• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Study: Is I²S interface better for DACs than S/PDIF or USB?

oryan_dunn

Member
Joined
Apr 24, 2021
Messages
31
Likes
41
External I²S is a solution lookin' for a problem, an answer to a question no electrical engineer was askin'...

Now, if you're the head of an audio company wondering what you can do to sell a new source and DAC and sell as a pair since it's a unique connection, then perhaps external I²S makes more sense...
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,626
Likes
10,202
Location
North-East
You never try it right so that your assumtion?

so arrogant for me that is

I have, and while I initially thought I heard a difference, I couldn’t reproduce it in a blind test.

Measurements showed higher noise level with DDC+I2S (SU-1 and USB100) than without, but well below audibility. That’s with two different DACs. I then modified SU-1 with improved voltage regulators and with new, lower phase noise oscillators and even powered them from a battery. Still nothing audible, and still somewhat worse measurements. Sorry.
 

sumpao

Member
Joined
Mar 29, 2021
Messages
48
Likes
4
I have, and while I initially thought I heard a difference, I couldn’t reproduce it in a blind test.

Measurements showed higher noise level with DDC+I2S (SU-1 and USB100) than without, but well below audibility. That’s with two different DACs. I then modified SU-1 with improved voltage regulators and with new, lower phase noise oscillators and even powered them from a battery. Still nothing audible, and still somewhat worse measurements. Sorry.

Thanks this is a quality comment.

Could you help me why below YT said there is an improvement in measurement when use Pi2AES

Do you think it device or measurement method?

 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,626
Likes
10,202
Location
North-East
Thanks this is a quality comment.

Could you help me why below YT said there is an improvement in measurement when use Pi2AES

Do you think it device or measurement method?


I've not looked through the whole video, but it appears that Pi2AES is helping remove some more of the interface noise than the more expensive DDC. That's entirely possible, if it has better power regulation and better isolation circuits. At least one of the shots actually shows increased jitter close to the test signal, but fewer random spikes away.

At the end, you can see that May DAC has lower jitter and noise when using its own inputs rather than any DDC over I2S when the PLL is engaged. One of two the DACs I tested with DDCs and I2S was Holo Spring, the predecessor to May. And even Spring over its own USB performed better than with the DDC+I2S.

You need to be careful about what inputs you're comparing. I think for most of the tests in the video the AES input was used. There is a possibility of introducing jitter and noise over AES and SPDIF coax if the receiver circuit in the DAC is not well designed. Optical has a much better isolation qualities and doesn't let the noise through, while USB has a nearly perfect source jitter reduction. But even in the video, the level of the differences is well below any audibility. The subjective reporting that the DDC made an audible differences is extremely unlikely to be real. The difference between -139dB and -142dB is simply not audible to anyone and can be due to a simple variation from one test session to the next due to external noise sources.
 

sumpao

Member
Joined
Mar 29, 2021
Messages
48
Likes
4
I've not looked through the whole video, but it appears that Pi2AES is helping remove some more of the interface noise than the more expensive DDC. That's entirely possible, if it has better power regulation and better isolation circuits. At least one of the shots actually shows increased jitter close to the test signal, but fewer random spikes away.

At the end, you can see that May DAC has lower jitter and noise when using its own inputs rather than any DDC over I2S when the PLL is engaged. One of two the DACs I tested with DDCs and I2S was Holo Spring, the predecessor to May. And even Spring over its own USB performed better than with the DDC+I2S.

You need to be careful about what inputs you're comparing. I think for most of the tests in the video the AES input was used. There is a possibility of introducing jitter and noise over AES and SPDIF coax if the receiver circuit in the DAC is not well designed. Optical has a much better isolation qualities and doesn't let the noise through, while USB has a nearly perfect source jitter reduction. But even in the video, the level of the differences is well below any audibility. The subjective reporting that the DDC made an audible differences is extremely unlikely to be real. The difference between -139dB and -142dB is simply not audible to anyone and can be due to a simple variation from one test session to the next due to external noise sources.

I find that first paragraph you talk about video

second and third paragraph you talk about your own test right?
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
You never try it right so that your assumtion?

so arrogant for me that is
Now you sound arrogant. This topic, I send it in, the Singxer* I2S/DDC device. But how nice of you to assume.
 
Last edited:

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
So you did own one and it not good?
Yep. I used it as an AES transport in the end, works very well at that. I saw no use for the I2S since it seemed to me just the same as direct USB input to a DAC. A good DAC should have no problems with its digital inputs :)

I'm sure the sales people would love to tell you you'd need a device like this and that it makes a difference though. I disagree.
 

sumpao

Member
Joined
Mar 29, 2021
Messages
48
Likes
4
Yep. I used it as an AES transport in the end, works very well at that. I saw no use for the I2S since it seemed to me just the same as direct USB input to a DAC. A good DAC should have no problems with its digital inputs :)

I'm sure the sales people would love to tell you you'd need a device like this and that it makes a difference though. I disagree.

My dac don't have AES/EBU want to try it too. so what about AES and USB direct did they have an improvement?

I assume you use U16
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
My dac don't have AES/EBU want to try it too. so what about AES and USB direct did they have an improvement?
There are powered speakers that have AES inputs that's all I wanted it for :p:D no such other improvements for me.
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,626
Likes
10,202
Location
North-East
Ok I get it now

In the video he did mention about clock did it matter?

Clock in a DDC can potentially matter if the DAC is poorly implemented. But, it is easier to implement a quality clock inside the DAC than to source it from another, external component (DDC) over some, unknown quality I2S interface and cable.

In the video, he compared DDC I2S to May DAC with the PLL turned off. That’s not a valid comparison because that disables a key jitter removal circuit in the DAC. When compared to the PLL-enabled May, the DDC produced more jitter.
 

sumpao

Member
Joined
Mar 29, 2021
Messages
48
Likes
4
Clock in a DDC can potentially matter if the DAC is poorly implemented. But, it is easier to implement a quality clock inside the DAC than to source it from another, external component (DDC) over some, unknown quality I2S interface and cable.

In the video, he compared DDC I2S to May DAC with the PLL turned off. That’s not a valid comparison because that disables a key jitter removal circuit in the DAC. When compared to the PLL-enabled May, the DDC produced more jitter.

In video he did mention TCXO ,OCXO and I found that there are another clock GPSDO which more accurate

DDC have so much for debate but he did mention he hear different (not mean it better)

I love to see conclusion someday
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,626
Likes
10,202
Location
North-East
In video he did mention TCXO ,OCXO and I found that there are another clock GPSDO which more accurate

DDC have so much for debate but he did mention he hear different (not mean it better)

I love to see conclusion someday

There's no question that the differences below -140dB that he measured in the video are completely inaudible. Claiming anything else requires a very carefully constructed and controlled experiment to demonstrate that this "super-human" ability is real -- it's simply beyond any known scientific fact and requires belief and faith, instead.
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
In video he did mention TCXO ,OCXO and I found that there are another clock GPSDO which more accurate

DDC have so much for debate but he did mention he hear different (not mean it better)

I love to see conclusion someday
Conclusion: obsessing over clocks is a fruitless endeavour not resulting in any happiness :p
 
Top Bottom