12G SDI cable as digital audio link

[yeti42]

Specifically a Belden 4794R obtained from Blue Jeans Cables (but not their UK outlet). For about £40 shipped from the States it's performing remarkably well between my Naim Core and Rega DAC-R. With the Rega it livens things up nicely compared to a Chord Fairway. I've also tried it into a Chord 2qute where it managed to let through more emotional comunication than a Shawline (of which the 2qute with Shawline had none).

 

[MJS]

No reason why this shouldn't work perfectly, the cable and connectors have the same electrical requirements and 12G SDI is very demanding.

 

[Patu]

I need digital interconnect cable between my TeddyPardo U2S USB bridge and Naim DAC. I've used Naim DC1 (BNC-BNC) for years and not given this any thoughts since I was quite sure there couldn't be any differences between coaxial cables or if there would, they'd be extremely small.

Well, after I heard clear difference between two different USB cables, I gave this one another thought. Someone mentioned the 12G SDI cables on Naim forum, preferring them to DC1. I googled what this term even means and ended up on BJC website. I've ordered few cables from them in past and everything's worked fine so I decided to give it a try. After all, the price is ridiculously low compared to branded cables.

4794R arrived quickly from US to Finland. I plugged it in and was very surprised of the performance. Definitely very different presentation from DC1, to which my ears have got used to over 4-5 years I've owned it. While DC1 sounds like Naim components usually sound, being quite aggressive with the typical front row sound and nice rolling feeling, Belden did things in more neutral way. It feels like the vocalists is pushed slightly back and the soundstage stretched little bit wider. Belden also digs much deeper in the extremes, hitting deeper in the low end and being more lively on the top end. Never was there any harshness or did the cable sound overly bright though. DC1 hasn't got very much use after Belden arrived.

I was so impressed by 4794R that I decided to give the better known 1694A a try. In same order I ordered one Canare coaxial cable, just to try it out. Its performance was quite a bit worse than DC1 or any Belden cables. 1694A is the coaxial cable BJC recommends for audio use. It has similar construction but instead of being silver plated solid core copper, 1694A doesn't have silver plating. There's some other differences also, specs can be found on BJC website. Once again, the cable arrived quickly and I plugged it in. Very similar sound signature than 4794R but to my ears, sound flows slightly better with 1694A and is more organic. Everything just sounds correct with this one. It's sound is somewhere between DC1 and 4794R, being closer to 4794R. Stunning performance all in all. I have 1694A plugged in now and no plans to unplug it. I'll give DC1 another go after a while but I don't think it matches the performance of 1694A.

 

[yeti42]

It's a shame Rega didn't fit at least one 75 Ohm BNC input to their DAC but at least the reflected waves should be absorbed at the Core.

 

[MJS]

We made a small handful of these a while back with 1394A cable, the last one is on eBay now (coincidentally, we put it up hours before this thread started).

 

[yeti42]

I've received a 1694A terminated rca and bnc yesterday, a bit harsh on first use compared to the 4794A but I'll give it a chance to settle down with continuous signal for a week before chanbing back. Ordered this through the bjc uk website but it still ships from the US so no real advantage to its existance.

 

[Patu]

I've received a 1694A terminated rca and bnc yesterday, a bit harsh on first use compared to the 4794A but I'll give it a chance to settle down with continuous signal for a week before chanbing back. Ordered this through the bjc uk website but it still ships from the US so no real advantage to its existance.

This was my first impression also. There is slightly smoother touch with 4794R but 1694A sounds livelier and more fun to listen to. I actually ordered another pair of 1694A, this time through eBay with the same Canare plugs. I wanted it shorter since this is the cable I'm gonna keep in the setup for now.

 

[shrink]

I've received a 1694A terminated rca and bnc yesterday, a bit harsh on first use compared to the 4794A but I'll give it a chance to settle down with continuous signal for a week before chanbing back. Ordered this through the bjc uk website but it still ships from the US so no real advantage to its existance.

I'm genuinely curious as to what properties of a digital cable you think are contributing towards harshness that you feel will disappear over time or use?

