SMSL D1 - ROHM DAC for everyone

[aewyli]

You can power the D1 externally, but you don't have to.
Any computer can power the D1 without any problems. The USB outputs on the computer are power-limited, but the D1 doesn't draw that much power.

Okay, the power draw of this device:

OFF = <50mW
ON USB data only = ~1.71W
ON dedicated power + USB data = ~1.64W

Initially, there seemed to be gain differences between running through USB data only VS powering seperately. The latter one is much proper sounding as how the D1 is supposed to sound like. By powering it through the data in, it has this unpleasant loudness effect for whatever reason.

However, this has been greatly resolved after installing the XMOS USB driver provided for the D1. The sound is now much proper and transparent by just running it through the USB data in. This device is truly fantastic, recommended.


- Tested with a neutral headphone on Windows 10 platform, running PCM playback on foobar2000 in exclusive mode

 

[RandomEar]

Initially, there seemed to be gain differences between running through USB data only VS powering seperately. The latter one is much proper sounding as how the D1 is supposed to sound like. By powering it through the data in, it has this unpleasant loudness effect for whatever reason.

However, this has been greatly resolved after installing the XMOS USB driver provided for the D1. The sound is now much proper and transparent by just running it through the USB data in. This device is truly fantastic, recommended.

I'm sorry but this is 100% placebo / sighted bias. The device does not know where the power is coming from. It does not matter how you power it as long as there is enough amps at 5 V (disregarding obvious problems with ground loops). If there isn't enough amps, it likely won't even boot successfully.

Please, please, please stop spreading such ridiculous misinformation.

 

[Roland68]

Okay, the power draw of this device:

OFF = <50mW
ON USB data only = ~1.71W
ON dedicated power + USB data = ~1.64W

Initially, there seemed to be gain differences between running through USB data only VS powering seperately. The latter one is much proper sounding as how the D1 is supposed to sound like. By powering it through the data in, it has this unpleasant loudness effect for whatever reason.

However, this has been greatly resolved after installing the XMOS USB driver provided for the D1. The sound is now much proper and transparent by just running it through the USB data in. This device is truly fantastic, recommended.


- Tested with a neutral headphone on Windows 10 platform, running PCM playback on foobar2000 in exclusive mode

I'm sorry but this is 100% placebo / sighted bias. The device does not know where the power is coming from. It does not matter how you power it as long as there is enough amps at 5 V (disregarding obvious problems with ground loops). If there isn't enough amps, it likely won't even boot successfully.

Please, please, please stop spreading such ridiculous misinformation.

However, this suggests a driver problem; I couldn't reproduce such differences on a Mac.

 

[Nutul]

However, this suggests a driver problem; I couldn't reproduce such differences on a Mac.

powering it through MAC's USB? Maybe they deliver power as it should be...?

 

[truth_soldier]

I'm sorry but this is 100% placebo / sighted bias. The device does not know where the power is coming from. It does not matter how you power it as long as there is enough amps at 5 V (disregarding obvious problems with ground loops). If there isn't enough amps, it likely won't even boot successfully.

Please, please, please stop spreading such ridiculous misinformation.

This is not necessarily always true. Besides the possible noise level difference between the two power supplies which can change the sound a little... if the two have different voltages the sound can be lower volume. I know because I saw this with a step-down regulator I was using was putting out 4 volts instead of 5. The D1 still worked fine but the sound was slightly lower in volume a rolled off a little in the high end compared running off just the PC USB. Mind you the difference was small - so small I didn't notice it for a while until I did some A/B listing, realized there was a difference, then measured and found the voltage of my external power supply was only 4v.

Surprising actually that the D1 works on only 4V and doesn't sound horrible, though not ideal obviously.
But people who prefer softer/warmer (rolled off) sound might actually prefer the sound.

