Behringer DCX2496 digital X-over

Jump to Latest
E

ergo

Member
Joined 2001
Some "good news" and interesting reading / studying material uploaded :)

Ergo's STOCK versus MODIFIED DCX2496 battle results

Perhaps this image is good for a teaser picture

jitter-detail.png


Regards,
Ergo
 
O

oettle

Member
Joined 2006
Hello Ergo,

Well done! Your various measurements are very interesting and now it’s for sure: Modding isn’t placebo! ;)
I’m also glad to see that I can still trust my ears although my brain sometimes is misinterpreting them.

I’m pretty surprised about how poor the jitter performance of the original DCX master clock (build with 24.576 MHz crystal and 74HCU04) is.

The only disadvantage is that headroom for further improvements becomes smaller. Perhaps AK4396 DACs or power supply or …?

Kind regards, Frank
 
E

ergo

Member
Joined 2001
I agree that we are getting closer to the technical potential this unit has. But that is at least for me only electronics side.

My personal roadmap for near future is to continue "voicing" my new speakers and especially speaker room interaction. And that is where the DCX is very useful.

I also bought Thuneau Frequency Allocator when it was on special price and I plan to test it out first using only the Phase Arbitrator part in PC and directing the SPDIF feed for the DCX for final crossover. I want to see if and what can be gained there.

I have been thinking a lot lately about subjective listening of my speakers and measurements in room, how and where those correlate etc. Too early to talk about yet, but if some of the ideas work out I will be posting it on my page in due time.

One thing I can not stress enough - it is real fun to have DCX and LspCad to play around with. The easiness and possibilities to try out are endless. I'm learing a lot almost every day.

Ergo
 
O

oettle

Member
Joined 2006
Many differences can be clearly heard, .... even when there are no differences! The placebo effect is too powerful to stop by force of will. If you gained an impression of clear improvements by listening to your dcx, modifying it with soldering iron, then listening to it again, that conclusion has to be ignored. Get an unmodified dcx, put the 2 side by side, and do some simple blind testing of friends, 10 sec blocks, quick switching, many tests, document all results, analyse stats. Ahh, too hard.
Click to expand...

Much simpler and more interesting would be posted waveform images or actual wav files of the analog signal exiting the dcx2496, before and after mods.
Click to expand...

I know you do not like modding under any circumstances.
I fear the only proof for you would be to hear the difference yourself. So just test it!

BTW: I lived in Sydney for about a year. Very friendly people there in Australia.

Frank
 
P

Pulse-R

Member
Joined 2005
tnargs said:

There is absolutely no connection between a few test measurements and the conclusion that "modding isn't placebo"! Do you think that amplifier measurements are proof that they sound different?
Click to expand...

I believe that if I can hear the difference, then its better.

Whether or not you can measure the difference I think relies a lot upon how sensitive your measurements, and exactly what you measure.

when will a full scientific study be done which determines how we hear, and why it sounds different? There have been many studies, and every aspect of sound reproduction (including aesthetics and price) has an impact on how we perceive sound.

p.s. I know that amplifiers sound different, otherwise we would all use $20 amplifiers to drive our (usually) expensive loudspeakers.
 
E

ergo

Member
Joined 2001
Do you guys know this software

Audio DiffMaker

It is a very cool freeware program that allows making a wav file recording from device before and after modification. Then it aligns the samples in time and in level and makes a differential signal wav file. So one can try to evaluate the magnitude of change a single capacitor, or opamp, etc makes to the sound. The difference track can be listened later (and amplified if the difference is very small).

I tried to use it for DCX testing also - but something did not work out. Either the additional processing in DCX influences enough to disturb the calculation or I overlooked something.

Anyway if any of you guys have a moment to mess around with that tool it would be interesting to hear for exaple the difference of signals before DCX and after DCX (DCX set to flat mode of course).

That method of analysis should be accepted even by people who believe any measurement is useless.

Ergo
 
O

oettle

Member
Joined 2006
Hi jimluu,

The volume control (preamp) should be at the end of the row between DCX and amps otherwise you reduce resolution. For best sonically quality my recommendation would be:

CD -digital-> DCX + volume control -> amps

The DCX has enough EQs (about 3 per output channel) when using 48dB crossover. If that's not enough you can use 24dB crossover to get more processing power for further EQs or add a DEQ. My experience is that for typical drivers 1-3 EQs are sufficient.

For volume control have a look a t Jans website: http://www.linearaudio.nl/

Frank
 
jeff mai

jeff mai

Member
Joined 2002
Here's a new one (maybe.) I've just tracked a problem to the digital interface between my computer and the DCX. It'll be playing along fine and then, preceded by a small tick noise, the sound stage disappears and the high frequencies disappear. Tonight, I finally resolved it by removing and reconnecting the digital input cable. It's been happening for months.

I've heard of some issues with the digital input. Is this symptomatic of the problem? I'm using a transformer with the correct impedance matching, but I don't think the chip driving the transformer is putting out a proper AES/EBU voltage which is a lot higher than SPDIF.

Thanks in advance for replies.
 
O

oettle

Member
Joined 2006
Hello jeff mai,

There is a well known problem with the CS8420 sample rate converter which causes sometimes a 10 kHz cut off (dull sound). This error is described in the Cirrus data sheet on page 49. I already contacted Behringer but you won’t get help.

There is a mod which fixes not only this problem but improves jitter performance and quality of sample rate conversion. Please see posts 1419, 1389 and 1521.

You can get this mod here: http://www.pilghamaudio.com/index.asp?pgid=7

Regards, Frank
 
You must log in or register to reply here.
Top Bottom