And then God said, "let there be light..." Thank you Paul for breaking this down into the simplest yet easy to understand way possible ... this was the BEST explanation I have heard so far (even from your channel), and now I fully understand why it's imperative for me to extend additional monies to secure an I2Ss capable DAC. Incredibly helpful. Thank you!
In a perfect world, the DAC runs a highly precise ultra low jitter clock without need of PLL or synchronizing with any external clock, and data is pulled asynchronously on-demand. This is how music and video streaming can work perfectly over the internet without jitter or drop-outs, even with non-audio gear feeding the data. Internet (IP) via whatever connection (ethernet, Wi-FI, optical etc.) supports asynchronous audio reliably. Also USB supports asynchronous audio reliably. Neither SPDIF, TOSLink or I2S supports asynchronous audio data. When a source device, such as a CD transport, wants to play "clock master" without actually having the direct control of the precise tick needed for jitter-free playback within the DAC, you end up relying on PLL and have to dance around managing two "clock masters". In the past where digital audio came from a source device that could not work as clock slave (such as a CD transport), you ended up with a synchronous interface to the DAC. So please consider the most audiophile architecture of tomorrow's digital streaming optimized system is end-to-end asynchronous data transfer with a precise DAC clock running the timing and no messing around with multiple clocks trying co-operate. This topic is a bit like how we buy food and eat it. Imagine your grocery store (source device) decides when to send you (DAC) food (audio data) without ever listening to you and you have to eat it (convert it to audio) when you get it. Your would have to accept eating at the rate of when the grocery store send you food without really knowing the precise timing of it. This is how SPDIF, TOSLink and I2S work. Then imagine an alternative where you could order food in advance (asynchronous) and always buffer enough food (audio data) to precisely have your meals as scheduled by your own clock. This is how asynchronous audio data transfer works over the internet and USB (and it's how we want to manage our meals with our own clock). A PLL is sort of like assigning someone to re-schedule your meals to whatever time you happen to receive your food. Yes there are ways to make synchronous audio jitter-free with more buffering and other techniques, but best is to run 100% asynchronous from source of music data to DAC. A correct asynchronous architecture also completely eliminates any need for a "re-clocker" device.
@@clickbeetle2720 No, I just like to express my opinion and hope some day I can buy streaming optimized audiophile gear rather than gear optimized to play vinyl and CD also able to support streaming without compromises. I have several systems and realize my KEF LS60 are heavily optimized towards streaming while my other systems have streaming as a less integral part of it. I2S between a streamer and DAC is less ideal than a USB interface where my streamer can feed the data asynchronously and with two-way awareness of the DAC capabilities (including DSD support). In the ideal world even DSD streaming is completely seamless and I just pick a track and if it happens to be DSD, it will play lossless in DSD. Without that sort of integration, DSD will never succeed.
I2S isn't a standard and was only meant for board level interconnections! My suggestion is to try I2S and compare it to your other connections and use the one that sounds better to YOU!
@@dontcare563 Yes, I2S is for Inter Integrated Sound transfer inside a product such as phone, tablet, CD player, TV or automotive head unit. There should be no sound to a digital audio interface provided it is engineered correctly avoiding data errors and audible jitter. There is no magic to it. Your online banking also don’t need a banking optimized Ethernet cable or router 🧐
@@Pete.across.the.street My point is that whatever interface is used to transfer music data to your DAC needs to transfer the data without errors and in time for the DAC to run the conversion clock precisely. Neither of that is hard nowadays and if one music interface sounds different than another interface, assuming the music source is identical, we have a problem of an engineering bug. Audible differences don't arise out of no root cause. And with similar reasoning, online banking doesn't need golden ethernet cables 🙂
My understanding is I2S (e.g. using an HDMI cable as an example) is the ONLY way to decoded DSD layer of the SACD... and there is not industry standard pinout hence a bit challenging to find affordable I2S capable DSD DAC...
I found it depends on the gear, sometimes i2s sounds better, sometimes coaxial does... It is system dependent in many cases, even with some PS Audio gear, let your ears decide.
Our entire hobby in all its configurations should, in theory, be boiled down to this simple statement. Too often, we want what sounds good or bad to our specific ears to be the same for everyone else’s ears - or our own musical taste and interest to be what others like too. And if not, we should be inclined to disagree without being disagreeable. That’s how things should work but too often they don’t, so it’s nice to have this reminder. 👍
Sounds like a multiplexer - demultiplexer when you’re describing strictly digital. A digital multiplexer is nothing more than a input signal selector, sort of the equivalent of a switch and a Demultiplexer does exact the opposite.
