Behringer DCX2496 digital X-over

AX tech editor
Joined 2002
Paid Member
Re: Bare feet runnin'

KBK said:
I don't have a schematic, I do what I have always done.... by the time I've figured out how to keep the unit operational, while being fully modified, is the time I have finally figured out how it is constructed. [snip]
I now have the RANE 26X multi-processor/computer controlled unit in for eval from the distributor. I will do a stripping of the unit to see what is inside, and then compare the two. A stock Rane against the modded Behringer. Rane makes quite a bit about it's in-house digital audio experts (at least the way things were voiced to me) and algorithms being critical, not the 96khz vs the Rane's 48khz. Due to the 96khz I/O, I believe the modded Behringer will show as being better overall, but we shall see. I have to make cables for the Rane, as it is bare wire connections. It is 8 in, 8 or 12 out, I believe. Some crazy number like that.

Yes, tracing is the only option here. I was being lazy and hoped someone had done and documented it. Oh well.

I am very curious as to your comparison results, will follow your posts with interests. By the way, did you replace the smps in the Behringer with a linear supply? If so, what were the improvements?

Jan Didden
 
ALERT: Offset

Hi folks,

Today I changed the amp that I was using after the DC2496. This amp is based in Gainclones (6 of them) with no capacitors in the signal path. That means the amps are DC coupled. I was measuring the amps when I noticed I had 210 mv DC offset at the output of the amplifiers. So I went to the input connector and I discovered that my Behringer unit has 10 mv DC offset, so If anybody is planning to connect the crosover with a DC coupled amp, beware.
I wonder if this is something particular to my unit or your units measure the same.

Cheers

Ric
 
I'm also waiting my DCX2496..... and letting my mind wonder for possible upgarde and tweaking scenarios even before seeing the unit :)

Anyway, as I plan to mostly use the cd player as a source I'm stuck with 44.1/16 bit signal and of course it puts one thinking how much would there be room of improvement if I could upsample (i mean really upsample not just oversample) the signal to 96/24.

I'm reading the datasheet for new BB SRC4192 and it states that the output signal is dithered from internal 28 bit signal. So if one would use this IC between CS8420 and the DSP and so feed the DSP with 96/24 signal...... what you guys think - at least to me this seems to be worthwhile to try. Especially if I can realise my other idea to use for example PCM1704 DACs as DAC part later on.

Or do you know any other real upsampler suitable for this application. Preferably DIY kind. Something like Tjoeb one?

Ergo
 
According to specification sheet the system sampling frequency is 96kHz, which means that indeed most likely the CS8420 always puts the information out in 96kHz sampling format but I'm afraid that if I feed the digital input with 16 bit data then the data goes through the whole chain also with only 16 bit precision (24 bit format but 8 bits unused).

I would like to find a solution that would really upsample the data (using interpolation) so that the data is feed to DSP in 24 bit precision......

Ergo
 
resolution

I don't think you'd loose any resolution going digital 16 bits.
You could try this: go digital into the unit with a signal at 0dbfs/16 bits. If the input of the Behringer shows 0db, then where's the loss? I keep the same 16 bits of resolution that I'm injecting. Notice that I say: keep the 16 bits, not "inventing" more resolution like 24 bits, because the original signal only have 16 bits.

The processing inside the unit is something different, I'll be performed at the resolution of the unit.

I'm right on this?
Ric
 
If you upsample to 96KHz you also need to add bits to avoid loosing resolution. My guess is that the unit upsamples to 24/96, thus "inventing bits" which is would have to. Any processing done on the resulting upsampled signal will avoid cutting into the original 16 bits.

Try this at home: Take regular 44.1/16 bit signal. Do processing on it in the 16 bit domain. Listen to result. It sounds awful.

Now do the same thing, but do all calculations for desired processing (such as for example room equalization) at 44.1/24 bits and burn the resulting 16 bits back to CD. Still sounds crap.

Petter
 
DEQ2496

Received mine to convert analog to AES/EBU for testing some boards. I have three Audio Precisions in my lab. THD+N is in the area of .005%. Am very happy with the performance and sound quality.

Tho SHARC AD21065 DSPs in it as well as mostly AKM A/D and D/A converters, far better than consumer gear uses. Loads of features too.
 
Anyway...

