• 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?

Propheticus

Senior Member
Joined
Nov 18, 2020
Messages
431
Likes
644
Location
Vleuten, Netherlands
It was never meant to be used as hi-fi / high-end quality
Neither were phone lines or Tv cable ever meant for high speed internet. 1Gbit/s cable internet works fine.
Never was molded bread meant for medical purposes, yet it (penicillin) works fine.
Never was the internet (darpanet) meant for sending cat videos....yet we all know what happened.

Data is data, as long as you buffer and dont have buffer underruns (do your own clocking) what's the problem?
 

Gradius

Addicted to Fun and Learning
Joined
Aug 17, 2019
Messages
655
Likes
419
Location
Iquique, Chile
Neither were phone lines or Tv cable ever meant for high speed internet. 1Gbit/s cable internet works fine.
Never was molded bread meant for medical purposes, yet it (penicillin) works fine.
Never was the internet (darpanet) meant for sending cat videos....yet we all know what happened.

Data is data, as long as you buffer and dont have buffer underruns (do your own clocking) what's the problem?

You are confusing a lot things here.

First, Internet today is very different from old days (from 80s and 90s). Even TCP/IP changed a lot in last decade (thanks to IETF). In 1991 I was already using Internet, you have no idea how it was back there... web/http was a DREAM back there. The best example is IPv6, it wasn't easy at all to making to work.

The very same for TV cable, isn't the same it used to be, they needed to change a lot things, mainly the protocol. Or you think the OLD tv cable would works with Internet just fine? Just plug & play? Of course not.

Your concept is pretty wrong. Isn't JUST "data as data".
 

Victor BR

Member
Joined
Aug 16, 2021
Messages
33
Likes
10
i2s was NEVER created to transmit over long distances. It was designated to be used inside on a PCB and stay there.

You already know i2s as a mean of transmission (long distances) is WORSE than USB. Why keep asking?!
No need to be rude my fine friend. Just to get the definitive answer to PS AUDIO claim that their I2S input is the pot of gold. And yes, short silver HMDI cable connecting their universal transport to their dac has a very good result and please remember test bench oscilloscope do not tell the whole story.
 

Gradius

Addicted to Fun and Learning
Joined
Aug 17, 2019
Messages
655
Likes
419
Location
Iquique, Chile
Would be curious to see AES compared to Spdif.

Almost no differences:

- Better lengths;
- Generally allows recording (normally spdif is protected by default);
- Better EMI resistance (but since is in digital domain, I doubt it will means anything in real world).
 

MDMD

Member
Joined
Jul 16, 2021
Messages
6
Likes
0
Almost no differences:

- Better lengths;
- Generally allows recording (normally spdif is protected by default);
- Better EMI resistance (but since is in digital domain, I doubt it will means anything in real world).
Thanks!
 

Roland68

Major Contributor
Joined
Jan 31, 2020
Messages
1,353
Likes
1,122
Location
Cologne, Germany
i2s was NEVER created to transmit over long distances. It was designated to be used inside on a PCB and stay there.

You already know i2s as a mean of transmission (long distances) is WORSE than USB. Why keep asking?!
Sorry, but the external i2s interfaces with HDMI connectors use LVDS as the physical layer. 2m, or even more if the required cable specification is complied with, are no problem. LVDS was developed precisely for such purposes and is used in many places in industry and computer technology (SATA, SCSI, PCIe, HyperTransport, DisplayPort, SpaceWire, RapidIO, etc.).
You cannot connect this interface directly to a normal onboard i2s interface either. An HDMI I2S LVDS to I2S converter module (or vice versa) is required for this.
 

Propheticus

Senior Member
Joined
Nov 18, 2020
Messages
431
Likes
644
Location
Vleuten, Netherlands
You are confusing a lot things here.

First, Internet today is very different from old days (from 80s and 90s). Even TCP/IP changed a lot in last decade (thanks to IETF). In 1991 I was already using Internet, you have no idea how it was back there... web/http was a DREAM back there. The best example is IPv6, it wasn't easy at all to making to work.

