"Audiophile" motherboards

[TonyJZX]

i'm not saying 'audiophile' motherboards arent good... they come at a cost and they come with a slew of "GAEMING" features I dont care about and at a quite elevated price...

I treat a pc or a laptop as just a means to store or stream music so hdmi or sp/dif out is all that is required... so an average motherboard that has all the features you want and none of the fluff you dont is all you need.

Its like how many phones have really excellent headphone output now... I mean that's nice but bluetooth has largely taken over... and really so has HDMI on just about every device.

 

[valerianf]

ESS DAC | Motherboard - GIGABYTE Global

www.gigabyte.com

For those who prefer an on board "audiophile" DAC.

 

[jhaider]

This looks like a very promising snake oil:

A PC with a linear power supply...

Didn’t watch the video, but I assume from the picture the nice little Mac is what you want and the large contraption is what you don’t want.

 

[Peluvius]

I did a bit of research before my last PC build and ended up with an X570 Master, which was promoted at the time as an "audiophile" capable solution (whatever that is actually supposed to mean). I couldn't find any source to output measurements however they did nominate the components used and these do have test data available. I just don't know how well these were implemented.

SS SABRE reference DAC (ESS 9118)
Hyperstream Dynamic Range (SNR 125dB)
High-Res music (32bit, 192KHz PCM)
Super low THD+N (THD+N -112db)
RealtekALC1220 Codec
WIMA and Nichicon Fine Gold capacitors
TXC OSCILLATOR

There is a good thread here as well:

Realtek ALC1220 - a look on PC sound

If you purchased new desktop PC in last 3-4 years you probably have one of those on your motherboard. Since I spent few last weeks trying to make it work as best I could, here are few findings: 1. Technical data about the chip were already summarized here...

audiosciencereview.com

I can't hear a difference between the USB output from this PC (into a MiniDSP flex) or USB output from my 14' Macbook pro into the same device.

 

[Sokel]

The main conversation is not about onboard audio,that's another matter.
It's more about the stuff like "clean" USB power or the ability to compensate for voltage drop on the specified USB ports (something that would benefit power-hungry dacs playing high sample rates) that they advertise,etc.
That would be interesting to measure.

Example:

 

[antcollinet]

The main conversation is not about onboard audio,that's another matter.
It's more about the stuff like "clean" USB power or the ability to compensate for voltage drop on the specified USB ports (something that would benefit power-hungry dacs playing high sample rates) that they advertise,etc.
That would be interesting to measure.

Example:

View attachment 235342

That looks like nonsense. How does a USB port reduce the voltage drop along a wire. That is a characteristic of the wire, not the port.

They can claim to out put a slightly higher voltage - but after that it will still fall off the same amount at the same cable length.

 

[Sokel]

That looks like nonsense. How does a USB port reduce the voltage drop along a wire. That is a characteristic of the wire, not the port.

They can claim to out put a slightly higher voltage - but after that it will still fall off the same amount at the same cable length.

Regulators?I don't know.
There are ways to compensate for voltage drop,actively monitoring them,etc.
IF they actually do it.
Need for measurements!

 

[TrevC]

My PC sounded awful via its audio outs, a cheap USB Behringer DAC fixed it. It now sounds as good as my other sources.

 

[antcollinet]

Regulators?I don't know.
There are ways to compensate for voltage drop,actively monitoring them,etc.
IF they actually do it.
Need for measurements!

The only way to do it would be to measure the voltage at the dac end of the cable and up the output voltage to compensate. As far as I know though, the USB standard has no facility for this.

 

[Sokel]

The only way to do it would be to measure the voltage at the dac end of the cable and up the output voltage to compensate. As far as I know though, the USB standard has no facility for this.

There are a lot of regulators doing that,Salas for example (which I use everywhere) but yes,the sense must be tied to the load to adjust.
The only way I can imagine is a constant monitoring-adjusting system,not difficult for gigabyte I suppose.

 

[antcollinet]

There are a lot of regulators doing that,Salas for example (which I use everywhere) but yes,the sense must be tied to the load to adjust.
The only way I can imagine is a constant monitoring-adjusting system,not difficult for gigabyte I suppose.

But you can only measure that with separate wires coming back from the load end to tell the source what the load end voltage is. These sense wires don't exist.

 

[Sokel]

But you can only measure that with separate wires coming back from the load end to tell the source what the load end voltage is. These sense wires don't exist.

Yes,they only measure at the port (if they do it) and adjust accordingly (delayed maybe?) edit: but only for the load (wire+device) presented there.
Only measurements can tell us the truth.

 

[Sokel]

Ok,I dig up a little and it seems it's configurable (through BIOS? ) and also (importantly) has the choice of disabling USB power lines for self-powered devises without having problems (or the need to trick it).
Of course that means it needs a little work measuring the actual Vdrop manually and adjusting accordingly.

 

[RandomEar]

Ok,I dig up a little and it seems it's configurable (through BIOS? ) and also (importantly) has the choice of disabling USB power lines for self-powered devises without having problems (or the need to trick it).
Of course that means it needs a little work measuring the actual Vdrop manually and adjusting accordingly.

That's still irrelevant. USB-fed DACs have to comply with the USB spec. Depending on the version, that means they are either limited to 2.5 W (2.0), 4.5 W (3.0), 7.5 W (3.2 Gen2 over USB C port) or use some form of USB PD, which essentially goes up to 100 W (240 W coming soon) but is rarely supported by motherboards. The USB spec also specifies maximum cable lengths, but they are rather conservative and you can usually get away with longer than the recommended cables. Either way: If the DAC complies with the USB spec, power delivery will be fine. That's it. End of story.