The more I think about how the unit works internally, the more I realize I do not have a clue about it.
The only thing I'm completely sure is that I'm using it alongside pro gear that cost +30K and it is sounding ok. Could it be improved? You bet. But in the mean time is working just right.
Today I put capacitors between my gainclones and the 2496 to take care of the 10 mv offset at the outputs. Also I'm working into band-limiting each band with R-C networks and that is sounding outstanding.
I think is key here to manage the levels right, trying to use every bit of resolution. I'm also attenuating the signal directly at the input of the amplifiers -hitting the input of the crossover as hot as the headroom lets me-and that helps too.

Ric
 
* First I would like to apologize to the moderator, I've posted a new thread because I've only paid attention to the date on the first post and thaught that this is an old thread and didn't want to bump it.

So I'm posting it here:

Hi folks,

I want to build a 2-way speaker (4 ohms 5/5" can speak revelator and morel supreme) with a sub (shiva) and do it actively.
I've found that the Behringer DCX2496 might be a good unit in order to do so.
I've read this long 9 pages thread and to tell you the truth some of the arguments there got me more confused.
My source will be an external sound card that also has SPDIF . My first question is will I be able to control the volume while using the SPDIF out on my sound card?
If not than I'll just use analog.

Is there a reason I'll have a problem connecting power amplifiers to it? I read that it has balanced outputs, will this be a problem?
The amp for the tweeters will be the RB-03 will it be okay?

Also I understand that an RTA can be helpfull, can you parhaps recommend on a pc based one (and recommend on a nice mic) or maybe tuning by ear can be good enough?
Will the behringer 8094 (If I remmember correctly) will be a better choice as an RTA and also is the auto mode (when it tunes itself) any good?
I don't like that fact that I'll be paying for the EQ while I'll have one in the DCX, but maybe if it's really worth it I'll consider it.

(I come from car audio and RTA in a car is preety much useless so we do most tuning by ear).

To sum it up,will it be a problem to hook it up? I want to pay attention to tweaking and tuning the xover, EQ and so on and not tweaking the thing to actually work.

Thanks!

Eyal
 
Behringer DCX 2496 Help!

Hi am new to this forum. I'm just about deciding wether Y/N to buy this digital Xover. I a not trained nor very experienced in electronics but have some experience in DIY audio mainly speakers.
Can anyone recommend the Behringer when I do not want to spend months tuning it. Want to start with axctive bi-amping of JBL ti10k speakers (4way Xover 250Hz-1kHz-4kHz) cirrently 2nd order passive standard Xovers) with 2 Rotel amps. Later maybe go to tri-amp.
Is it essential to start room-measurements before (to choose filter type, slopes etc) and after or can you get quite far from more settings and some adjustment 'by the ear'?
Don't want to start upgrading the thing except maybe PSU so can anyon recommend it as fair quality as it is? Or would i be better of building an analogue Xover like the ESP 24dB one?
Thanks for your expert advice! (janneman?)
 
Petter: yes 3way application

Hi thanks 4 reply. yes I know the Behringer allows 3-way stereo not 4 way.
I plan to start 2-way (1 amp to parallel twin woofers) and 1 amp for Low midrange-Mid-High. Have it hooked up now in this way but passively which of course is not 'the real thing'.
First I would like to gather some experience 2-way then move to 3 way - am in doubt if best would be:
1=woofers
2= low midrange
3=Mid + high

or
1= woofers
2=low mid + mid
3=tweeter
Current passive Xover points are 250 Hz-1kHz-4kHz 2nd order (the standard ones)
I guess the tweeter will need some DC protection if it would be the 3rd channel with no passive filter in between?
 
While people are in the answering questions mode;)

Has anybody direct experience wrt using the Behringer instead of a dedicated active x/o made with opa2134 op-amps, wima caps, halco resistors type of thing ?

I'm about to start building such a thing, and would go for the behringer if people think it would sound better. (note sound better, not easier to get results etc....)

Cheers

Rob
 
I do crossover design work and this is the first time I have heard a digital crossover. The lack of phase distortion throughout the crossover region, and freedom from having to lock your head in a vise (these are MTM designs usually!) is very, very nice. I of course have completely rebuilt the unit. Almost. Sort of. Mostly.

The Rane digital unit (RPM26Z DSP Multi Pro) has arrived, and now I have to give it a shot as well.
 