 

[RandomEar]

This is not necessarily always true. Besides the possible noise level difference between the two power supplies which can change the sound a little... if the two have different voltages the sound can be lower volume. I know because I saw this with a step-down regulator I was using was putting out 4 volts instead of 5. The D1 still worked fine but the sound was slightly lower in volume a rolled off a little in the high end compared running off just the PC USB. Mind you the difference was small - so small I didn't notice it for a while until I did some A/B listing, realized there was a difference, then measured and found the voltage of my external power supply was only 4v.

Surprising actually that the D1 works on only 4V and doesn't sound horrible, though not ideal obviously.
But people who prefer softer/warmer (rolled off) sound might actually prefer the sound.

As I said: "enough amps at 5 V". You are right on the noise, it's the same as ground loops. You will hear it immediately even if no sound is playing.

 

[aewyli]

However, this suggests a driver problem; I couldn't reproduce such differences on a Mac.

I could have worded my statement differently. There is something off with the USB IN if one plan to use only this, like many would, out of the box. The closest example I can make is that it sounded like "Loudness Equalization" but it is not this at all, not necessarily louder in volume output now. One would find this is not an issue or even realizing it but having listened to proper DACs, this is unpleasant for me.

By installing the XMOS USB driver v6.0 specifically for the D1, this is now non-existent as if the D1 is back to how it should have been. Although, when I first powered it externally prior to installing the driver just to measure power usage, I did not notice this "loudness".

I realized this through PCM playback on foobar2000 in exclusive mode in fixed(max) volume with my headphone amp having this 'click' volume mechanism, untouch. Simply end playback, turn off D1 and disconnect USB, install the driver and reconnect back then re-select exclusive mode then I'd able to tell the sound powering through USB IN has the same volume output but no more peaking loudness.

I have tested this again just to be sure by uninstalling the XMOS driver and did a full system restart. I can now confidently say this, the sound of powering via no-driver USB IN, is untuned. Simply reconnect back with external power, without driver installed, has no peaking loudness crap it is just proper. The D1 also does not support 16-bit out of the box.

#Windows10 #foobar2000 #exclusivemode #headphone

 

[RandomEar]

This is extremely unlikely to be a problem with the D1 or the XMOS driver itself. Neither should alter the sound in any way. It's more likely that there is some form of Windows APO active on your system, which gets circumvented by using a stand alone driver. You can check it in the sound settings. If you have EAPO installed, you should also make sure that it is enabled for the XMOS driver output.

EDIT: I just saw that you are using exclusive mode in foobar. Technically, that should avoid all Windows APOs. Not sure what the problem is in that case. Nonetheless, neither the DAC itself nor the XMOS driver do any equalization or something similar. At best, I could imagine some level difference between the paths causing the impression of sound differences.

 

[Farva0066]

unit variation on the production line is a well known fact, i remember that intel cpu were tested out of the production line and certified for different speed based on their individual performance coming from random per unit differences, same thing for grado drivers headphones or ray tubes on the gremlin amp that are not measured but sorted by a human ear

A silicon wafer is going to have variability in the quality of individual chips - even tho they all use the exact same design. In the example of Intel, if a chip comes out "perfect" (6.0ghz on 2 Pcores, all other cores hit target at target voltage) it is an i9 for $589. If it can't hit the i9 standard, they will fuse off some e-cores, and give it a lower clock speed (5.4ghz on P cores) and sell it as an i7 for around $400. Alot of this is pure product segmentation: maybe one of the e-cores *was* defective, so the entire chip got got a lower frequency; or, the P-cores couldn't hit the right frequency so perfectly fine e-cores got fused off.

This may not seem relevant, but I'm looking at noise floor as the equlivant to frequency - not all chips will hit the top-of-line spec, but they are otherwise fine (I doubt there are many examples of an entire section of a chip being defective on older nodes - and in this application it wouldn't be able to operate at all in all likelihood).

TL;DR same chip, different bin. "different tuning"... sure, but as a function of manufacturing - not a deliberate choice to sell the "different tuned" chip for a fraction.

 

[VientoB]

So now we're comparing very high frequency CPU chips with DAC chips?