I'm confused. I thought that digital on XLR was AES/EBU. I've never seen a PS Audio equipment with that implementation but I might be wrong. So if he's talking about XLR (analog) how is that compared with I2S? Unless is trying to understand which option would be better, coming out with XLR from PW SACD or using it as a transport by connecting both units I2S with a HDMI cable. I'm lost.
Your initial thought is correct. This video wasn't specific that way. XLR (analog) is not apples-to-apples with I2S. XLR (digital) -- a.k.a. AES/EBU -- is. That all said, if we exclude XLR (analog) from the conversation, Paul is essentially saying that I2S is better than AES/EBU because there is no multiplexing when using I2S.
IMHO @shuntachi is fully correct: in addition to Paul's explantions, I2S is also natively some kind of balanced. Anyway, the HDMI cable quality matters according to my own experience, the best is to find one designed for audio I2S, e.g. those proposed by Italian Company RICABLE (pls note: I'm not linked with them at all !).
Also, you can use a HDMI cable but there are also specific I2S cables with HDMI connectors that only utilize the four (iirc) wires needed for I2S instead of all the wires.
Coax rca/bnc always sounds most natural to me. Its all just marketing to sell more with the latest and greatest input for the latest and greatest resolution when 99% of your music library is even recorded at that resolution. Upsampling when the information is not even there to uncover. In the 90s, it was fiber optic(not toslink) in the uber high end before it was phased out because too expensive, replaced by the cheaper toslink which became common and that stuck around despite being universally hated. Im no expert, but Fiber optic used glass and lasers vs plastic and led light transmission. (i know they make toslink in glass as well). Shouldn't that be the real best way to do digital connections even today? I believe some of the uber high end dac makers use it still. Off the top of my head i want to say Playback Designs and MSB? They also pass DSD. They now do seperate clock and signal with multimode vs the single mode fiber optic that was used in most of the 90s devices. I suggest yall give toslink another try with a solidly built glass cable, BUT clean both tips(they sell cleaning pens that clean both the cable tip and the inside of the connector), and use coupling gel on both ends to reduce losses. It's a gel you just dip the tip ends in and then stick into your connector.
AES/EBU ?? over I2S for CD Transport >>> DAC >>> Pre-Amp>>Power Amp - *N.B. I do *NOT* stream audio*...* Also can anyone, please advise* - Would using a DDC in the chain improve the sound?, i.e., CDT >>DDC>>DAC>>PRE-Amp >>POWER Amp = Better, clearer AUDIO QUALITY? (I 'might' try streaming at a later date... but it's not on my current - 2 years into future agenda" PRESENTLY = C.E.C. 2-Belt CDT>> Possible Denafrips IRIS or GAIA DDC >> Denafrips Venus 12th DAC >> ROKSAN BLAK as Pre-Amp >> ROKSAN CASPIAN M2 Power Amp >>>PMC's *I live in Japan, so a lot of Brands are either >unavailable< or TWICE/THREE times the cost in UK/N. America/Mainland Europe*. *All advice and/or suggestions - *not too technical, please*... *I'm a retired Mech. Engineer (60 + yrs..... young ; - ) * N.B. # 2 - I do *NOT* listen to SACD or HD CD's.... and do not intend to at any point in future* *Many Thanks!!*
And then God said, "let there be light..."
Thank you Paul for breaking this down into the simplest yet easy to understand way possible ... this was the BEST explanation I have heard so far (even from your channel), and now I fully understand why it's imperative for me to extend additional monies to secure an I2Ss capable DAC.
Incredibly helpful.
Thank you!
It was probably an AES/EBU vs. I²S question...
Very educational. Thanks, you filled in a few blanks for me there.