"Cleaner" power is equally bullshit, because the USB spec determines what is acceptable and the DAC - if compliant to spec - has to accept and handle everything within spec and run fine. Further, the chips in DACs don't run on 5 V and there are regulators for 3.3 V, 1.5 V or whatever is required. Those are responsible for generating "clean" power at 3.3/1.5/whatever Volt and - as long as the 5 V on the USB input is within spec - the regulators have to handle any residual noise or deviation.

I'd go as far as to say that anything on the motherboard side which alters the USB voltage or power delivery system risks running it out of spec and should therefore be considered detrimental to the system.

 

[Sokel]

That's still irrelevant. USB-fed DACs have to comply with the USB spec. Depending on the version, that means they are either limited to 2.5 W (2.0), 4.5 W (3.0), 7.5 W (3.2 Gen2 over USB C port) or use some form of USB PD, which essentially goes up to 100 W (240 W coming soon) but is rarely supported by motherboards. The USB spec also specifies maximum cable lengths, but they are rather conservative and you can usually get away with longer than the recommended cables. Either way: If the DAC complies with the USB spec, power delivery will be fine. That's it. End of story.

"Cleaner" power is equally bullshit, because the USB spec determines what is acceptable and the DAC - if compliant to spec - has to accept and handle everything within spec and run fine. Further, the chips in DACs don't run on 5 V and there are regulators for 3.3 V, 1.5 V or whatever is required. Those are responsible for generating "clean" power at 3.3/1.5/whatever Volt and - as long as the 5 V on the USB input is within spec - the regulators have to handle any residual noise or deviation.

I'd go as far as to say that anything on the motherboard side which alters the USB voltage or power delivery system risks running it out of spec and should therefore be considered detrimental to the system.

Agreed 100%.In an ideal world.
In an ideal world will also have zero threads about USB carrying all shorts of noises,all dacs would draw their specified current according the protocol they use,etc.
Also the spdif an optical jitter measurements would be nice and not all over the place as we have seen in a gazillion measurements coming from PC side.
Ideal performance obviously is not the case,there are too many threads reporting exactly that.
People rely on optical for no noise and jitter is going silently bananas!
People can hear CPU/GPU activity,fans,HDD's,etc.Not to mention ground loops,etc.

Wouldn't be nice if we had measurements to see if at last someone is attending all that?

 

[antcollinet]

Agreed 100%.In an ideal world.
In an ideal world will also have zero threads about USB carrying all shorts of noises,all dacs would draw their specified current according the protocol they use,etc.
Also the spdif an optical jitter measurements would be nice and not all over the place as we have seen in a gazillion measurements coming from PC side.
Ideal performance obviously is not the case,there are too many threads reporting exactly that.
People rely on optical for no noise and jitter is going silently bananas!
People can hear CPU/GPU activity,fans,HDD's,etc.Not to mention ground loops,etc.

Wouldn't be nice if we had measurements to see if at last someone is attending all that?

I would say close to all reports of audible noise on USB driven connections come from ground loops. This has nothing to do with the motherboard design whether that is power supply or anything else.

Ground loops are system problems, and depend on the flow of interconnects between devices and how each device is grounded (or not). And the noise is not created in the USB connection - but in the analogue connections further down the line. The USB connection only forms part of the loop for the ground circulating currents that cause the noise.

The motherboard design can't help here unless it offers galvanically isolated USB, and as far as I am aware, none do.

So no, audiophile motherboards will do nothing for USB connections.

 

[Sokel]

I would say close to all reports of audible noise on USB driven connections come from ground loops. This has nothing to do with the motherboard design whether that is power supply or anything else.

Ground loops are system problems, and depend on the flow of interconnects between devices and how each device is grounded (or not). And the noise is not created in the USB connection - but in the analogue connections further down the line. The USB connection only forms part of the loop for the ground circulating currents that cause the noise.

The motherboard design can't help here unless it offers galvanically isolated USB, and as far as I am aware, none do.

So no, audiophile motherboards will do nothing for USB connections.

Let's see a sloppy example:

PC1

PC2

Comparison

The above are two of my PC's.
Each measurement is running them throught a (USB connection) loopback to E-MU.
Running the same stuff,trying to measure them when idle.
Ok,I'm lucky not to have problems.
But are not the same and I'm sure that if I have measured a previous one I had there would be a party all over the place.

 

[antcollinet]

Let's see a sloppy example:

View attachment 235354

PC1

View attachment 235355

PC2

View attachment 235356

Comparison

The above are two of my PC's.
Each measurement is running them throught a (USB connection) loopback to E-MU.
Running the same stuff,trying to measure them when idle.
Ok,I'm lucky not to have problems.
But are not the same and I'm sure that if I have measured a previous one I had there would be a party all over the place.

None of that stuff is audible. You are showing variations in a noise floor far below the level that can be heard.

 

[Sokel]

None of that stuff is audible.

Nor is 115 SINAD (or far lower with music),but...
Wouldn't be nice everything to be in order from a technical view?

 

[RandomEar]

Let's see a sloppy example:

View attachment 235354

PC1

View attachment 235355

PC2

View attachment 235356

Comparison

The above are two of my PC's.
Each measurement is running them throught a (USB connection) loopback to E-MU.
Running the same stuff,trying to measure them when idle.
Ok,I'm lucky not to have problems.
But are not the same and I'm sure that if I have measured a previous one I had there would be a party all over the place.

I'm really not sure what is shown here or what setup is behind this. But USB is a digital connection by design. If this were a digital connection, there would be no difference in the results. So it looks to me like there is an analog step in there with different noise levels. If so, that would be totally different from a USB-connected DAC, which relies solely on digital audio transmission and doesn't care about USB noise (up until the point where the connection fails, that is).