Digital Co-axial (SPDIF) Output Quality?

[Lugs]

Hello,

Looking at the vast range of streaming options available. There is a quite a choice and prices vary wildly.

I would not use the on-baord DAC in any such device.

Is the co-axial digital output (spdif) quality consistantly the same between products?

Are cheap models plagued by jitter.

Regards

 

[matt49]

There are some boxes that are known to have relatively poor jitter performance (e.g. Sonos Connect), though it's not clear whether this affects SQ.

 

[davedotco]

SPDIF is a heavily compromised method of moving data between separate components.

There are clocking issues, impedance mismatches, noise transmission and half a dozen other issues that degrade the data stream though fortunately a good dac can, by various methods, reduce these issues to manageable proportions.

SPDIF is not used within components, neither is it used in professional applications, personally if I could set a system up without using it, I would do so.

 

[ID.]

SPDIF is a heavily compromised method of moving data between separate components.

There are clocking issues, impedance mismatches, noise transmission and half a dozen other issues that degrade the data stream though fortunately a good dac can, by various methods, reduce these issues to manageable proportions.

SPDIF is not used within components, neither is it used in professional applications, personally if I could set a system up without using it, I would do so.

that said, I don't think I can hear any difference between USB and optical with my Pioneer, and the lock range can be set pretty narrowly for either connection.

 

[abacus]

Scroll down for s/pdif http://www.soundonsound.com/sos/feb07/articles/interfacing.htm

Hope this helps

Bill

 

[davedotco]

That's fair comment.

A lot of these 'digital issues' seem inaudible in real world scenarios, most of the audible digital problems I have come up against are due unsuitable cables or errors in setup.

Given the current state of the digital art, keeping it simple seems to be a sound practice.

Like you I have a Mac, a late model laptop and demonstrations at dealers where I have used both Toslink and USB are inconclusive. There may be differences but in my situation they were trivial.

In general when selecting and setting up a hi-fi, I try and maintain a regime of 'good houskeeping', basically just doing things in the correct way, keeping it simple and getting the details right. I have always found such consistency to be very much worthwhile.

 

[davedotco]

Double bubble......*dash1*

 

[hybridauth_Facebook_100004211376004]

The difference between coaxial and optical are marginal at worst.I use a chord optichord from my Amazon fire tv box into Arcam ir dac and it sounds just fine,I've used coaxial and optical through a cd transport and it didn't make any audible difference..... to me anyway.

 

[DocG]

I would not use the on-baord DAC in any such device.

As Dave pointed out, the connection between the actual streamer (the transport, if you like) and the onboard DAC is not using SPDIF. It's usually through I2S (ditto for most CD players, BTW), a standard that doesn't squash the clock data into the signal. So using a separate DAC (with an SPDIF-cable), you lose this advantage. Better compare both options (onboard vs. separate DAC), using your ears, before deciding which way you choose.

Edited for spelling.

 

[davedotco]

I would not use the on-baord DAC in any such device.

As Dave pointed out, the connection between the actual streamer (the transport, if you like) and the onboard DAC is not using SPDIF. It's usually through I2S (ditto for most CD players, BTW), a standard that doesn't squash the clock data into the signal. So using a separate DAC (with an SPDIF-cable), you lose this advantage. Better compare both options (onboard vs. separate DAC), using your ears, before deciding which way you choose.

Edited for spelling.

Some years ago a company called Audio Alchemy, famous for it's budget 'Dac in a Box' add on dac attempted to introduce I2S as a digital interface between components.

Sadly no CD player/transport manufacturers took this up which left AA with awkward interface boxes and non standard connectors. This was a shame as it showed some promise.

Rsconstructing the clock from the SPDIF data stream is the big issue, different solutions have been tried including the use of a 'master clock' often with separate connections, fi-fo buffers and now, with the advent of the asyncronous dac, total reclocking.

As I said earlier, best to do without it if you can. All in one streamer/dac/preamp is the way forward though as yet there are few such items available at sensible prices.

 

[Mike_Schmidt]

In the near future I was going to pick up a Naim 5XS to go with my cyrus pre DacQx but was going to the spidif as that is the only digital output. Why would naim have a output that is not any good for sound quality?

 

[Benedict_Arnold]

The debate between optical and co-axial is an old chestnut around here. Only speaker cables (and, latterly, HDMI cables) raise more debate.

Here is my 2 cent's worth though.

The source material is either going to be optical (in the cases of CDs, DVDs or Blu-Rays) or digital in the form of computer files, data streams, whatever. Every time the signal gets translated from optical to digital or digital to optical, there is the chance, however small, of an error.

