I2S: Is it a good external interconnect? Give your opinion and findings!

[Tog]

they knew that endlessly fussing about I2S is a harmless but ultimately futile activity? Alas probably not...

 

yours, in the know?,tog

 

[timequest]

To Gordon:

 

Then why does the new PS Audio PWT/PWD sound so (extremely) good utilizing their proprietary I2S interface? After extensive auditioning of these components, your statement (for me) truly impacts your credibility; or reflects your stubbornness – which is probably the case because I believe you are a pretty credible person.

 

 

[CG]

For almost everybody, there's an audio solution that works for them. It's like desserts, mates, and art.

 

I2S wasn't designed for this sort of application, but people can make it work for them for what they want to accomplish.

 

SPDIF was a test signal interface that people adapted to make work for them for what they wanted to accomplish.

 

This doesn't mean that either is the *best* solution. But, it may be good enough in the overall system context to get the sound that the designers are trying to attain.

 

I was just talking to somebody about this. It seems that Americans tend to like REALLY BIG BASS, even if it is inaccurate and not "real." Others like REALLY PENETRATING TREBLE - same thing. Other people have different preferences. I think that concept applies here. Not everybody really wants accurate. You can't even get that if you wanted it, because it all starts with microphones that are limited in what they can do.

 

[CG]

"Then why does the new PS Audio PWT/PWD sound so (extremely) good utilizing their proprietary I2S interface?"

 

Since you've heard this product, could you describe what separates it from the rest of the contenders you've heard? I'm very curious... Any comparisons?

 

[Audio_ELF]

Timequest ... You asked "Then why does the new PS Audio PWT/PWD sound so (extremely) good utilizing their proprietary I2S interface?"

Maybe the operative word here is proprietory. Also from things I've read there is a question as to how well the i2s link works when usinging High Resolution material - where tollerences are going to be tighter over they syncronisation of clock and data signals. Anyway, all you (or anyone has said) is that it sounds best with i2s - maybe that's because the other outputs from the transport are not so good. You can't judge except as a whole so you haven't proved i2s is a better interface generally, just on a specific device.

 

CG - wasn't SPDIF a "consumer ready" adaptation of AES? It was my understanding from the beginning CD players were designed with capability of adding an external DACs. Is not SPDIF part of the "Red Book" standard? Meridian were one of the first with an offboard DAC (IIRC) connected to the first commercial Phillips CD player. Or did you mean that Sony / Phillips adapted it from a test signal interface?

 

Eloise

 

[The Computer Audiophile]

We always need to consider the marketing aspect of using one interface or another. For example, if a DAC chip accepts an I2S signal, then wouldn't it be great to offer consumers an unconverted I2S signal directly from the transport? That's not a question seeking an answer on the forum here :~) I can see consumers thinking this is the best interface ever because there is "no" conversion to the layman.

 

[idiot_savant]

I'd agree with Chris, hi-fi has long been about the shortest signal path, fewest components in the way, which are all truisms for a lot of aspects.

I'd point people to Einstein, here: "Everything should be made as simple as possible, but no simpler"

in which I'd say the I2S between boxes is "too simple".

 

As for the PS audio player, never heard one, or even seen one in the flesh, but have seen the review in Hi-Fi News - they seemed to think it sounded great, but interestingly it's "Native Mode" via I2S had loads of jitter - the upsampler mode ( which it looks like it feeds through an ASRC, like benchmark ) looks OK, and lots of apologies in the measurements for "Native Mode" being poor

 

your friendly neighbourhood idiot.

 

[Tog]

My letterbox accepts junk mail and pizza leaflets but thats not what I want. Simply because a dac can accept I2S doesn't mean it will be better.

 

yours, acceptingly,tog

 

[Audio_ELF]

I wasn't sure if Chris's comment "For example, if a DAC chip accepts an I2S signal, then wouldn't it be great to offer consumers an unconverted I2S signal directly from the transport?" was suggesting that an audio company adding an i2s connection was more to do with marketing and being apart from the crowd than actual engineering sense.

 

Eloise

 

[CG]

"CG - wasn't SPDIF a "consumer ready" adaptation of AES?"

 

I was led to believe that this interface was originally added so that the Sony and Philips engineers could have a way to test the transports of the original CD players. Then it took on a broader application. "That's not a bug - It's an undocumented feature! Let's sell it!"

 

But, that could just be some urban legend.

 

In any case, this was all done in the era of "perfect sound forever" when the standards were first adopted. Make no mistake - all of the standards for CDs were based around what could be done and what could be sold, right then. (Some time take a wander through Google to discover the history of how the bit rate and all that was really selected.) Commercial considerations were the top priority.

