another one of those areas where we have measurable problems yet over the years some very fine sounding systems have used "bad" techique.
The SPDIF interface was one never meant to see duty outside the test bench and was not part of the original digital signal chain with the CD player being a "one box" format that used a direct digital streeam to the DAC chips.
The problem was like with ALL consumer grade gear the more you put in a single box (think HT receivers) the more shortcuts and cost saving measures are taken so there was a need for an "outside the box plug and play" device to replace the mostly crap internal analog sections of the average CD player.The first I know of was made by PS audio.
Shortly after there was a profusion of DACs touted as the single biggest improvement you could make to the sound of the CD which meant you had to somehow get the digital stream out of the player and to the new interface and by default it ended up the spdif connection even though the I2S was a superior method and was pushed by both Audio Alchemy and sonic frontiers.why did it not become the standard for external digital ?
Cost and laziness.Why bother adding expensive connectors and complicated circuits when any idiot could toss in a transmitter/receiver connected by a cheap ass RCA connector and used off the shelf coaxial cable ?
So the implentation of the "improvement" many times was no better than what was already inside the player but to say that would not do if you want to sell another gadget to the consumer who is mostly an idiot and will beleive anything if you complicate the ad copy enough.
The spdif connection
IS jitter prone but this can be minimised with correct designs and that means not only correct impedances,true 75 or even 50 ohm coax connectors which is IMPOSSIBLE to have in an RCA because og the proximity of the center connection to the outter connection (75 ohms means it must be wide) plus there is cable length to consider and if it is too short is where the reflections enter in and why mine are all 2 meters even though the DAC sits within inches of my transports.
BTW-wanna know a dirty little secret ? All this could be avoided if there was a proper buffer storing the digital content before it was actually sent to the DAC chip input but that is too easy
Relying on the real time stream means playing back any problems real time while playing from a re-clocked buffer eliminates that particualr problem and why computers are potentially superior to all but the best CD players.
So flawed yes but still we manage to get damn fine sound which says good implementation of even a flawed idea is better than a hack using the best of technology or ALL digital would sound identical being a ones and zeroes format.
USB is a recent addition and one that at first was no more than an easy method to get past the mostly crap sound card though if you read back into the history of it was maybe the most ill used audio technology of all time with the norm being crap speakers with a dedicated crap internal amp and a USB input before someone realised there was potential for good sound and that only in the last two years or so even though USB has been with us as a viable format for close to a decade !
The first device that impressed me as a true high end audiphile type device was made by Wheatfield Audio in 2001 and though it never went into production I actually have the product literature that I grabbed when I auditioned a prototype at the 2001 Stereophile sponsored "Home Entertainment Show" at the N.Y. Hilton.
I was impressed,very excited and was impatiently waiting for a product that never seen full production so i went and built my own imitation using no more than the clues from the literature in combination with the Burr-Brown data sheets for the PCM2702.
That was 2001 yet the idea of USB as a serious format fell off the planet until revived by Gordon Ranking of Wavelength Audio.When it was he who touted this as SUPERIOR in combination with a computer based sound system to any CD based or Transport/DAC based system USB was "discovered" even though Pete Millett had already shown the way years earlier.Just shows what print/internet exposure will do on whether something is noticed or not or even considered a "good thing" or not.
Anyway.
Millett used off the shelf stock parts and as far as I know pretty much followed the eval board schemo/layout using high quality parts but Wavelength does NOT use a plug and play USB chip like the ones found in 99.9% of the available devices.
Having a background in programming he used chips that actually take a bit of work to get going rather than the "solder it in,plug it in,play it" devices we DIY hobbyisits mostly use.He took this "new" interface and mated it to the "old" already in the product lineup Wavelength DAC and because the DAc was already one with great sonics the product well received.
Now every DIY audio forum has at
least five or six design threads active and every knucklhead who can use circuit board CAD software is an audio design engineer which makes the general observation "why should I pay so much for the Wavelength DAC when all these other USB DACS cost pennies to the dollar" when they should be asking "how did he manage to make such a fine USB interface usable with a quality dac ? what chips ? What programming ?What software model ?"
So guys like me are pretty much stuck with the off the shelf BB/TI chips and concentrate on making that sound good while true pioneers do things their own way,as it always has been,but that does not mean these simple little boxes can not sound great !
I have sitting before me on the bench right now waiting for a new round of prototyping three PCM2704 and three PCM2707 chips.The prior will be my best attempt at an all in one portable USB to Analog inteface/home USB to SPDIF box,the second my final attempt at a dedicated USB-I2S-AD1861 DAC but to be honest,all the added complexity of the USB/I2S/Analog is not a definitive path to better sound over the simple PCM2704 "just use it" interface as the Fubar DAC proves.sometimes imple is just best even though on paper another method may SEEM better.
It is the implementation that counts and if that means making the best possible device with a seemingly inferior part (on paper) over a bad implementation of a
potentially superior device the better way is a no brainer and why I crack up laughing when I read folks that have no clue what is involved and who have never actually did a head to head LISTEN proclaim one device superior to another just because the specs say it should be so.
don't get caught up in the trap man.Down that path leads a lot of loot thrown at a problem that never existed with the race to sonic perfection mostly a theory none of us have a shot at attaining unless we get past the ad copy and even then it comes down to personal taste and system requirements.
My very first "lame" PCM2702/OP627 battery operated USB DAC has yet to be bettered in the "real world" in my system and my only mistake was getting rid of it before I realised the potential of USB as a viable high fidelity medium,a recent "discovery" that came four years late for me