Best Smartphone for audiophile | Page 455 | Headphone Reviews and Discussion - Head-Fi.org

Jimster480 · Feb 18, 2017 at 1:55 AM

kyrts said:
Kinda reminds me of the sound test graphs in devicespecifications.com reviews, all look alike and don't give a clue how good or bad or average real life performance of audio circuit is going to be like.

They sure look cool on that devicespecifications.com webpage though.

Just my 2c.

I just went to that site and I don't see any such graphs.

The graphs this guy makes are actually very informative.

GouryellaIV · Feb 18, 2017 at 2:16 AM

Ive updated the audiophile smartphone list, only found 3 Leagoo smartphones with dedicated audio chips so far for 2017.

LG G6 is to be released with quad dacs but not confirmed yet, theres more audiophile smartphones to come

@robervaul

SergeSE · Feb 18, 2017 at 5:10 AM

jimster480 said:
Wow that is awesome.
I would love to see the LG V10/V20/G5/G6 and the HTC 10/M9/A9/M7 and Apple iTouch 6G/4G/7.

it would be awesome if you could throw some other cheap DAC/AMP's into the mix like the FiiO E10k/K1/E18K

I realize this is alot of devices to ask to test with but maybe if you can find people who own them, it would be really interesting to see how many devices stack up.

At the moment the testing procedure and the test signals are not final. When I finish research I will try to crowdfund buying devices for testing. I need more freedom in operating with a device and at lest a few days to complete testing and be sure that it was made correctly. It is not possible with friends' devices.

AnakChan · Feb 18, 2017 at 5:33 AM

shai347 said:
Yeah I assumed that.. Still hopeful..
That will be a real treat..

I want to go balanced with my Shure SE846 so it will be cool to have that option on a slim good performance device.

The Onkyo Granbeat will probably have better sound quality but it's such a brick with mid range performance.. Plus, who knows how the Smartphone department will be from Onkyo for the first time making a Smartphone..

At the Fujiya Winter Potaken Festival today, I managed to try the Granbeat and placed some basic thoughts in that thread.

gugman · Feb 18, 2017 at 5:58 AM

on geekbuying.com you can pre-order Le Max 2 X829 with a coupon code LETVX829 for 199.99$

fairx · Feb 18, 2017 at 2:16 PM

gouryellaiv said:
Ive updated the audiophile smartphone list, only found 3 Leagoo smartphones with dedicated audio chips so far for 2017.

LG G6 is to be released with quad dacs but not confirmed yet, theres more audiophile smartphones to come

@robervaul

Thanks for the effort mate. Is there any link I can get the list in high quality pic or html / text

GouryellaIV · Feb 18, 2017 at 2:28 PM

fairx said:
Thanks for the effort mate. Is there any link I can get the list in high quality pic or html / text

Click the image and click "Original", should be in high res

SergeSE · Feb 18, 2017 at 5:09 PM

jagwap said:
Interesting test.

What is interesting is that they are so similar.

Have you done a quick check with and without the 33 ohm resistors? If there is a big reduction of the difference every time, the load may be showing up something in the test jig. I would expect the loading to change the result due to crosstalk and loading of the amplifier, but it may be the test configuration may be limiting the best number available.

More likely is that as the 1kHz signal is often good, it is the frequency extremes that are showing the difference. A small difference in bandwidth would cause a phase shift. This would cause white noise and program noise to be much worse.

(By the way, the Pink Floyd Discovery Editions are not great. Compressed compared to the originals. Animals and Final cut are particularly bad. http://dr.loudness-war.info/album/list?artist=pink+floyd&album= )

I did some tests with different loads including real headphones some time ago. All loads worsen the readings, what seems to me natural. I don't know how the map would look like in case of loadless testing. Probably the Df values will keep their mutual positions but will shift a few dB towards -Inf. In any case such loadless measurements are pretty useless for end users.

And yes, complicated transfer function of a device distorts waveform of white noise (and music material) more significantly than of pure sine.

BTW, Df values of different devices do not seem to me very similar, each device seems to have its own "degradation signature" (a set of Df values with different test signals).

In the next version of Df testing a 2 hours set of various tracks will be used instead of the PF album.

Maglebrius · Feb 18, 2017 at 8:30 PM

gouryellaiv said:
Ive updated the audiophile smartphone list, only found 3 Leagoo smartphones with dedicated audio chips so far for 2017.

LG G6 is to be released with quad dacs but not confirmed yet, theres more audiophile smartphones to come

@robervaul

Are you sure about x3 and x3 lite having the same internals?

Also you may add bqAquarius having a dedicated dac

andrewski · Feb 18, 2017 at 9:31 PM

jagwap · Feb 18, 2017 at 10:17 PM

sergese said:
I did some tests with different loads including real headphones some time ago. All loads worsen the readings, what seems to me natural. I don't know how the map would look like in case of loadless testing. Probably the Df values will keep their mutual positions but will shift a few dB towards -Inf. In any case such loadless measurements are pretty useless for end users.

And yes, complicated transfer function of a device distorts waveform of white noise (and music material) more significantly than of pure sine.

BTW, Df values of different devices do not seem to me very similar, each device seems to have its own "degradation signature" (a set of Df values with different test signals).

In the next version of Df testing a 2 hours set of various tracks will be used instead of the PF album.

The distortion I'm talking about it the phase shift that different bandwidths set by the unit under test. All of them are likely to have a low pass filter after the DAC, and perhaps the amp, and some may have a high pass filter in the low frequencies. This will cause different phase shifts for each unit, which will be measured as a error. It is an error, but is it one you want to include? It will swamp the others distortions I suspect.

I wasn't criticising the Pink Floyd as a test. Just saying for a re-issue the quality is disappointing compared to some

eio · Feb 18, 2017 at 10:49 PM

sergese said:
no, Df is computed using input and output waveforms in time domain.

yes, but what does the diff signal look like in spectrogram? how much difference is in the audible range? are there some clock shifts over time?

SergeSE · Feb 19, 2017 at 5:10 AM

eio said:
yes, but what does the diff signal look like in spectrogram? how much difference is in the audible range? are there some clock shifts over time?

All clock/phase shifts are eliminated during Df computation. Frequency analysis of diff signal (first 4:17 of The Dark Side of The Moon, device: Highscreen Boost 3):

Diffrogram of this excerpt (red - more distortion of waveform):

Spectrogram of diff signal:

MusicFi · Feb 19, 2017 at 5:41 AM

could someone help me with this?

http://www.head-fi.org/t/837216/which-smart-phones-case-is-this

Mustang80 · Feb 19, 2017 at 6:21 AM

I see a lot of comments about the ZTE Axon 7 mini, but what about just the Axon Mini??

Accoring to thier site, it has an AKM 4961 - Are they any good?

Latest Thread Images

Jimster480

Headphoneus Supremus

GouryellaIV

100+ Head-Fier

SergeSE

Reviewer at SoundExpert.org

AnakChan

Don't let him borrow your M-100 (click)

gugman

1000+ Head-Fier

fairx

500+ Head-Fier

GouryellaIV

100+ Head-Fier

SergeSE

Reviewer at SoundExpert.org

Maglebrius

New Head-Fier

andrewski

500+ Head-Fier

jagwap

1000+ Head-Fier

eio

100+ Head-Fier

SergeSE

Reviewer at SoundExpert.org

MusicFi

Head-Fier

Mustang80

100+ Head-Fier

Users who are viewing this thread