Introduction
I created this thread to post and discuss measurements and data related to audio fidelity of the headphone output. Note that audio fidelity does not automatically correlate with audio quality. Audio quality is a very subjective measure. Some people like their audio to be very bass heavy while others prefer accentuated mids and/or highs. Yet others (like myself) enjoy a very "neutral" sound signature. Audio fidelity -- at least the way I define it -- on the other hand can be measured: Audio data is fed to the DAC/amp and measured at the output. For perfect fidelity, the recorded output is identical to the input. Of course, this ideal cannot be achieved. Deviations from the ideal can be measured, documented, and discussed. Particularly, not all deviations from the ideal are actually audible. I predict some heated discussions on audibility in this thread going forward...
Test Equipment
The device under test (DUT) is a Pixel XL that sports a Qualcomm WCD9335 codec/amp discrete audio chip. The headphone output is split with a Y-wire. One signal path goes into the measurement rig, a Rohde & Schwarz UPV Audio Analyzer. Depending on the measurement, the other end either stays open (not connected or N/C) or connected to my Sennheiser IE800 IEMs. Those IEMs have a virtually frequency-independent impedance of 16 Ohm, a sensitivity of 125 dB/Vrms at 1 kHz, and a ruler-flat phase response. These IEMs are not only one of the best "sounding" headphones I know, but actually very easy to drive as the specs that I've listed suggest.
Test Signals
For the tests conducted so far, I have used two signals that I created with Matlab:
stereo wave file with full-scale (0 dBFS) 1 kHz sine wave in the left channel and silence (zeros) in the right channel (16 bit, 48 kHz sampling rate)
stereo wave file with white noise mastered to take advantage of the full dynamic range (16 bit, 48 kHz sampling rate)
These signals were played with the stock Android music player, with all audio effects disabled.
In some cases I swapped the Pixel XL with an iPhone 5s for comparison purposes. All other hardware stayed the same.
Measurement Results
This section summarizes the measurements detailed in the second post.
Output Impedance
The Pixel's output impedance was measured to be 4.8 Ohm. To compare, the output impedance of the iPhone 5s' amp is 2.0 Ohm.
Comments: The Pixel's output impedance violates the often-quoted 1/8 rule in some cases. The rule says that most amps work best with headphones that have an impedance that is at least eight times its own output impedance. According to this rule, headphones should ideally have an impedance of more than 40 Ohm. Nevertheless, the amp seems to work with my 16 Ohm IEMs without any issues (other than the level drop, see below). The reason, I think, is that my IEMs make use of a single dynamic driver rather than multiple balanced armature ones like many other IEMs. Multiple balanced armature drivers are much harder to drive and may cause minor problems (slight modification of the frequency response) with the amp in the Pixel
Frequency Response (see plots in the second post)
The frequency response is ruler flat, whose shape does not change at all once my IEMs are plugged in as a load. The gain dropped by 2.3 dB after plugging in the headphones. The reason for this drop is the interaction of the headphone's impedances with the output impedance of the amp. Higher impedance headphones will see a smaller drop in volume when plugged into the Pixel.
The channel imbalance is a negligible 0.1 dB at 1 kHz.
Comments: As discussed above, the only potential issue I can see here is that the combination of this amp with low-impedance multiple armature-based IEMs may lead to slight modifications in the overall frequency response, maybe on the order of +/- 2 dB max.
Maximum output level
When the volume on the Pixel is turned up all the way, the maximum attainable output voltage with a full-scale 1 kHz sine wave is 0.427 Vrms, without any load attached. I have found that the output volume was capped digitally to -7 dBFS in the mixer by Google. The maximum theoretically attainable output voltage of virtually any DAC/amp combo devices is 1 Vrms. Once a custom recovery has been made available for the Pixel, it should be trivial to boost its maximum output voltage to 1 Vrms.
Comments: No issues here.
Volume level steps with volume rocker
As discussed above, at max volume the Pixel delivers 0.427 Vrms, which is volume step 15.
Lowering the output volume from the maximum setting, 15, with the volume rocker results in the following steps (rounded):
Volume setting: 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0
Change in gain: 0dB, -3dB, -3dB,-3dB,-3dB,-3dB,-3dB,-4dB,-4dB,-3dB,-4dB,-4dB,-4dB,-7dB,-7dB,-54dB
Noise Floor
The noise floor was measured to be 115 dB, which gives a theoretical dynamic range of (currently) 108 dB.
Comments: No issues here.
