DAC (HQ R) which I would like to use for my mains. It can send signal to XLR and RCA output at the same time so using it to play mains full range with subs filling the bottom is not a problem....but XO at 200Hz sounds much better. The ESL become much more universal speakers and even rock sounds great. Unfortunately I do not know how to make it work so that I can use Jriver or any other player as a crossover with DSP and use two separate DACs...
If your nominal input is at -10dbFS and you apply no extra gain or EQ then it would be essentially impossible to create a clipping condition within the DCX. Your statement implies there are peaks within your digital data stream that are in excess of 0dbFS. Is that what you're saying?
I find myself in agreement with Pano here. Your statements regarding the DCX clipping don't make a lot of sense.
Regardless, back to your original query in the thread title. The answer is "yes."
Dave.
I find myself in agreement with Pano here. Your statements regarding the DCX clipping don't make a lot of sense.
Regardless, back to your original query in the thread title. The answer is "yes."
Dave.
Last edited:
Your statement implies there are peaks within your digital data stream that are in excess of 0dbFS. Is that what you're saying?
Regardless, back to your original query in the thread title. The answer is "yes."
Dave.
Re DCX, I'm not going to argue about it anymore, not here anyway. You can read this thread where many people reported the same issue even when specifically used "Bypass" option:
http://www.diyaudio.com/forums/digital-line-level/15943-behringer-dcx2496-digital-x-over-309.html
Yes, let's get back to the original question. Software crossovers work, but will they work with two separate DAC?
Yes, I'm well aware of the other thread and the various idiosyncrasies of the DCX2496. You're talking to one of the experts on it.
Regardless, I agree with some of the posters that JRiver is a good approach. You can indeed steer a multi-channel data stream to disparate external DAC's, if you so choose. Obviously, you need the hardware configuration to support that.
Personally, I like the autonomous stand-alone approach from units like the nanoDIGI unit from miniDSP, but there are other options as well.
Cheers,
Dave.
Regardless, I agree with some of the posters that JRiver is a good approach. You can indeed steer a multi-channel data stream to disparate external DAC's, if you so choose. Obviously, you need the hardware configuration to support that.
Personally, I like the autonomous stand-alone approach from units like the nanoDIGI unit from miniDSP, but there are other options as well.
Cheers,
Dave.
I looked at nanoDigi (and Najda) but it still sends digital signal out and 2 DACs wouldn't be in sync. It would make it easier to measure the misalignment though.
Two identical DAC's would be in sync with the nanoDIGI unit. (one of the many aspects I've tested on that unit.)
However, two different DAC's....especially if utilizing ASRC's....might not be in sync. However, that issue would exist regardless of the type of "software-based" platform you choose. Any of the applications you might consider will allow to alter the relative time delays to adjust for it. It gets a bit trickier when dealing with ASRC's and varying sampling rates though.
Cheers,
Dave.
However, two different DAC's....especially if utilizing ASRC's....might not be in sync. However, that issue would exist regardless of the type of "software-based" platform you choose. Any of the applications you might consider will allow to alter the relative time delays to adjust for it. It gets a bit trickier when dealing with ASRC's and varying sampling rates though.
Cheers,
Dave.
baMarek. Just trying to help. The problems you describe are exactly gain structure problems and need to be addressed in whatever digital crossover you are using. They are not unique to the DCX. Just wanted to be sure you understand that and that others reading the thread also understand it. You have other gain problems outside the DCX, too.
I do know that the ESL57s need a healthy signal, I have direct experience with the 57s. (mostly with double stacked tho, which is a lot louder at the same voltage). I also used to run Squeezebox->DEQ->DCX-Class-D and tube. My DCX does have transformer outputs, which gives 6dB over the single ended outs. And less noise.
Since the large majority of recorded tracks hit 0dB on the peaks, that leaves no headroom for EQ unless you attenuate the incoming signal. That's true no matter what sort of digital EQ you use - hardware or software. For that reason out of JRiver I normally run -10dB to -12dB which leaves enough headroom for work in the DEQ and DCX. Sending 24 bit out with 10dB attenuation still leaves me with almost 40dB more dynamic range than the 16 bit recordings. So no worries. And no clipping. I do try to keep the levels up tho, just to optimize gain.
As far as software driven crossovers, I've only ever done that with a multi-channel DAC, so can't be of any help there. My attempts to used 2 different soundcards for measurements have failed badly. Certainly someone here can make it work.
I do know that the ESL57s need a healthy signal, I have direct experience with the 57s. (mostly with double stacked tho, which is a lot louder at the same voltage). I also used to run Squeezebox->DEQ->DCX-Class-D and tube. My DCX does have transformer outputs, which gives 6dB over the single ended outs. And less noise.
Since the large majority of recorded tracks hit 0dB on the peaks, that leaves no headroom for EQ unless you attenuate the incoming signal. That's true no matter what sort of digital EQ you use - hardware or software. For that reason out of JRiver I normally run -10dB to -12dB which leaves enough headroom for work in the DEQ and DCX. Sending 24 bit out with 10dB attenuation still leaves me with almost 40dB more dynamic range than the 16 bit recordings. So no worries. And no clipping. I do try to keep the levels up tho, just to optimize gain.
As far as software driven crossovers, I've only ever done that with a multi-channel DAC, so can't be of any help there. My attempts to used 2 different soundcards for measurements have failed badly. Certainly someone here can make it work.
Two identical DAC's would be in sync with the nanoDIGI unit. (one of the many aspects I've tested on that unit.)
However, two different DAC's....especially if utilizing ASRC's....might not be in sync. However, that issue would exist regardless of the type of "software-based" platform you choose. Any of the applications you might consider will allow to alter the relative time delays to adjust for it. It gets a bit trickier when dealing with ASRC's and varying sampling rates though.
Cheers,
Dave.
I have seen reports (can’t find it now) where someone had timing issue using the same DACs when playing hi-res files but maybe he was just unlucky.
Question, if you have the same DACs and timing is not a problem, wouldn’t you prefer to use software XO instead of extra gear like nanoDIGI in the path?
Marek
Pano, I appreciate your help, hoenstly. All I was trying to say is that I realise I have gain issues also outside DCX but DCX with striped off output stage is far from perfect in my setup. That is why I am looking at software driven XO before I start looking to replace DCX with fully digital XO of some sort. If software is not an option for different DACs then I would probably go for two identical DACs in order to have the bass as fantastic as it is at the moment. What kind of problems did you have when using 2 different soundcards? 2 USB DACs would effectively act as 2 separate soundcards.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.