 

Just as with vinyl, modern audio technologists have gotten far better performance out of each media storage form than was ever thought of by the originators. Every now and then I marvel at just how really good sound reproduction actually is. It's far better than any of us have any right to expect it to be. It may not be the same as the real thing, which is probably good since I wouldn't really want Keith Moon playing the drums in my living room, but it's far better than just a vague resemblance of a live performance. (This next part will likely get stricken, so read it while you can...) I'll put it this way - it's far closer to the real thing than pornography is to its counterpart.

 

[idiot_savant]

I'm pretty sure that is what Chris was saying, Eloise.

 

To some ( a large extent ) you have to sympathise - given that most of the hardware between DACs/Transports is so similar ( CD-Pro2 for transport, Burr-Brown DAC/Analog Devices DAC in the DAC ), apart from the metalwork/number of DACs/valves, how does one manufacturer distinguish themselves from another in the market place? Actually doing the job properly ( proper analogue stuff on the output stages, actual care taken over clocking the DAC ) doesn't really appear on checklists - every DAC/transport manufacturer will say they have "excellent" output stages & "low jitter" ( though surprisingly few do )

 

It just annoys me that the supposed advantage of I2S between boxes is to lower jitter, where I have never seen an I2S implementation of this kind that does anything but exacerbate it!

The PS audio implementation is, I think, proprietary in that it uses an HDMI cable - this is actually not a bad idea ( if you're set on I2S,anyway) , as at least HDMI cables/ports need to have excellent electrical characteristics because real HDMI is super fast.

 

your friendly neighbourhood idiot

 

[cfmsp]

 

 

"and lots of apologies in the measurements for "Native Mode" being poor"

 

Well, if it ain't one thing, it's another.

 

so, perhaps there's more than bluster in Gordon's thoughts on the use of i2s as a protocol.

 

I for one am not surprised that a proprietary implementation of I2S could be made to sound acceptable, but for what purpose, and at what expense?

 

My guess is that the 'purpose' has as much to do with marketing (we've got this new whiz-bang approach that will render all others obsolete) as with engineering. In an already crowded field, what's a poor engineer to do. And a dogged engineer could likely succeed. As reference, consider Gordon's implementation of Async USB. No doubt, prior to his success with Async, it was thought that USB was NOT an audiophile-worthy interface (if one wants the highest quality). Apparently, some still think so.

 

As for the expense, well a $6000 transport/dac combo ought to sound good, indeed it ought to sound CONSIDERABLY better than anything costing just half as much, let alone the number of good sounding DACs that cost $1000-$1500, one quarter to one sixth. [Yes, I know, the $6k also includes a transport, but it's redundant in my view, as most of us already have a quite sufficient 'transport', keyboard included.

 

I'm convinced that with unlimited budget, it's relatively easy to get (either build or buy) good sound. For reference, consider dCS. They can make a SOTA device with any protocol known to mankind. But being of limited means, I'm not interested in SOTA assaults, I'm more interested in the most intelligent design approach, i.e., the approach that will most easily produce the best quality sound reproduction. I don't believe I2S meets the standard set by Firewire, Async USB (and perhaps networked players) with regards to the probability of getting highest quality sound for a reasonable cost.

 

When Gordon says this or that protocol sucks (my words, not his), I take that to mean that it requires considerable extra effort, and therefore expense, to reach high standard of quality and therefore is likely being utilized for reasons other than it being the best engineered approach.

 

your mileage may vary, and hopefully does, as then we'll have something to discuss.

 

clay

 

 

 

 

 

 

 

[Audio_ELF]

Moving on from i2s vs SPDIF, where direct FireWire isn't an option, is there any advantage in using Dual AES with 24/192 (and 24/176.4) rather than single wire AES. There's not many interfaces tht support it on a computer (Lynx AES16 and AFI1 come to mind and SonicStudios Model 303/4 in the future) and a few high end DACs, but does it reduce jitter issues or increase other problems?

 

Eloise

 

[idiot_savant]

Hi Eloise,

 

in theory anyway, by having the clock rate effectively half the data rate, the absolute level of jitter could be as much as halved.

I'm not sure it will in the real world, but the fact that most implementations of SPDIF/AES are optimised at something less than 192k, I would imagine there would be some improvement, depending on implementation as always.

The "other problems" I imagine would be to do with implementation ( people not agreeing how the data should be spread across the two wires, for instance ),

 

your friendly neighbourhood idiot

 

[The Computer Audiophile]

I've heard that 176.4 / 192 over single wire may not even be within the AES spec. Can someone verify? I know dCS requires dual wire for those sample rates.

