Evaluation of the DCX2496


Hello,
I like to give you the results of my evaluation for the DCX2496 as digital crossover.

Until now my system was a "classical" system with Sony DVP-NS 930 SACD-Drive, preamp and stereo poweramp both from ALPS. The speakers I use are Quadral Vulkan (with the legendary Technics ribbon tweeters) with built-in passive crossover (standard) and a terminal with direct access to each of the three chassis. All that did not sound bad at all but I wanted to do something more sophisticated to improve the sound of my chain.

In Stereoplay I read an article about a speaker system for studios with active amplification for each chassis and a digital crossover. The test sounds very very good but the speakers were priced in the 20.000 Euro region which was too much for me. So I decided to activate my Quadral Vulkans by using one power amp for each chassis and a digital crossover.

At first step I tested the DCX2496 (great price) in my current chain to decide if I could use it or not. My ALPS preamp has the possibility to insert devices in the chain. I took the DCX2494 and inserted it (all filters off) to find out how it influences the sound. With the ALPS preamp I have the possibility to enable and disable the inserted device so I really could do an A-B-test. I tested with a lot of different CDs and SACDs. What I did was a real blind test. I let my wife switch the inserted DCX2496 in and out and I tried to identify the state only by listening. Believe me or not, there was no audible difference between DCX2496 in or out. I just had no chance to identify the difference. Even with SACDs, where the difference to CDs is clearly audible, I could not hear any difference regarding the DCX2496. This result is really encouraging because with the DCX2496 there is a A/D and D/A converter in the chain.

So I decided to use this device as my new digital crossover. Next I looked for a power amplifier. Soon I buried my thoughts of using a digital amp because the cost was too high for me and diy seems too complicated and unsafe. I decided to buy the USHER Reference 6 (list price 2900 Euro). This amp has six separate (analog) amplifiers with 125 watts each. I use one amp for each chassis in the Vulkans. Right before the USHER I set up a 8-channel passive volume control from THEL (CP-2500, 350 Euro) which I put in a 19' 1HE enclosure (http://www.thel-audioworld.de/bauteile/regler/Potis.htm). Currently I use only six channels of the volume control. I connected the passive volume control with low capacity wires from Profigold (PGA 4201) to the power amp. Right before the volume control I inserted the DCX2496. The input of the DCX2496 I connected via digital link to my CD-player. The quality of the CD-player does not matter because the DCX2496 does reclocking (44.1Khz in, 96Khz out) and so jitter is no problem anymore!

Before testing the whole thing I repeated the above described insertion test with the USHER power amp. This sounds better (Oh I love the smooth heights of the USHER, it sounds like a tube amp) but even here I could not hear any difference between DCX2496 in or out.

Now I need a method to do A-B comparison between my old system and the new active concept. Therefore I installed a pack of 12A relays in the back of my Vulkans , allowing me to switch between the internal passive crossover which was fed by the ALPS amp chain and the USHER with the DCX2496.

Ok, but how to set up the DCX2496? First guess: use 180Hz and 5Khz as cut-off frequencies and 48db/oct Linkwitz-Riley Crossovers. Oh, it sounded so horrible!!!! All I love in my chain was gone. I adjusted the DCX2496 played around but with no success. The more I played the worst it got.

Next I turned my living room in an audio lab (take care of your wife, she might not love it at all!). Using my PC with it's M-Audio 24/96 audio card I measured the frequency response of the built-in passive crossover directly at the speaker chassis (by wire, not by microphone). So I got an impression how the digital crossover has to be tailored. I found out Quadral used 12db/oct filters and a lot of other components to equalize the frequency response. Well, the DCX2496 has built-in equalizers for each speaker channel which I tailored the way I got exactly the same frequency response as the passive crossover. This was a lot of work!!

The A-B-test after this procedure was much better. The activated speakers sound as I expected, nearly the same as with the passive crossover. But there was a difference! The bass was more tight, the heights more brilliant. I could clearly hear the improvement of the sound. I could not tell if this comes from the USHER or the active concept. Well what else could be done?

Well, the DCX2496 gives me the ability of time correction. Its setup is quite simple. I used a measurement microphone from Behringer, connected it to the DCX2496 and let it run. Half a minute later the delays were corrected and I was ready for my next A-B-test.

WOW!!! I could not believe it. Imagine the difference you get if you switch from mono to stereo. This is comparable with the difference you get if you switch from my old system to the activated system with corrected delay. The stage opens widely. You hear each instrument exactly at the position it belongs to. All sounds so open and so transparent, just fantastic. The system really cries for more volume. Increasing the volume you never get the impression that it is too much. The sound stays open and clear and gives a lot of fun! I never got tired listening at high volumes (Hope you have prepared a cinema ticket for your wife. Give it to her now!)

With that now much tighter bass I had the expression there were some room resonances in the bass region. Again I took the microphone and made a mls-measurement of the frequency response at my listening position. I did two measurements one for the right channel, one for the left, imported the data to MS-Excel and visualized it on a diagram. I was not astonished to see that there were differences between the right and the left channel in the bass region. This depends of the placement of the speakers in my living room. I also found two peaks (more than 15db) one at 151Hz and one at 58Hz. I programmed the parametric equalizers of the DCX2496 to suppress these two peaks. On the frequency curve now the response was quite flat but the A-B-test was frustrating. The whole warmth of the bass was gone. The sound was synthetically. I decided to find out the right suppression value by carefully listening and experimenting with different levels of suppression. This is the point where the whole thing starts to become subjective. I decided to stop at a suppression value of -6db. This fits my taste of a nice bass.

So what did I learn?

I learned that the impression of "room" is a function of correct phase (delay) adjustment. I believe this is also the reason why SACDs give a better room impression than CDs. Because of their higher time resolution they can better reproduce the phase correlation and the better the phase correlation the better and precise is the room impression.

I also learned that the absence of a passive crossover with all its damping elements like coils, capacitors and resistors gives a more tight compact and precise sound reproduction. Now I will smile about speakers with socalled bi-wiring/amping terminals. In those speakers there are still the passive elements between the chassis and the amplifier and the advantage of direct coupling is gone.

In respect of the DCX2496 I learned that a digital crossover is always worth the effort. For me there is no audible difference with DCX2494 in and out. This means to me the device is absolutely ok. I did not need more expansive hardware if I could no more hear the difference.