Total Harmonic Distortion + Noise (THD+N, see plots in the second post)
Measurements of THD+N were done on both iPhone 5s and Pixel at their respective maximum volume as well at a volume setting that corresponds to an output voltage of around 20 mV, with which the IE800 would produce around 90 dB SPL:
Pixel at 20 mV: THD+N is -71.4 dB
Pixel @ max volume: THD+N is -81.3 dB
iPhone 5s at 20 mV: THD+N is -78.7 dB
iPhone 5s @ max volume: THD+N is -84.1 dB
Comments: By comparing the plots in the second post, it can be seen that the Pixel has higher non-linear distortions than the iPhone 5s, especially at low output voltages. The reason for this is that Google and I believe any other Android manufacturer opts to maximize analog gain while controlling overall level with digital gain only. This is not ideal. It would be better to hit the codec with the highest digital gain possible (but no more than 0 dB), keep the analog gain low initially and control overall loudness with the analog gain only. This way the full dynamic range of the DAC can be utilized. This is the paradigm that virtually any home/car stereo systems utilizes. Based on the plots below, I'm guessing that Apple is doing it, too.
Intermodulation Distortion
Preliminary inspection of SMPTE IMD measurements suggest no problems either unloaded or loaded with my IEMs. Measurements to follow sometime next week.
Stereo Crosstalk
Stereo crosstalk very much depends on the interaction of amp, headphone, and analog circuit design of the phone. Here, at least on the surface, the Pixel does not seem to be showing strong numbers. With my headphones attached and throughout the volume range offered by the Pixel, I measured the stereo crosstalk to be 45 dB. I compared this number with the iPhone 5s. Interestingly, stereo crosstalk on the iPhone shows a strong dependence on the playback level. Again with the IEMs attached the crosstalk ranged from 43 dB at the lower volume settings up to 60 dB at the higher end. I repeated the measurements with full-size cans, the Sennheiser HD 540 (300 Ohm). With those headphones attached, the stereo crosstalk of both the iPhone 5s and the Pixel are north of 80 dB.
Comments: To put things into perspective, though, at the playback levels that in the long run do not cause permanent hearing loss (less than 90 dB SPL at the ear drum) with the IEMs, both iPhone 5s and Pixel have very similar stereo crosstalk performance, i.e. around 45 dB. Carefully controlled double-blind tests should be considered to determine at what point crosstalk actually becomes audible.
Comments on the Pixel mixer
There is one thing that strikes me as odd after having examined the Pixel mixer (/system/etc/mixer_paths.xml): The amplifier is set up as "CLS_H_LOHIFI" and not as "CLS_H_HIFI", which is the default for the WCD9335 in general and the HTC10 in particular. While I don't know what exactly the differences between those two settings are, I will play around with them as soon as a custom recovery becomes available. I'll also look into enabling hardware IIR filtering via the Qualcomm codec and update my biQuads app.
Supporting plots done with stock mixer can be found here.
Pixel "frequency response" at 20 mV output voltage. Note that the 2.3 dB level drop to to the presence of the IE800 has been compensated for (with this output voltage my IE800 delivers around 90 dB SPL):
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Pixel "frequency response" at maximum output voltage (0.427 Vrms). Note that the 2.3 dB level drop to to the presence of the IE800 has been compensated for (with this output voltage my IE800 delivers around 115 dB SPL):
Pixel THD+N at 20 mV output voltage (only the left graph is relevant):
Pixel THD+N at maximum output voltage (only the left graph is relevant):
iPhone 5s THD+N at 20 mV output voltage (only the left graph is relevant):
iPhone 5s THD+N at maximum output voltage (only the left graph is relevant):
reserved (for future measurements based on a modified mixer)
Damn. Thanks for doing a deep dive on this.
Greatly appreciate you taking the time to do the measurements.
I'm assuming Normal Pixel would have mirrors measurements as my listening observations are in line with what you have presented.
Although on my Pixel it seems like the Single loud speaker also has the same 'Boomy' bass boost applied.
I have Campfire Jupiters and Shure SE535 with Brown Knowles dampeners and the Shures just pair better with the stock Pixel.
Looking forward to future Audio tweaks on this device.
With both headphones I have been using and brothers Shure SE215: I have noticed that the lowest volume level is still what I would consider listening level. (Where as on Note 7 sound was not audible until about 25% volume from off)
Thanks for providing such quality measurements and interpretation.
Since the physical structure of the plain Pixel is different, it is possible the crosstalk is not the same. Would be interesting to compare if anyone has one to test.
mariano3113 said:
I'm assuming Normal Pixel would have mirrors measurements as my listening observations are in line with what you have presented.
