Hi -

I have a piece of outboard hardware I have been routing from/to my current interface and Cubase. For years I've been doing it via my interface's analog connections, just sort of pretending that the s/pdif connections on each unit weren't present.

I'd like to think about using those s/pdif connections now, just to see if maybe the sound quality would be a little better with less A/D conversions (4 of them I might be able to avoid, if my thinking is right - D/A out of the interface > A/D into the hardware > D/A out of the hardware > A/D back into the interface ... ?)

Here is a description of the digital i/o of the hardware unit:

Quote:

---

The digital I/O is via unbalanced phono sockets, but the data can be switched between AES-EBU or S/PDIF modes, as required. There are three clocking options: internal 44.1kHz and 48kHz, or the rate at the external digital input. The digital input can also be adjusted in level via a menu, and the output resolution dithered to 24, 20, 16, or 8 bits, allowing optimised connection with a wide range of systems.

---

Some things I've been thinking about:

1) Re: the "cables" connecting the interface and the hardware unit: are any old "phono" cables OK, or are there quality issues in terms of signal transmission (I've read that I should buy "coaxial" ones for a better signal, like this http://www.tvcables.co.uk/cgi-bin/tvcables/PH402.html - but I don't have the perspective to know if that is just snake oil like magic crystals, etc.)?

2) Also re: the s/pdif cables: "Unbalanced" - I know about the importance of balanced/unbalanced for audio signals, but is that important for digital as well?

3) As far as clocking goes, should I just tell the outboard hardware to use the clock on my interface, rather than the other options listed above?

4) "Dithering" - if I am going to return to my interface via s/pdif digital, am I right in thinking I should NOT dither at this stage, leaving it instead for the final mixdown to CD?

Thanks for any thoughts on this!

--------------------
Alexis -Cubase 6.5.0/SX3.1.1.944, XP SP2, 4GB RAM (1GB not accessible, but used just to balance the computer so it doesn't tip over); Delta 66 in Omni i/O Studio; Motif8; UAD-1

what is your piece of "outboard hardware"?

--------------------
http://soundcloud.com/seaapes

TC Helicon Voice One. Circa 2002.

Thanks,

--------------------
Alexis -Cubase 6.5.0/SX3.1.1.944, XP SP2, 4GB RAM (1GB not accessible, but used just to balance the computer so it doesn't tip over); Delta 66 in Omni i/O Studio; Motif8; UAD-1

you could connect the s/pdif out from the tc to the s/pdif in of your interface
do you put a mic straight into the tc when recording?
your a/d conversion would take place in the tc rather than in the daw interface...and you'd probably free up an input on your interface as well

try it....



--------------------
http://soundcloud.com/seaapes

Quote Wease:

---

you could connect the s/pdif out from the tc to the s/pdif in of your interface
do you put a mic straight into the tc when recording?
your a/d conversion would take place in the tc rather than in the daw interface...and you'd probably free up an input on your interface as well

try it....

---

No go, Wease - no preamp in the tc unit.

Thanks for your help here, and in the other thread re: combining interfaces.

--------------------
Alexis -Cubase 6.5.0/SX3.1.1.944, XP SP2, 4GB RAM (1GB not accessible, but used just to balance the computer so it doesn't tip over); Delta 66 in Omni i/O Studio; Motif8; UAD-1

Quote alexis:

---

Some things I've been thinking about:

1) Re: the "cables" connecting the interface and the hardware unit: are any old "phono" cables OK, or are there quality issues in terms of signal transmission (I've read that I should buy "coaxial" ones for a better signal, like this http://www.tvcables.co.uk/cgi-bin/tvcables/PH402.html - but I don't have the perspective to know if that is just snake oil like magic crystals, etc.)?

2) Also re: the s/pdif cables: "Unbalanced" - I know about the importance of balanced/unbalanced for audio signals, but is that important for digital as well?

---

Since you've already heard about the interfacing and clocking, I'll just try to respond to your other questions about cabling.

Regarding 1): No, don't use "phono" cables. You want to use "composite video" or "component video" type coaxial cables for S/PDIF. Such video cables are better for avoiding degradation of digital audio signals, because of lower capacitance than typical phono/RCA/Cinch-Jones audio cables (which are also "coaxial" cables btw).

Regarding 2): A problem which can still arise when using unbalanced cabling for carrying digital audio is "ground loops" between connected devices, although usually this is not as problematic for digital signals since the noise (mostly audible hum) on a "ground loop" is typically at low audio frequencies (power mains frequency and its harmonics) and not anywhere near the far higher frequencies of digital audio signals, so interfaces can easily filter this noise out without affecting the digital signal. AES/EBU can (preferably) use balanced cabling, or unbalanced coaxial cabling like "consumer grade" S/PDIF. While one can usually get by using unbalanced video cables for S/PDIF without encountering problems, if your device and interface both offer the option and both have AES/EBU connectors, you may want to use balanced cabling, especially for longer cable runs. Another option, if your gear offers it, is S/PDIF over "TOSLINK" optical (no ground loops!).

HTH,
=Goddard

Thank you, Goddard, for the great explanation! ^

--------------------
Alexis -Cubase 6.5.0/SX3.1.1.944, XP SP2, 4GB RAM (1GB not accessible, but used just to balance the computer so it doesn't tip over); Delta 66 in Omni i/O Studio; Motif8; UAD-1

Re ground loops and co-ax S/PDIF.