A theoretical answer to the previous post:

If you use digital sources, there are several benefits to using a digital crossover:

1. More flexibility, especially wrt. time delays
2. Easier to change
3. Likely higher performance than passive solutions, particularly at low frequency where series inductance of bass element required in passive crossover reduces the damping factor of that element and does harmful things in general.
4. If you split up the spectrum there is less energy in each band than the full signal. Thus, you can and should apply a digital gain to each portion to maximize the ability of the DAC chip used. This should be highly effective in improving sound quality other things being equal. It should come as no surprise that using 3 dac chips instead allows one to achieve theroretical and hopefully practical gains.

The final results will depend on the execution of both methods. I suspect that you will have a hard time beating a digital setup if using digital sources. You might want to spend your design time modifying and upgrading a digital setup instead of building a passive system.

Petter
 
Petter said:
A theoretical answer to the previous post:

If you use digital sources, there are several benefits to using a digital crossover:

1. More flexibility, especially wrt. time delays
2. Easier to change
3. Likely higher performance than passive solutions, particularly at low frequency where series inductance of bass element required in passive crossover reduces the damping factor of that element and does harmful things in general.
4. If you split up the spectrum there is less energy in each band than the full signal. Thus, you can and should apply a digital gain to each portion to maximize the ability of the DAC chip used. This should be highly effective in improving sound quality other things being equal. It should come as no surprise that using 3 dac chips instead allows one to achieve theroretical and hopefully practical gains.

The final results will depend on the execution of both methods. I suspect that you will have a hard time beating a digital setup if using digital sources. You might want to spend your design time modifying and upgrading a digital setup instead of building a passive system.

Petter

Hmmm,

That's the thing see - I'm using vinyl as a source and I'm a bit scared of putting it into digital stuff.- Hence my q's about sound quality. I'd rather stick it out and build an analogue x/o if it will cause less 'harm' than turning it digital.

Cheers

Rob.
 
A real life answer

RobWells said:


I'd rather stick it out and build an analogue x/o if it will cause less 'harm' than turning it digital.

Use of digital crossovers for home systems is still in it's infancy and having growing pains.

There are Pro units out there that you will not have to worry about causing _harm_ to your signal, like anything else, price is an important factor and not necessarily linear to benefits _grin_

If you _need_ any delay - the benefits will outweigh the negatives. If you will benefit from EQ, depending on how much, you likely will be better off with a digital crossover.

I'm using an unmodified Behringer to crossover to active subs. I was quite surprised to see how much difference I could hear between even 2 hz change in crossover point. Ditto for EQ of 1.5 DB in one place and 2 DB in another.

Would any of this mean as much to you as is does to me? Can't say. Even if our systems, rooms, etc. were the same - we're all going to hear it differently and have different opinions.

Can I and others hear the Behringer? Don't know - my system is in transition and my preamp isn't debugged - thankfully, I hope I'll be listening with a new pre this weekend.

It's hard to verbalize how much something improves when integration is seamless, time aligned and tweaked with moderate EQ - The improvement is so good that I will accept _some_ hopefully slight loss of clarity because the overall effect is more magical. Trying to describe the effect of these things is somewhat like trying to describe "soundstage" - get it all tweaked right and it easier, smoother, creamier and more " magical" - at least it is to me. It's like it removes a harshness you were unaware of until it was gone.

If you try it, you may well find unexpected benefits as I did.

FWIW, I'm using the analogue inputs for volume and although I'm using a digital source am finishing up a tube preamp to use with the Behringer.

Hopefully, these guys will have the mods figured out by the time I am able to get round to it _another grin_

YMMV

Regards

Ken L
 
Thanks Ken, a thought provoking post.

I've got an ultracurve here, and can hear the 'harm' that puts into my signal ;) If I turned down the source I'd get a 'digital buzz' type effect aswell. I guess thats why everyone is intending to mod the unit without hearing it first. I think the x/o has a higher sampling rate ?


""FWIW, I'm using the analogue inputs for volume and although I'm using a digital source am finishing up a tube preamp to use with the Behringer.""

I started looking at 6- gang pots to attenuate after the unit, to ensure the fullest signal into the unit. (place a source selector before the unit, attenuate after)

Cheers

Rob