Click to expand...
Click to collapse
bjrmd said:
Since the physical structure of the plain Pixel is different, it is possible the crosstalk is not the same. Would be interesting to compare if anyone has one to test.
Click to expand...
Click to collapse
I can imagine that the Pixel would measure exactly the same as the Pixel XL. It would make a lot of sense for Google/HTC to use the same board layout as a cost savings measure, even though the two devices have different physical dimensions.
chdloc said:
I can imagine that the Pixel would measure exactly the same as the Pixel XL. It would make a lot of sense for Google/HTC to use the same board layout as a cost savings measure, even though the two devices have different physical dimensions.
Click to expand...
Click to collapse
What I meant (and it was just a possibility) is illustrated here
If you look at the sony z5 regular vs compact, the crosstalk is 8 dB different. As you have mentioned, it's the grounding circuitry that determines crosstalk (not the dac) and perhaps a different physical layout makes a difference.
bjrmd said:
What I meant (and it was just a possibility) is illustrated here
If you look at the sony z5 regular vs compact, the crosstalk is 8 dB different. As you have mentioned, it's the grounding circuitry that determines crosstalk (not the dac) and perhaps a different physical layout makes a difference.
Click to expand...
Click to collapse
Oh, I apologize for being so vague in my response. As reported, Google/HTC spent only 9 months in designing the Pixels. On such a short development timeframe, it would make a lot of sense to clone as much of the hardware design as possible. Provided the circuit board carrying the codec physically fits into the regular Pixel's body (a big "if" as the regular Pixel is 6.2 mm narrower), I'm willing to bet that the two phones are identical from a headphone audio perspective. Otherwise, of course, all bets are off.
chdloc said:
Introduction
Output Impedance
The Pixel's output impedance was measured to be 4.8 Ohm. To compare, the output impedance of the iPhone 5s' amp is 2.0 Ohm.
Comments: The Pixel's output impedance violates the often-quoted 1/8 rule in some cases. The rule says that most amps work best with headphones that have an impedance that is at least eight times its own output impedance. According to this rule, headphones should ideally have an impedance of more than 40 Ohm. Nevertheless, the amp seems to work with my 16 Ohm IEMs without any issues (other than the level drop, see below). The reason, I think, is that my IEMs make use of a single dynamic driver rather than multiple balanced armature ones like many other IEMs. Multiple balanced armature drivers are much harder to drive and may cause minor problems (slight modification of the frequency response) with the amp in the Pixel
Frequency Response (see plots in the second post)
The frequency response is ruler flat, whose shape does not change at all once my IEMs are plugged in as a load. The gain dropped by 2.3 dB after plugging in the headphones. The reason for this drop is the interaction of the headphone's impedances with the output impedance of the amp. Higher impedance headphones will see a smaller drop in volume when plugged into the Pixel.
Click to expand...
Click to collapse
Just to be clear, the output impedance is a problem. Here is the frequency response driving Unique Melody Merlins:
More than a minor problem there.
Here it is with a 32Ω load.
For comparison, here is a Sony ZX1 with the same two headphones used as load:
With the Merlins, the ZX1 between 30Hz-15KHz is +.5/-1.25dB with the heavy roll starting at around 7KHz, not ideal, but not much music content up there.
The Pixel XL is around +1dB/-1.5dB, but with a ton of rolling in the middle. And that's what you will see with balanced armature drivers.
Also note that the hump at around 80Hz with the PS1000 load is around double what it is on the ZX1, but the PS1000 impedance goes a little nuts right there: http://www.innerfidelity.com/images/GradoPS1000.pdf
And a comparison of the Astell&Kern AK300, Sony ZX1, and Nexus 6 with the Merlins as a load. N6 was actually the best performer of the lot!
Reginalb124 said:
Just to be clear, the output impedance is a problem. Here is the frequency response driving Unique Melody Merlins:
Click to expand...
Click to collapse
Thanks for sharing your measurements!
For the sake of others looking at this, however, I think it is important to point out that the output impedance may be a problem and is by no means guaranteed.
Not all folks using this phone or looking to buying one use such high-demanding IEMs. Remember that my IEMs, the Sennheiser IE800, do not cause any measurable change in the overall frequency response, other than a frequency-independent level drop of 2.3 dB. I'm very satisfied how my IE800 sound with the Pixel.
