Can you hear a smartphone DAC? Well can you PUNK?

[BillG]

Those specs are optimistic versus what I measured.

They're not specifications, but rather measurements, and there's a link to the equipment and methods used at the bottom of the page. Perhaps if you'd post yours for comparison...

 

[Blumlein 88]

They're not specifications, but rather measurements, and there's a link to the equipment and methods used at the bottom of the page. Perhaps if you'd post yours for comparison...

I found like they did, that loading the output had significant effect. And that it had very low output though sufficient for your typical very sensitive ear buds. So I loaded it with 1500 ohms. My results weren't so far from their loaded with headphones results. Except for FR. It was flat to 15 khz and dropped rapidly above that. Loaded the noise level was also higher than what they are showing.

 

[grizzog]

Looks like I'm late to the party on this one. I didn't know which was the smartphone, but I could hear a difference between A/B so I just stayed with those two. For me, I enjoyed that I could tell that there was a difference, not necessarily picking out which is "better." It may also be down to the playback gear. When I identified the difference, I was using the Sennheiser HD660s. I then tried my ATC SCM7 V3 and the difference was more apparent. For me, it's the transition of the cymbals that happen between 2-3 seconds in the file that I think I hear the difference. B seems to just "stop" whereas there is more decay on the sound in A.

 

[GaryH]

Looks like I'm late to the party on this one. I didn't know which was the smartphone, but I could hear a difference between A/B so I just stayed with those two. For me, I enjoyed that I could tell that there was a difference, not necessarily picking out which is "better." It may also be down to the playback gear. When I identified the difference, I was using the Sennheiser HD660s. I then tried my ATC SCM7 V3 and the difference was more apparent. For me, it's the transition of the cymbals that happen between 2-3 seconds in the file that I think I hear the difference. B seems to just "stop" whereas there is more decay on the sound in A.

Foobar blind ABX Comparator with at least 10 trials or it didn't happen

 

[GaryH]

The phone has an SNR of about 75 db. A tiny response droop at the low end (which probably isn't meaningful with 99% of music). It has distortion around -90 db which is inaudible. It has an interesting frequency response. It is almost fully flat until right about 15 khz. It then rolls off to -10 db by 20 khz. Likely uses one of those half band reconstruction filters. This also is likely inaudible to the mostly middle-age clientele of the audiophile world.

Would be great to have these recordings redone at 48 kHz, because Soomal's measurements of the Nexus 6P show its frequency response is flat up to 20 kHz at that sample rate, so that could determine if the differences some (may) hear are just down to the FR at 44.1 kHz sample rate. Or another smartphone with known to flat FR to 20 kHz if you no longer have the 6P.

 

[Blumlein 88]

Would be great to have these recordings redone at 48 kHz, because Soomal's measurements of the Nexus 6P show its frequency response is flat up to 20 kHz at that sample rate, so that could determine if the differences some (may) hear are just down to the FR at 44.1 kHz sample rate. Or another smartphone with known to flat FR to 20 kHz if you no longer have the 6P.

I don't think I still have the 6P anywhere. This was over 5 years ago. Lots of current phones require a dongle as the headphone jack is gone. If you need a dongle get the $9 one from Apple.

 

[GaryH]

I don't think I still have the 6P anywhere. This was over 5 years ago. Lots of current phones require a dongle as the headphone jack is gone. If you need a dongle get the $9 one from Apple.

I meant it more as a test to debunk the mantra that a so-called 'low quality' DAC as found in a phone is audibly worse than a standalone DAC, so any device with THD more than around -90 dB and flat frequency response up to 20 kHz would do.

 

[JJB70]

To me there are two questions:

-can you discern a difference; and

- is such a difference discernible in normal listening as opposed to intensive critical listening to identify differences?

For me the answer to the first is generally no. For the second the effort it takes to discern those differences I can identify tells me that they are utterly irrelevant when listening to music.