The very same for TV cable, isn't the same it used to be, they needed to change a lot things, mainly the protocol. Or you think the OLD tv cable would works with Internet just fine? Just plug & play? Of course not.

Your concept is pretty wrong. Isn't JUST "data as data".

I'm not really. Hopefully you understand I know very well that all those examples are not used in their old original form. The point is they can/are used for other purposes than their original one...with some (or a lot of) changes.

The same is true for i2s over HDMI. Not originally meant for it, but works.
If it's your best choice for transport inter_device or whether it has any benefits over usb is another matter.
 

Jimbob54

Grand Contributor
Forum Donor
Joined
Oct 25, 2019
Messages
11,063
Likes
14,694
Any chance to measure PS AUDIO DAC over I2S input x USB????
No. The ps audio dac like almost everything else measured here will have been sent in by a member, long long ago. What is the reason anyone would send it back in?

What do you think would change if the same signal is sent to any dac via i2s vs USB or indeed spdif? There may be small measured changes but I highly doubt an audible difference unless one of the other inputs is broken.
 

audio2design

Major Contributor
Joined
Nov 29, 2020
Messages
1,769
Likes
1,830
please remember test bench oscilloscope do not tell the whole story.

When we are discussing signal recreation and not preference, yes, yes it does .... But we use high performance audio analyzers not oscilloscopes. Just mentioning an oscilloscope reveals a common misguided talking point not a real argument.
 

Victor BR

Member
Joined
Aug 16, 2021
Messages
33
Likes
10
So, what should be the best DAC CHIP actually based on bench tests and what kind of implementation should cause less harm to its performance?
I always thought OP AMPS as the worst components in a DAC. Is it true guys?
I
 

EdW

Senior Member
Joined
Jul 5, 2020
Messages
329
Likes
413
Location
Cambridge, UK
So, what should be the best DAC CHIP actually based on bench tests and what kind of implementation should cause less harm to its performance?
I always thought OP AMPS as the worst components in a DAC. Is it true guys?
I
AKM4499 vs ESS9038 PRO? If the DAC hardware surrounding the IC are well implemented both are superb. The ESS chip may be a little more tricky to design with (remember the ESS IMD hump?) but if we compare the Topping D90 vs the D90SE (AKM4499 vs ESS9038 PRO) we see that the D90S has a few dB more SINAD - some of this is down to John Yang’s superb attention to detail of course! Will you hear any difference between the 2 DACs? Absolutely not!

The AKM4499 does however have 2dB headroom for intersymbol overs I believe (sorry don’t have a reference here) so its performance maybe a little closer to the ESS9038 PRO in a real world situation if we are splitting hairs here. But arguably a recording shouldn’t have intersymbol overs.

Sadly AKM had a serious fire at their fabrication facility and many of their ICs are unavailable.

Most State of the Art DAC implementations choose to use IC opamps for buffering etc because opamp designers have realised performance above that achievable with discrete transistor implementations and at a fraction of the cost. Stand-alone preamps using opamps like the TI OPA1612 usually display distortion levels lower than DACs using the same opamps and this is as might be expected from the data sheets so we can conclude that opamp performance should not be the limiting factor in DAC performance - the limitation is the DAC IC itself and of course the skill of the designer implementing the complete DAC.
 

audio2design

Major Contributor
Joined
Nov 29, 2020
Messages
1,769
Likes
1,830
So, what should be the best DAC CHIP actually based on bench tests and what kind of implementation should cause less harm to its performance?
I always thought OP AMPS as the worst components in a DAC. Is it true guys?
I

No, but truly competent analog design / designers is surprising rare. It takes more than just understanding the basics but being able to visualize error currents and how they impact performance. Due to the small market size and/or total margin in this market, most companies don't have the advanced PCB design tools let alone circuit analysis and certainly not integrated tools (nor knowledge to use) that can show where your performance flaws are most likely coming from.
 

Victor BR

Member
Joined
Aug 16, 2021
Messages
33
Likes
10
I stand with the crazy notion that numbers are more than measurements when it comes to audio. Some X Files is really going on here.
 
Top Bottom