In a perfect world, the DAC runs a highly precise ultra low jitter clock without need of PLL or synchronizing with any external clock, and data is pulled asynchronously on-demand. This is how music and video streaming can work perfectly over the internet without jitter or drop-outs, even with non-audio gear feeding the data. Internet (IP) via whatever connection (ethernet, Wi-FI, optical etc.) supports asynchronous audio reliably. Also USB supports asynchronous audio reliably. Neither SPDIF, TOSLink or I2S supports asynchronous audio data. When a source device, such as a CD transport, wants to play "clock master" without actually having the direct control of the precise tick needed for jitter-free playback within the DAC, you end up relying on PLL and have to dance around managing two "clock masters". In the past where digital audio came from a source device that could not work as clock slave (such as a CD transport), you ended up with a synchronous interface to the DAC. So please consider the most audiophile architecture of tomorrow's digital streaming optimized system is end-to-end asynchronous data transfer with a precise DAC clock running the timing and no messing around with multiple clocks trying co-operate. This topic is a bit like how we buy food and eat it. Imagine your grocery store (source device) decides when to send you (DAC) food (audio data) without ever listening to you and you have to eat it (convert it to audio) when you get it. Your would have to accept eating at the rate of when the grocery store send you food without really knowing the precise timing of it. This is how SPDIF, TOSLink and I2S work. Then imagine an alternative where you could order food in advance (asynchronous) and always buffer enough food (audio data) to precisely have your meals as scheduled by your own clock. This is how asynchronous audio data transfer works over the internet and USB (and it's how we want to manage our meals with our own clock). A PLL is sort of like assigning someone to re-schedule your meals to whatever time you happen to receive your food. Yes there are ways to make synchronous audio jitter-free with more buffering and other techniques, but best is to run 100% asynchronous from source of music data to DAC. A correct asynchronous architecture also completely eliminates any need for a "re-clocker" device.
Or put another way, the writer doesn't claim it has any audable effect, so we agree then!
@@clickbeetle2720 The difference is if your DAC pulls the data asynchronously on-demand or someone else is pushing the data synchronously.
I don’t care, I just want beautiful music. 🔥👍
@@clickbeetle2720 No, I just like to express my opinion and hope some day I can buy streaming optimized audiophile gear rather than gear optimized to play vinyl and CD also able to support streaming without compromises. I have several systems and realize my KEF LS60 are heavily optimized towards streaming while my other systems have streaming as a less integral part of it. I2S between a streamer and DAC is less ideal than a USB interface where my streamer can feed the data asynchronously and with two-way awareness of the DAC capabilities (including DSD support). In the ideal world even DSD streaming is completely seamless and I just pick a track and if it happens to be DSD, it will play lossless in DSD. Without that sort of integration, DSD will never succeed.
@@DaleCrommie Yeah, I’m talking as an engineer but as consumer, I agree. When I enjoy music, it’s about the music itself and how it sounds.
I just purchased an IS2. Thanks, Paul.🔥
I2S isn't a standard and was only meant for board level interconnections! My suggestion is to try I2S and compare it to your other connections and use the one that sounds better to YOU!
@@dontcare563 Yes, I2S is for Inter Integrated Sound transfer inside a product such as phone, tablet, CD player, TV or automotive head unit. There should be no sound to a digital audio interface provided it is engineered correctly avoiding data errors and audible jitter. There is no magic to it. Your online banking also don’t need a banking optimized Ethernet cable or router 🧐
Wheels weren't meant for automobiles when they were invented, but here we are.
@@Pete.across.the.street My point is that whatever interface is used to transfer music data to your DAC needs to transfer the data without errors and in time for the DAC to run the conversion clock precisely. Neither of that is hard nowadays and if one music interface sounds different than another interface, assuming the music source is identical, we have a problem of an engineering bug. Audible differences don't arise out of no root cause. And with similar reasoning, online banking doesn't need golden ethernet cables 🙂
My understanding is I2S (e.g. using an HDMI cable as an example) is the ONLY way to decoded DSD layer of the SACD... and there is not industry standard pinout hence a bit challenging to find affordable I2S capable DSD DAC...
I found it depends on the gear, sometimes i2s sounds better, sometimes coaxial does... It is system dependent in many cases, even with some PS Audio gear, let your ears decide.
How about DSD layer of an SACD disc ?
And, "better" is different for everyone. . .
Our entire hobby in all its configurations should, in theory, be boiled down to this simple statement. Too often, we want what sounds good or bad to our specific ears to be the same for everyone else’s ears - or our own musical taste and interest to be what others like too. And if not, we should be inclined to disagree without being disagreeable.
That’s how things should work but too often they don’t, so it’s nice to have this reminder. 👍
@@housepianist And then you have the "Cult of Measurements"... Spot on sir.
@@briand.1694 It is. But many will disagree.
Good question from Mitch in Central Oregon. ✉Mazel tov.
Sounds like a multiplexer - demultiplexer when you’re describing strictly digital. A digital multiplexer is nothing more than a input signal selector, sort of the equivalent of a switch and a Demultiplexer does exact the opposite.
