Meridian Unplugged - Powered by MrTechGuy.co.uk Forums Discussion HiFi & HT Chat Digital is digital ? No so sure...

In anticipation of getting a new DSP room correction unit that just uses optical or USB input and not coaxial (Dspeaker anti-mode 2.0) , I took my Squeezebox touch that outputs both optical and SPDIF simultaneously and attached both into my Z-sys RDP-1 digital pre/parametric.

Z-sys RDP-1

This allows instant switching between sources and considering there is only one source , it effectively allows comparison between co axial and optical. I Use Kimber D60 as SPDIF cable and Kimber OPT1 for optical.

Ostensibly , both inputs should sound as close as dammit to each other... but this in not the case .. the optical has more drive and a more defined bass and a more forward focussed sound , the SPDIF has more "air" or perhaps more low level detail , a slightly more diffuse sound and less bass drive but it seems to go lower...

I can't say whether this is the cable itself , the output of the squeezebox , or the input of the Z-sys affecting things , but it's sure not subtle or a psychoacoustic placebo effect.
Neither sound is objectionable .. for fast paced music the optical is my preference , for more slow pace , the SPDIF is a little more to my taste.

I'm using 24/96 Meridian DSP6000 speakers and a 568.2 processor after the Z-sys - processor in direct mode (I only use the processor for music , just l/r speakers and it is set to upsample any input to 24/96 and outputs the same to match the speaker DAC - nothing else)

I did some blind testing just now - just roped my daughter in and told her to change at random and not always from A to B.. problem is that changing A to B or back has a slight pause while the Z-sys gets a lock and A to A is seamless so you can tell when doing a "false" switch..but I was able to identify whether it was A or B every time.

I firmly believe in at the very least level matched AB testing when evaluating a component if possible as well as trusting my ears.. single blind testing too - but double blind or ABX testing is mostly impossible for me and some components cannot even be AB tested.

My next test is to use Optical out of my Z-sys (which has all digital outputs active at all times )and into the 568.2 and just set up another source as direct but using optical and switching the 568.2 between sources , to see of it also changes with respect to optical and coax...

Hi

Interesting; but this is the kind of thing I almost wish I'd never read.

Hopefully it's something to do the Squeezebox outputs or the inputs.

Neil

I have just hooked up Toslink only out my SBT to my Z-Sys , disconnected Coaxial. Then took both coax and toslink out to my 568.2 processor and set up the processor to have identical configured inputs , one using optical , the other coaxial and once again roped my daughter into it....
This time , no matter whether you press the same input twice or change them , there is a pause while the 568 gets lock , so essentially a "false" switch from A to A is not identifiable.. can still say which is which with a 100% accuracy...even with false switching

The Z -sys does no DSP in bypass mode all it will do is allow me to format convert - ie an optical in can be output as coax , optical and AES/EBU..

The Meridian processor has anti jitter circuitry and allows output at 24/96 which I have enabled , so it does some DSP (as do the speakers , also has anti jitter circuitry and X/over etc in the digital domain before it hits the dacs in them)
Wondering if perhaps this DSP has anything to do with it?

As to testing I'm pretty much aware of expectations when doing it yourself , so tried to eliminate them using my daughter and the false switching etc.

I had a few audiophile pals over some years ago and set up 3 presets on my Z-sys .. told them one had more treble , one had more bass and one was flat , asked them which they preferred when switching between the 3 and got differing answers as to preferences..thing was there was NO difference between any of the presets..they all were flat..they were somewhat flabbergasted when I revealed this...

Makes me curious how this compares to a MDMS system (either) direct linked to the 6k's, and via the 568 and a 800 series (818 or 861 depending if you need room correction and multichannel or 2 ch only).

Still, great to see that despite what many believe, owning DSP speakers doesn't restrict for tweaking fun in the chain of sources. In that respect digital IS digital as long as the sample rate is compatible with the speakers.

Oh yes, is here a "why" for the testing you are doing or is it for fun, which is a very good reason at itself?

Happy listening!
Cheers,
Jaap Jan

The why is that I have various ways of implementing the digital format and cabling within some of my stuff's input/outputs and my new purchase and was trying various of them prior to getting my Dspeaker antimode dual core room correction/parametric to find out what I prefer or what sounds best.

I like fiddling too. I had DSP's since 97 or so and have messed plenty, changing drivers, using various DSP equalisers and so on.

Sure, digital audio is half analogue, that's why it can sound different.

Digital audio consists of digital audio data and digital audio clock(s).

The data is digital and the clock is analogue, and is easily corrupted. That's why cables and connectors and everything else matters

(at least, while we have to use the wrong digital audio architecture....)

Nick

Yeah, 'digital is digital' is not true in the terms we're talking about for SPDIF transmission down cables for consumer audio use, the answer above is correct, and unfortunately both coax and toslink are flawed in their different ways (e.g. coax, as used for SPDIF transmission, has an inherent impedance mismatch at either end of the cable, because some bright spark at Sony/Phillips in the design phase decided to use a phono RCA connector on either end of a 75 Ohm cable, when an RCA is usually in the order of 30-something Ohms, so you get reflections down the cable - why didn't they just use BNC, Who knows?).

Remember also that when you're using coax/toslink both will have different output conversion stages and receivers at the other end, so I'm not surprised at all you're hearing a difference.

[edit: I believe one of the valuable aspects of SpeakerLink as a transmission format is that it is less prone to these effects, and therefore more accurate?]