 

[manisandher]

Esoteric and PMI also require dual wire at these sample rates.

 

What sample rate can the Berkeley DAC handle with its AES?

 

Mani.

 

[The Computer Audiophile]

The Berkeley Audio Design Alpha DAC handles up to 192 via single wire AES.

 

[timequest]

There is a lot of misinformation being bandied about here regarding the “Perfect Wave” devices. I suggest if anyone is interested that they check out the forum at PS Audio. The reviewer in HiFi News – according to PSA, did not measure jitter after the Digital Lens (rated at “well below 1ps). While the external I2S interface is certainly an interesting design aspect of the PWT/PWD, it is the Digital Lens that is the truly remarkable feature of said products (the Bridge forthcoming).

 

As for the sound, it is stunning – simply incredible. We compared it directly to ARC’s latest (the CD 7??), the Transporter, the PS Audio DL-III and the DL-III-Cullen Stage IV mod. I am very familiar with products like the DAC Magic, DAC-I Pre and others in the $1,000 to $5,000, including Bel Canto Products. To compare the aforementioned with the new Perfect Wave combination is simply not a fair comparison. This is truly something special. Listen for yourselves….

 

 

[The Computer Audiophile]

"There is a lot of misinformation being bandied about here regarding the “Perfect Wave” devices. I suggest if anyone is interested that they check out the forum at PS Audio. The reviewer in HiFi News – according to PSA, did not measure jitter after the Digital Lens (rated at “well below 1ps)."

 

To be honest, the comment about jitter rated well below 1ps is absolutely wrong. I'm not aware of how one can measure jitter that low with the current tools in use. Plus, measuring jitter itself is incredibly difficult to do accurately and not for marketing purposes. I'm not saying this one done for the PS products. I'm just stating the facts as told to me by senior AES fellows who are some of the most respected engineers in audio.

 

[idiot_savant]

Do they claim that the digital lens is in the transport?

As I read it, the measurements taken were transport->dac, via I2S, then you can't complain about jitter measurements taken at the point it matters ( the DAC output ) rather than where it matters less ( output of the transport ).

If the digital lens is in another box, and improves things, then they should have supplied one to the reviewer, surely?

 

I will say, that the jitter measurements ( while not very good ), would be the best I2S measurements I've seen.

 

oh, and 1ps.... I'd be extremely impressed, but I can't see it happening...

 

 

your friendly neighbourhood idiot

 

[timequest]

This is what is being stated by the man himself at the PSA website. I believe they use the manufacturer’s specs for the hardware (less than 1ps). I recall it being said that they couldn’t measure anything that low, but they do make the statement about “well less than 1ps jitter.

 

Look, the thing sounds fabulous and my initial response was to Gordon, who made a really poor decision to post such a statement on an open forum. I’m not smart enough to understand whether, or not the reason the Perfect Wave devices sound so good has anything to do with the implementation of I2S – according to PSA it does. The concept of marketing is not lost on me since that is what I make my living at.

 

Why such a resistance on this site to embrace what the whole industry has been abuzz about? I’m telling you, these things are in a different league. And they don’t cost $70,000 like some of the ridiculous digital front end systems available today.

 

 

[idiot_savant]

I have no problem believing that the PS audio stuff sounds great. I do have a problem with a manufacturer asserting an impossible figure that they can't measure themselves. Especially if, on (first review,admittedly) the measurements show something else? I've gone on record here with my views on I2S, so I'll say no more about it.

 

I'm not sure what resistance you're perceiving - unless it's resistance to marketing claims not being met by measurements?

 

As for $70,000 front ends, I've heard (on other forums) about the latest stereophile, but haven't read it. I'd agree, it sounds rather expensive...

 

your friendly neighbourhood idiot

 

[Purite Audio]

Why am I am always the last to know about the latest buzz!

 

[idiot_savant]

Should the industry be "buzzing" with the latest output from a company who makes power conditioners?

 

Sorry,

 

couldn't resist,

 

your friendly neighbourhood idiot

 

[The Computer Audiophile]

I don't have any resistance to the PWD and PWT. I've been writing about them since before CES in January 2008. I love the concept and I'm always happy when companies produce products that are fairly inexpensive and please fellow audiophiles.

 

I am still a skeptic of claims that seem to be contradictory to what I've been told by many other people much more learned that I. Hopefully this skepticism will lead to a great answer from someone who could back up the -1ps measurement or someone who knows a lot about the PDW & PWT implementations. I'm just out for more and correct information no matter if it's good, bad, or indifferent. Marketing speak of -1ps is just too hard for me to believe and I'll need some education as to why I2S is good on the PWD / PWT and not used on much else in high-end audio. That's all.