I'm confused. I thought that digital on XLR was AES/EBU. I've never seen a PS Audio equipment with that implementation but I might be wrong. So if he's talking about XLR (analog) how is that compared with I2S? Unless is trying to understand which option would be better, coming out with XLR from PW SACD or using it as a transport by connecting both units I2S with a HDMI cable. I'm lost.
Your initial thought is correct. This video wasn't specific that way. XLR (analog) is not apples-to-apples with I2S. XLR (digital) -- a.k.a. AES/EBU -- is. That all said, if we exclude XLR (analog) from the conversation, Paul is essentially saying that I2S is better than AES/EBU because there is no multiplexing when using I2S.
If you want to get really good, build the streamer and the DAC in one chassis and run I2S on U.FL cables not longer than 4".
Each signal, clock and data, of I2S is differential pair in the HDMI cable. It does reject the common mode noise.
You have no idea what you are talking about!
IMHO @shuntachi is fully correct: in addition to Paul's explantions, I2S is also natively some kind of balanced. Anyway, the HDMI cable quality matters according to my own experience, the best is to find one designed for audio I2S, e.g. those proposed by Italian Company RICABLE (pls note: I'm not linked with them at all !).
Does HDMI keeps things separate like I2S?
The HDMI protocol and I2S one are two different things.
I2S can use HDMI connectors, but that's it.
Unless you really ask about the HDMI protocol?
Also, you can use a HDMI cable but there are also specific I2S cables with HDMI connectors that only utilize the four (iirc) wires needed for I2S instead of all the wires.
Coax rca/bnc always sounds most natural to me. Its all just marketing to sell more with the latest and greatest input for the latest and greatest resolution when 99% of your music library is even recorded at that resolution. Upsampling when the information is not even there to uncover.
In the 90s, it was fiber optic(not toslink) in the uber high end before it was phased out because too expensive, replaced by the cheaper toslink which became common and that stuck around despite being universally hated.
Im no expert, but Fiber optic used glass and lasers vs plastic and led light transmission. (i know they make toslink in glass as well). Shouldn't that be the real best way to do digital connections even today? I believe some of the uber high end dac makers use it still. Off the top of my head i want to say Playback Designs and MSB? They also pass DSD. They now do seperate clock and signal with multimode vs the single mode fiber optic that was used in most of the 90s devices.
I suggest yall give toslink another try with a solidly built glass cable, BUT clean both tips(they sell cleaning pens that clean both the cable tip and the inside of the connector), and use coupling gel on both ends to reduce losses. It's a gel you just dip the tip ends in and then stick into your connector.
Would the customer's combination of PS Audio kit send the SACD layer content as DSD or PCM via the I2S connection ?
I believe DSD as SACD layer is DSD. PCM is CD Layer.
It's more of an a codec than a format..
So the take away is SPDIF does not work and you should not use it. 😅
AES/EBU ?? over I2S for CD Transport >>> DAC >>> Pre-Amp>>Power Amp - *N.B. I do *NOT* stream audio*...* Also can anyone, please advise* - Would using a DDC in the chain improve the sound?, i.e., CDT >>DDC>>DAC>>PRE-Amp >>POWER Amp = Better, clearer AUDIO QUALITY? (I 'might' try streaming at a later date... but it's not on my current - 2 years into future agenda"
PRESENTLY = C.E.C. 2-Belt CDT>> Possible Denafrips IRIS or GAIA DDC >> Denafrips Venus 12th DAC >> ROKSAN BLAK as Pre-Amp >> ROKSAN CASPIAN M2 Power Amp >>>PMC's *I live in Japan, so a lot of Brands are either >unavailable< or TWICE/THREE times the cost in UK/N. America/Mainland Europe*. *All advice and/or suggestions - *not too technical, please*... *I'm a retired Mech. Engineer (60 + yrs..... young ; - ) * N.B. # 2 - I do *NOT* listen to SACD or HD CD's.... and do not intend to at any point in future*
*Many Thanks!!*
Same question here! Thanks
Ddcs can help cheap dacs. I find better results just upgrading the clocks in the dac with something like the ian canada clocks
@@Pete.across.the.street Thanks!
I spy with my little eye a subwoofer.... is this a PS prototype?
Upon further review, look like REL
marketing keeps inventing things... whats wrong with good old optical and coaxial?
Won't work with SACD's DSD layer due to copyright.
@@ptg01 mmm ok thanks
First!
4th!!!!
Third.