There might be a bit of a language issue, but that is not how I read it. It is a pity that marko has chosen not to clarify what his intent actually is. I am more than happy to be corrected if I misinterpreted the posting.
Double Blind Testing is useless for audio, many an audiophile guru has pointed out that it is pointless using a test that shows there are actually no differences to be heard and the sighted view that the sound had changed was just expectation bias and perceptions being fooled.
Only long term sighted testing is of any use, confirmation of changes in the sound often being confirmed by a non audiophile, I abuse my wife in the kitchen.
Only long term sighted testing is of any use, confirmation of changes in the sound often being confirmed by a non audiophile, I abuse my wife in the kitchen.
My only intent was to share my personal experience of modifying the DCX.
I got out of line trying to make people check their sources of information and I regret that.
We have a language barrier, even thinking in english doesn't do justice to expressing my opinions here.
You are responsible to choose if my opinion matters to you and if it doesn't please move on.
I'd also like the moderators to transfer the posts about double blind testing to a new thread.
And that's not nice!
I got out of line trying to make people check their sources of information and I regret that.
We have a language barrier, even thinking in english doesn't do justice to expressing my opinions here.
You are responsible to choose if my opinion matters to you and if it doesn't please move on.
I'd also like the moderators to transfer the posts about double blind testing to a new thread.
I abuse my wife in the kitchen.
And that's not nice!
In this case, I think I am just trying to understand what your opinion actually is.
Changing the discussion ever so slightly, for years I've been wondering about specifications for DSP gear, especially the familiar specs. You wouldn't buy a pre-amp without looking at the specs.
The answer I usually get is, "You're a moron. Don't you know DSP gear can have no distortion, etc.".
Actually..... by the usual measurement yardsticks, the DCX2496 is about as good as anyone could possibly want. I am not sure that's the whole story of performance, but as I say, frequency response, distortion, noise, and so on look really good.
A historic trend in HiFi has been the pursuit of measurements that capture and fully characterize audiophile performance.
So where does that put the DCX2496? And the endless efforts to improve it?
I have an unmodified DCX and I have electrostatic speakers, from 130 Hz. I'd say anybody who uses their time and money fixing a DCX while continuing to play their music on cone speakers, certainly is not fixing what really needs fixing.
Ben
The answer I usually get is, "You're a moron. Don't you know DSP gear can have no distortion, etc.".
Actually..... by the usual measurement yardsticks, the DCX2496 is about as good as anyone could possibly want. I am not sure that's the whole story of performance, but as I say, frequency response, distortion, noise, and so on look really good.
A historic trend in HiFi has been the pursuit of measurements that capture and fully characterize audiophile performance.
So where does that put the DCX2496? And the endless efforts to improve it?
I have an unmodified DCX and I have electrostatic speakers, from 130 Hz. I'd say anybody who uses their time and money fixing a DCX while continuing to play their music on cone speakers, certainly is not fixing what really needs fixing.
Ben
Last edited:
I dislike the discussions about double blind tests here, there are always people who ask for those tests to be done followed by those who tell those tests are worthless. I'm in bethween and think some of them have significant results and some don't. I'm trying not to generalise and remain unbiased.
There are measurments posted on the net before and after modifications where the modified unit shows better results.
I'll dig out the sites about modifications when I get to my desktop.
The DSP part is good on the DCX but the rest is not made with HiFi in mind. The topology of the input and output stage is questionable, the power supply is noisy and the clock is adequate at best. Modifying it makes the good parts really shine.
I own the DCX for seven years now, I found it a very useful prototyping tool but found a passive crossover and a stand-alone DAC sounded better most of the time.
Then a year ago I borrowed a USB to AES/EBU interface and liked the sound much better.
That inerface was expensive to me so I went DIY and started reading about the possible modifications and kits being offered.
Now I've reached a point of diminishing returns where further improvements would require expensive kits and I like the sound as it is.
As for electrostats I've listened to Quads at my local dealer a lot, liked the reproduction but disliked the price and the low max. volume they could achieve full range. The sweet spot was microscopic.
During that time I made a choice to build cone dipole speakers and live happily since.
Sent from my C6903 using Tapatalk
Fair enough - and I agree there are good and bad double blind tests, and there are good and bad sighted listening, but as long as our brain is part of the system, we need ways to take that into account (or out of the equation).
Much appreciated. Since I haven't seen any such evidence before, I am eager to see your measurements as soon as convenient for you.
Ben
Double-blind is gilding the lily not usually needed, if the testee is blind.
"Good" and "bad" isn't the formal way to talk about tests.
Ordinarily, the challenge is placed on the testee to show they can tell the difference between condition A and B to a degree that seems convincing depending on circumstances ("95%" is just an arbitrary number they use in grad courses.)
If the testee can tell A and B apart, we say that "something" about A is different than B. In the case of a DSP, maybe there's a boost of 3dB at 7kHz, accidentally along with other changes. Maybe A is 2dB louder over all. That would be detectable but if we knew that's the case, we would never say the DSP is improved.* That's a "bad" or artefactual test what shows a difference.
The other side is tests which show no difference. They can only be bad if the test lacks what is called "power". That is, the test doesn't present enough time or sounds on which to sense the difference between A and B. In that case, we'd have to judge if the power was enough and we'd say, ".... you (or 6 people) have listened for an hour and can't tell A from B let alone agree that one is better than the other".
Ben
*OK, some people (customers?) would say, "Gosh more sparkle to that DSP now". Yes, it sounds better but a tone control does the same thing.
Last edited:
An externally hosted image should be here but it was not working when we last tested it.
ak4113 and ak4118 module in test
For the digital section, yes. For the analog section, no.
I've done the blind tests, the measurements, etc. Improvements to, or replacement of, the analog section and the power supplies make real improvements to the final output.
Some say receiver and clock upgrades make a big difference. I have not tested that.
I've posted some of that way back in this thread somewhere.
lets also mention: modify the dcx so that you get 2 or 3 spdif-outs instead of 2 or 3 analog outs. at that point, you can argue that it now is a 'math numbers' box and the psus and analog sections don't enter into it, anymore.
I did a 2 spdif-out mod so I have digital in and digital out. I could care less what the psu is like, as long as it has not affected the digital streams. after several years of daily use, I can't see any issues with this concept (maybe someday I'll do a board for that mod; mine is a one-off perf board build).
I did a 2 spdif-out mod so I have digital in and digital out. I could care less what the psu is like, as long as it has not affected the digital streams. after several years of daily use, I can't see any issues with this concept (maybe someday I'll do a board for that mod; mine is a one-off perf board build).