 

[Guest432]

Impedance mismatched reflections or standing waves are completely irrelevent in hifi because of the frequencies involved.

You can have a completely mismatched impedance between the source and DAC and even using a high carrier freq for SPDIF of 96 KHz the wavelength would be over 3000 metres, that's one hell of an interconnect needed to be impacted by reflection losses or interference.

It is just another load of old foo bollox and anyone who says they can hear a difference between different competently designed digital cables is imagining it.

 

[Yomanze]

Impedance mismatched reflections or standing waves are completely irrelevent in hifi because of the frequencies involved.

You can have a completely mismatched impedance between the source and DAC and even using a high carrier freq for SPDIF of 96 KHz the wavelength would be over 3000 metres, that's one hell of an interconnect needed to be impacted by reflection losses or interference.

It is just another load of old foo bollox and anyone who says they can hear a difference between different competently designed digital cables is imagining it.

Your maths is wrong as you're just looking at the sampling rate. Here you go, a lesson for you:

"The S/PDIF digital audio interconnect format is described by a sampling rate of the 64-bit audio data frames. Each 64 bit frame contains a single audio sample for both the left and right channels along with various status bits. Therefore for the sampling rates of 48kHz and 96 kHz discussed here, the information bit rates are 3.072 and 6.144 Mbit/sec respectively. However since the encoding used in the S/PDIF format is biphase mark code, a 1 bit would have 2 transitions in the bit interval. This means that the clock rate is twice the information bit rate. The coaxial S/PDIF standard specifies an electrical level of 0.5Vpp at a termination impedance of 75 ohm. What receiver bandwidth is required for proper detection of bit streams for 48kHz and 96 kHz? From basic binary data recovery analysis, it is known that the required bandwidth is HALF the transmission bit rate. Therefore, the receiver must have a minimum bandwidth of about 3 MHz to properly detect a 48 kHz S/PDIF stream, or 6 MHz for the 96 kHz case."

http://www.jensign.com/SPDIFLink/

In any case, it's more than just impedance mismatches, slower rise times need to be considered as well as as return loss...

 

[Julf]

Therefore, the receiver must have a minimum bandwidth of about 3 MHz to properly detect a 48 kHz S/PDIF stream, or 6 MHz for the 96 kHz case[/b]."

Both are very low frequencies from a RF perspective. Any basic coax cable should do a reasonable job at those frequencies, and precise impedance matching is not required.

In any case, it's more than just impedance mismatches, slower rise times need to be considered as well as as return loss...

What effect does return loss have on the signal (in your opinion)?

 

[Patu]

I didn't expect the deniers to come out so strong on this forum. I thought the atmosphere is more open and subjective here. Our local forum with threads considering digital transmission are usually flamed by people who deny the differences because bits are bits and other bs theories. I really don't care how or why there are differences between digital cables but if I hear the difference with my own ears in my own setup, it's all that matters. In this case, ~25 € cable works better than ~400€ Naim cable so I'm quite happy about my findings.

 

[Julf]

So instead of discussing the factual arguments, you prefer to label anyone disagreeing with you "deniers", and refer to "bs theories"? I am sure that qualifies as having an open mind...

 

[Guest432]

Your maths is wrong as you're just looking at the sampling rate. Here you go, a lesson for you:

"The S/PDIF digital audio interconnect format is described by a sampling rate of the 64-bit audio data frames. Each 64 bit frame contains a single audio sample for both the left and right channels along with various status bits. Therefore for the sampling rates of 48kHz and 96 kHz discussed here, the information bit rates are 3.072 and 6.144 Mbit/sec respectively. However since the encoding used in the S/PDIF format is biphase mark code, a 1 bit would have 2 transitions in the bit interval. This means that the clock rate is twice the information bit rate. The coaxial S/PDIF standard specifies an electrical level of 0.5Vpp at a termination impedance of 75 ohm. What receiver bandwidth is required for proper detection of bit streams for 48kHz and 96 kHz? From basic binary data recovery analysis, it is known that the required bandwidth is HALF the transmission bit rate. Therefore, the receiver must have a minimum bandwidth of about 3 MHz to properly detect a 48 kHz S/PDIF stream, or 6 MHz for the 96 kHz case."

http://www.jensign.com/SPDIFLink/

In any case, it's more than just impedance mismatches, slower rise times need to be considered as well as as return loss...

