That sums it up nicely.
I really have no way to do a direct comparison between the Rendu and something similar.
Also, unfortunately, our test equipment is not set up to perform sensitive tests on that sort of device either.
If we were actually set up to accurately measure what's going on, once and for all, I'd make the time.
And, as far as my subjective opinion of how it sounds, that isn't worth much more than anyone else's (well, maybe a little more)
I would also point out something that is probably pretty obvious to people who know me by now.
I tend to be a bit pedantic when it comes to accuracy and euphonic differences.
For example, I calibrate my computer monitors - with a meter, I do
NOT "set them so they look nice".
In those terms I am a rather aggressive objectivist.
To me, DACs like our DC-1, and the DACs in the XMC-1, are pretty close to accurate and perfect
Of course, nothing is perfect, and there's always room for improvement, but they're pretty darned close.
(The DACs we used in the RMC-1, and in the soon-to-be-released DC-2, are a tiny bit better.)
And, when I've actually tested devices that reduce jitter, and similar system tweaks, I've always found the audible differences they make to be quite subtle.
Therefore, when I read descriptions of "huge differences" here's how I interpret them......
We started out with something pretty close to perfect...
Therefore, it's quite possible we could make tiny improvements...
But making
MAJOR improvements in accuracy is simply not possible...
Therefore, if the Rendu produces "major differences in sound", then it
MUST be doing so at the cost of accuracy.
They must be doing something that reduces accuracy in a way that many people find to sound pleasing (we call that euphonic coloration)
(If you're "really close to perfect to begin with" then you simply cannot "get a lot closer".)
Now, in the context of a device like the Rendu, this leads to several conclusions....
At a personal level, I'm quite certain that, If I agree that the Rendu sounds significantly different, that difference will be the result of some sort of euphonic coloration.
And, personally, I don't find this desirable.
At another level, a few people have suggested that "we should find out what they did and incorporate it into our products".
The idea seems to be that, at some level, they've discovered some magic trick or secret that nobody else has found yet, and which we could copy if we reverse engineered it.
To be quite honest, this is about as likely as world peace; people have taken their products apart and found no secret parts inside.
And, to be equally honest, there's nothing they could do
in software that would magically make the bits sound better without actually changing the bits.
Now, if it turned out that they were doing an awesome job of reducing jitter, and the difference was audible....
Or if they really do have excellent noise filtering, and it really produces an audible improvement....
Then this might suggest areas where we could focus our efforts in future designs.
However, the effect would be subtle, and there are no specific bits and pieces that could simply be copied; it just doesn't work that way.
And, if it turned out that they are doing something that simply produces some sort of euphonic difference, then we're right back where we were with tube gear.
All of the measurements tell us that, in almost every measurable way, tube equipment is inferior to solid state equipment.
Yet some people find the differences pleasing, and very much enjoy tube equipment.
However, to be quite honest, at Emotiva we concentrate on producing neutral and accurate equipment.
Therefore, we aren't interested in copying euphonic modicifations, either from tube equipment, or Rendu boxes.
(We might sell tube gear, but we won't add a switch labelled "tube sound" to one of our solid state amps.)
I should also point out another interesting possibility.
What if, instead of reducing jitter, they're actually adding massive amounts of specific types of jitter?
Like tube distortion, jitter is in fact audible when present in quantity, and produces an effect that some people might actually find quite pleasant.
(Jitter comes in an infinite number of types and flavors. Perhaps they've just hit upon one that a lot of people think sounds good.)
(Remember that wine is simply water that's been contaminated with rotting fruit; it's often all a matter of perspective.)
-----------
Let me add an aside here... just to provide a bit of perspective...
When we test jitter, we do it like almost everyone else does, by inference.
We input a test signal, then analyze the output, and confirm that the distortion that jitter causes isn't there.
We don't measure the jitter; we measure the symptoms of jitter; and confirm that they aren't there in significant amounts.
This gives us a very sensitive indication if we have a jitter problem - but doesn't allow us to measure it directly.
(This is what the oft-quoted "J-test" actually does.)
We have quite a few of those cool $60k industry standard Audio Precision AP585 test sets which we use for production and development...
But we haven't purchased any of the extra $10k plugins we would need to add to enable us to measure jitter directly...
It simply isn't a priority.
(And very few companies do so... even ones who make and sell expensive DACs.)
------------
What
KeithL hasn't said (and correct me if I'm wrong, Keith) is that his system is currently set up without any Ethernet streaming at all. Therefore, he'd have to totally and simultaneously change a BUNCH of things to be able to test a streaming device. Even then, how would he know whether any change he heard was due to all the OTHER significant changes or to the streaming device itself? Some folks just don't need a streamer...