A minutes work with my trusty Fluke told me that the RCA out on both my Fast track pro and my NI KA6 were NOT connected to chassis. This is also the case for the S/PDIF out on my 2496 cards.

The RCA input IS however bonded to chassis in each case and I wonder why?

If a loop is a problem then yes, optical can fix it and converters can be had for less than £10 these days tho' it might be cheaper and less hassle (no faffing about with power) to use a transformer balun each end?

Dave.

Quote ef37a:

---

Re ground loops and co-ax S/PDIF.

A minutes work with my trusty Fluke told me that the RCA out on both my Fast track pro and my NI KA6 were NOT connected to chassis. This is also the case for the S/PDIF out on my 2496 cards.

The RCA input IS however bonded to chassis in each case and I wonder why?

---

Perhaps to avoid ground loops? Seems a sensible convention.

Quote Exalted Wombat:

---

Quote ef37a:

---

Re ground loops and co-ax S/PDIF.

A minutes work with my trusty Fluke told me that the RCA out on both my Fast track pro and my NI KA6 were NOT connected to chassis. This is also the case for the S/PDIF out on my 2496 cards.

The RCA input IS however bonded to chassis in each case and I wonder why?

---

Perhaps to avoid ground loops? Seems a sensible convention.

---

Nay, good Sir nay! If you want to avoid hum loops you float both ins and outs but I agree, so long as everyone stays with that connection chassis won't be linked.

Mind you the situation is probably a rare one since much S/PDIF equiped kit OTHER than AIs/computers such as CD/DVD machines are class 2, insulation and thus earth free. Then AFAIK all Mdisc recorders are class 2 AND optical?

Dave.

Quote ef37a:

---

Nay, good Sir nay! If you want to avoid hum loops you float both ins and outs but I agree, so long as everyone stays with that connection chassis won't be linked.

---

Isn't there a similar convention with MIDI connections? Ground is only connected at one end? I thought that was because of ground loops?

Hmm (or should that be "hum"?), seems Martin has dealt with this very issue previously!

http://www.soundonsound.com/sos/feb09/articles/qa0209_2.htm

http://www.kvraudio.com/forum/viewtopic.php?t=304425

It's like deja vu all over again...

Afaik, it is a common practice (if not actually required per the S/PDIF specs) to ground the screen/shroud at the output and isolate at the output (e.g. use a transformer ala RME's 'ground-free' and 'galvanically-isolated' designs). But I've also seen the screen/shroud terminal at the input being simply floated via a cap to ground.

Roland advertise a "ground lift functionality" for some of their interfaces, but dunno how they actually implement that.

The electrical specs for S/PDIF are in IEC60958-1 Rev3 (they used to be in IEC60958-3 but were moved to part 1 during the last revision). I'll leave it to someone else to look them up (see around p15).

Pass me that orange-tipped cable, Speedy!
=Goddard

Edited by Goddard (08/05/12 01:54 PM)

Quote ef37a:

---

Re ground loops and co-ax S/PDIF...

If a loop is a problem then yes, optical can fix it and converters can be had for less than £10 these days...

Dave.

---

Seems like this whole week of topics for me has been mind-addling , but can I ask one more newbie question please ... "converter" as in S/PDIF to optical (and then maybe back again if the receiving hardware unit requires S/PDIF)? Would that take care of digital ground loops?

Thanks!

--------------------
Alexis -Cubase 6.5.0/SX3.1.1.944, XP SP2, 4GB RAM (1GB not accessible, but used just to balance the computer so it doesn't tip over); Delta 66 in Omni i/O Studio; Motif8; UAD-1

Quote alexis:

---

Seems like this whole week of topics for me has been mind-addling , but can I ask one more newbie question please ... "converter" as in S/PDIF to optical (and then maybe back again if the receiving hardware unit requires S/PDIF)? Would that take care of digital ground loops?

---

Yes - but try a wired connection first.

James.

--------------------
JRP Music - Audio Mastering and Restoration.
http://www.jrpmusic.net

Yes, a co-ax-opto-opto-coax converter string will prevent any electrical currents, 50Hz or otherwise, flowing between the two pieces of equipment because the interconnect is a length of transparent plastic.

BTW, if the notion of an extra "double conversion" bothers you this is not like another set of A/D-D/A converters, the electrical signal merely turns an LED on and off and 'tother end a phototransistor perfoms the opposite function. A chip is used merey to boost the currents, no buffers, clocks or latency is involved (AFAIK!).

Dave.

Quote Goddard:

---

Afaik, it is a common practice (if not actually required per the S/PDIF specs) to ground the screen/shroud at the output and isolate at the output (e.g. use a transformer ala RME's 'ground-free' and 'galvanically-isolated' designs). But I've also seen the screen/shroud terminal at the input being simply floated via a cap to ground.

---

Erk, mistyped. That should have gone "...ground the screen/shroud at the input and isolate the output..."

Oh, and the cap floating the input terminal shroud mentioned earlier is for bypass (so, technically, effectively an AC ground but won't show as connected to chassis when you put an Ohm meter probe to it).


There, that looks better.
=Goddard