So if you start with a CD (for example), in a CD transport linked to a separate DAC linked by an optical cable, the signal goes:

CD > CD Transport Optics > Optical > CD Transport Digital Electronics > Back to Optical > Optical cable > DAC Optics > Back to Digital Signal > DAC Electronics > Onwards

For an electronic or co-axial interconnect the signal goes:

CD > CD Transport Optics > Optical > CD Transport Digital Electronics > Co-Axial Cable > DAC Electronics > Onwards

So, by using a co-axial cable you've saved the digital electronic signal from being converted back into an optical signal back into a digital electronic signal. That eliminates two potential sources of errors.

For a digital electronic source, such as a FLAC or WAV file, a data stream over theinternet, whatever, the process is the same: to transmit over an optical cable requires the signal to be converted into an optical one, then back into an electronic one. So, again, by using a coaxial cable you eliminate two conversion processes and the chances of an error should therefore be reduced.

All that having been said, I can't hear the [expletive deleted] difference either and its as much down to personal choice, the equipment you use (and what sockets the boxes have), and the converters and connectors as anything else.

 

[Mike_Schmidt]

Amazing repsonse greatly appreciated.

 

[davedotco]

The debate between optical and co-axial is an old chestnut around here. Only speaker cables (and, latterly, HDMI cables) raise more debate.

Here is my 2 cent's worth though.

The source material is either going to be optical (in the cases of CDs, DVDs or Blu-Rays) or digital in the form of computer files, data streams, whatever. Every time the signal gets translated from optical to digital or digital to optical, there is the chance, however small, of an error.

So if you start with a CD (for example), in a CD transport linked to a separate DAC linked by an optical cable, the signal goes:

CD > CD Transport Optics > Optical > CD Transport Digital Electronics > Back to Optical > Optical cable > DAC Optics > Back to Digital Signal > DAC Electronics > Onwards

For an electronic or co-axial interconnect the signal goes:

CD > CD Transport Optics > Optical > CD Transport Digital Electronics > Co-Axial Cable > DAC Electronics > Onwards

So, by using a co-axial cable you've saved the digital electronic signal from being converted back into an optical signal back into a digital electronic signal. That eliminates two potential sources of errors.

For a digital electronic source, such as a FLAC or WAV file, a data stream over theinternet, whatever, the process is the same: to transmit over an optical cable requires the signal to be converted into an optical one, then back into an electronic one. So, again, by using a coaxial cable you eliminate two conversion processes and the chances of an error should therefore be reduced.

All that having been said, I can't hear the [expletive deleted] difference either and its as much down to personal choice, the equipment you use (and what sockets the boxes have), and the converters and connectors as anything else.

There is no chance of data error in the way you describe, digital data transfer is effectively error free.

The issue is jitter/ timing and noise. In severe cases the dac will not be able to 'lock' onto the incoming data stream. On the other hand most good modern equipment will handle this with aplomb.

You may also notice that Naim prefer BNC connectors for co-ax connection, this is an impedance thing and may reduce jitter.

 

[Mike_Schmidt]

I did notice Naim had the BNC but didnt take notice until read this thread but everything is clear now. the only problem is I still need a RCA on the coax end for my cyrus pre. I have narrowed down to two Naim 5XS or Cyrus CD signature.

 

[Benedict_Arnold]

There is no chance of data error in the way you describe, digital data transfer is effectively error free.

The issue is jitter/ timing and noise. In severe cases the dac will not be able to 'lock' onto the incoming data stream. On the other hand most good modern equipment will handle this with aplomb.

You may also notice that Naim prefer BNC connectors for co-ax connection, this is an impedance thing and may reduce jitter.

Fairy nuff, if you don't consider jitter and noise as errors in the sense that a 1 becomes a 0 or vice versa, but my points still stand: (1) the use of a fibre-optic interconnect introduces two additional processing steps which could cause "jitter and noise", as you put it, if not errors per se; (2) most of us won't be able to tell the flipping difference with most kit anyway; (3) that what you use can depend on what "outs" and "ins" you have available; and (4) interconnect threads are like speaker cable threads - doomed to squabbling from the get-go.

 

[matt49]

The debate between optical and co-axial is an old chestnut around here. Only speaker cables (and, latterly, HDMI cables) raise more debate.

Here is my 2 cent's worth though.

The source material is either going to be optical (in the cases of CDs, DVDs or Blu-Rays) or digital in the form of computer files, data streams, whatever. Every time the signal gets translated from optical to digital or digital to optical, there is the chance, however small, of an error.