My maths is fine, my transmission theory was wrong as I don't particularly care what the specifics are for spdif. As an ex R Sigs radio tech I know about impedance matching and that it is not an issue at low freqs.

Even using your highest number of 6 MHz it would need an interconnect of over 50 metres to be an issue.

Rise time is also really only an issue in high speed digital ccts.

Why is return loss an issue? (the losses on a mismatched 1 metre cable would be difficult to measure or see on a scope)

 

[Patu]

So instead of discussing the factual arguments, you prefer to label anyone disagreeing with you "deniers", and refer to "bs theories"? I am sure that qualifies as having an open mind...

During my 17 year journey in hifi, it's always been the same with digital components. Transports sound the same because bits are bits, digital cables are the same because bits are bits and so on. People seem to think that in digital chain, everything's been founded and invented already and everything's measurable. There's no room for improvement anymore. People also seem to think that it's 1/1 possible to mimic human ear and brains with some measuring mics and gadgets. Sounds absurd to me.

I've tried to explain how it really makes no difference how many "factual arguments" someone shows me if I still hear the difference when I sit down on my sofa and turn up the music. There's no point in blind testing since that's not how I listen to my music on daily basis. You need to audition components in the same situation you'd be using them. Nothing else matters really.

When you open your mind and audition this stuff yourself, it's very easy to hear the differences. I didn't believe this stuff either before I decided to give tranports, USB-cables and digital cables a try.

 

[Guest432]

During my 17 year journey in hifi, it's always been the same with digital components. Transports sound the same because bits are bits, digital cables are the same because bits are bits and so on. People seem to think that in digital chain, everything's been founded and invented already and everything's measurable. There's no room for improvement anymore. People also seem to think that it's 1/1 possible to mimic human ear and brains with some measuring mics and gadgets. Sounds absurd to me.

I've tried to explain how it really makes no difference how many "factual arguments" someone shows me if I still hear the difference when I sit down on my sofa and turn up the music. There's no point in blind testing since that's not how I listen to my music on daily basis. You need to audition components in the same situation you'd be using them. Nothing else matters really.

When you open your mind and audition this stuff yourself, it's very easy to hear the differences. I didn't believe this stuff either before I decided to give tranports, USB-cables and digital cables a try.

CD Transports don't sound the same because they have differing abilities to extract the data from the CD.

 

[Julf]

There's no point in blind testing since that's not how I listen to my music on daily basis.

I admire your open mind and willingness to question assumptions.

Despite my own 40-year journey in audio, electronics and technology I still haven't managed to convince myself of the infallibility of my senses. When I am building a shed, I still have to use a measuring tape instead of just relying on my eyes. The same goes for my ears.

 

[Purité Audio]

CD Transports don't sound the same because they have differing abilities to extract the data from the CD.

Cd specs were pretty robust, I have never heard any differences between transports unless broken.
Keith

 

[Patu]

CD Transports don't sound the same because they have differing abilities to extract the data from the CD.

Yes, even USB-transports have huge differences. I use one in front of my Naim DAC.

I admire your open mind and willingness to question assumptions.

Despite my own 40-year journey in audio, electronics and technology I still haven't managed to convince myself of the infallibility of my senses. When I am building a shed, I still have to use a measuring tape instead of just relying on my eyes. The same goes for my ears.

That's totally fine. I accept the fact that I might be imagining the differences. But it really doesn't matter since the difference is still there and I can sense it. It's there in the real world situation where I listen to my system and music every day. How and why I hear the difference is irrelevant. Even if I'd do the blind test, I'd still have to do the final decision in open audition and once again I'd hear (or imagine) the difference.

 

[Purité Audio]

That is the beauty of unsighted comparisons, if unsighted you can't tell any difference you can just relax knowing that you can't hear the difference.
Keith