So if you start with a CD (for example), in a CD transport linked to a separate DAC linked by an optical cable, the signal goes:

CD > CD Transport Optics > Optical > CD Transport Digital Electronics > Back to Optical > Optical cable > DAC Optics > Back to Digital Signal > DAC Electronics > Onwards

For an electronic or co-axial interconnect the signal goes:

CD > CD Transport Optics > Optical > CD Transport Digital Electronics > Co-Axial Cable > DAC Electronics > Onwards

So, by using a co-axial cable you've saved the digital electronic signal from being converted back into an optical signal back into a digital electronic signal. That eliminates two potential sources of errors.

For a digital electronic source, such as a FLAC or WAV file, a data stream over theinternet, whatever, the process is the same: to transmit over an optical cable requires the signal to be converted into an optical one, then back into an electronic one. So, again, by using a coaxial cable you eliminate two conversion processes and the chances of an error should therefore be reduced.

All that having been said, I can't hear the [expletive deleted] difference either and its as much down to personal choice, the equipment you use (and what sockets the boxes have), and the converters and connectors as anything else.

There is no chance of data error in the way you describe, digital data transfer is effectively error free.

The issue is jitter/ timing and noise. In severe cases the dac will not be able to 'lock' onto the incoming data stream. On the other hand most good modern equipment will handle this with aplomb.

You may also notice that Naim prefer BNC connectors for co-ax connection, this is an impedance thing and may reduce jitter.

The conventional view is that optical SPDIF has higher jitter than coax SPDIF because of the conversion of electrical to optical signals and vice versa.

Whether this matters is another question.

 

[davedotco]

Fairy nuff, if you don't consider jitter and noise as errors in the sense that a 1 becomes a 0 or vice versa, but my points still stand: (1) the use of a fibre-optic interconnect introduces two additional processing steps which could cause "jitter and noise", as you put it, if not errors per se; (2) most of us won't be able to tell the flipping difference with most kit anyway; (3) that what you use can depend on what "outs" and "ins" you have available; and (4) interconnect threads are like speaker cable threads - doomed to squabbling from the get-go.

Digital transmission is pretty well understood because it is a human construct, it was designed to operate that way. Toslink optical is both robust and error free, it is also largely free of noise and hum, however it is limited in terms of bandwidth.

For electrical connections, the bnc connector is a better match impedence wise but only really works at both ends of a cable. If you are at all practical, you can often replace the phono socket with a 75 ohm bnc.

 

[davedotco]

The conventional view is that optical SPDIF has higher jitter than coax SPDIF because of the conversion of electrical to optical signals and vice versa.

Whether this matters is another question.

That is not my understanding, but then with all things digital my understabding is somewhat patchy.

What I do recall though is some research carried out by a fairly serious hi-end manufacturer I was working with showed the optical connection to be far superior, though to be fair they were using propriety hardware rather than 'off the shelf' Toslink converters.

As is so often in hi-fi, there is more than one 'right' answer and it is usually the compedence with which a particular solution is implemented that is critical.

As I have said in a number of threads, when putting together a system, I am a great believer in doing things the 'right' way, good housekeeping as I put it. Keep it simple, do it right, I often find the cummulative effects of such practice to be very worthwhile.

 

[matt49]

That is not my understanding, but then with all things digital my understabding is somewhat patchy.

What I do recall though is some research carried out by a fairly serious hi-end manufacturer I was working with showed the optical connection to be far superior, though to be fair they were using propriety hardware rather than 'off the shelf' Toslink converters.

As is so often in hi-fi, there is more than one 'right' answer and it is usually the compedence with which a particular solution is implemented that is critical.

As I have said in a number of threads, when putting together a system, I am a great believer in doing things the 'right' way, good housekeeping as I put it. Keep it simple, do it right, I often find the cummulative effects of such practice to be very worthwhile.

I agree entirely that 'good housekeeping' is what it's all about. A well designed combined streamer/DAC/preamp has a lot to be said for it. If the streamer/renderer has to be separate, USB makes more sense to me than SPDIF, although some folk think the USB power line is a whole new world of pain. *unknw*

These results suggest that the jitter performance of SPDIF coax is seven times better than that of SPDIF optical. Again, whether it matters is moot.

The advantage of optical is that it provides electrical isolation, and in some (many?) systems that may outweight optical's relatively poor jitter performance.

Personally, I've never been able to hear a difference between coax and optical in my systems.