As your own plots show, your IEMs cause "issues" in all configurations, even with the Astrell&Kern dedicated music player! Note that you are arguing over frequency response swing differences of less than 1 dB. Keep in mind that the amplifier-headphone system is largely linear so you can take the additional small frequency (and likely phase) swings due to the headphone-amp interaction and fold them into the already existing large swings of the headphones themselves (linear system theory).
My point is that those tiny frequency swings, +/- 1 dB, by themselves likely do not cause sufficient changes in the headphone's overall sound signature to be problematic. Unless the artificial bass boost causes the bass to become muddy.
The final word, however, will have to come after carefully conducted double-blind listening tests have been performed. I'm not going to making a lot of friends around here by saying that comparing the audio quality of two phones/devices without carefully matching the levels within a fraction of a dB and performing the tests blindly only result in subjective opinions without much merit.
My Shure SE535's sure love this phone, sounds a bit better to me than my Nexus 6.
My SE846 with an impedance of 9 ohms sound great, although I do use the built in EQ a bit to create a v-shaped sound signature which I prefer. Maybe I'll do a comparison with my Grace m9XX DAC/amp to see if I notice any major differences with a flat EQ on both.
The "problem" with low impedance headphones is not necessarily the low impedance as @chdloc has indicated. It's the variability with frequency (because each driver and crossover has different impedance effects) that can mess with the frequency. The two graphs below are from a Note 4 (1.4 ohm) with a Shure se535 (variable but higher impedance) and the se846 (variable low impedance).
Here is the impedance graph of the se846-note the variability:
As can be seen, the frequency shifts are higher with the se846--but if the headphone impedance was the same throughout, it would look flat.
The se846- more exaggerated:
The se535 (much higher impedance but still variable)-not too bad:
I have measured the same curve with my Dragonfly red (.02 ohm) and it is totally flat
I would guess the exaggeration would be more with the Pixel (5 ohm). This may sound "better" to some, but not to others-as noted above.
My plan (after I get my preorder) is to remeasure the frequency response and just design some biquads to "fix it" the way I like it. But, the average user who is not adjusting this may notice a different sound signature.
It is not a deal breaker, but I was a bit disappointed in Google/HTC for not doing a better job. My Note 2 was close to 1 ohm, S6 1.4 ohm and Iphone noted above.
bjrmd said:
The "problem" with low impedance headphones is not necessarily the low impedance as @chdloc has indicated. It's the variability with frequency (because each driver and crossover has different impedance effects) that can mess with the frequency.
Click to expand...
Click to collapse
It is actually a combination of a highly variable and low impedance. Higher impedance headphones with varying impedance will have less of an effect as the amplifier's damping factor -- which is a function of its output impedance -- will be able to "deal" with it.
But still, the number of users that have such high-demanding IEMs, like your SE846 or the AKG K3003, is likely very low. That's why I wanted to counter the general statement of doom and gloom expressed earlier in this thread. Having said that, you guys with demanding IEMs may want/need to resort to either utilizing correction filters, buying a different phone with a lower impedance amp, or investing in an external DAC and/or amp.
Thanks for the effort you put into this. I've got a couple sets of ciems from Noble Audio that I'm itching to try once my XL comes in.
GSMarena review up-no real surprises and basically agrees with above. Crosstalk -62 dB, not as good as HTC 10, Iphone but better than the LG V20 (special ESS dac).
All in all, with respect to audio a solid performer but certainly not spectacular.
Life is full of compromises.
This is from my Google Pixel (non XL)
Removed LO from LOHIFI (To my ears it seems to have reduced the 'Boomy’ bass)
About to test Elemental kernel.
Edit:
another XML edit: "RX HPH Mode" value="CLS_H_LP" compared to HTC 10 & Note 7 (Snapdragon) "RX HPH Mode" value="CLD_H_HIFI"
Before:
After:
mariano3113 said:
This is from my Google Pixel (non XL)
Removed LO from LOHIFI (To my ears it seems to have reduced the 'Boomyâ?? bass)
Click to expand...
Click to collapse
I've made that change, too, but I have a hard time hearing a difference (my headphones don't sound boomy to begin with).
Did you reboot after making the change? Measurements will have to wait until at least Monday.
Did you actually edit your mixer_paths.xml in /system or did you go the system-less route by binding a modified mixer in /su/etc?
BTW, biquad filtering, stereo recording, and a very first shot at dual loudspeaker playback seem to be working (the earpiece is a *lot* quieter than the main speaker, however), all done via mixer_paths.xml changes. I'm still having trouble, though, getting the biquads modification to work reliably when going the system-less route...
This is a great thread. I love getting the best out of my devices and sound has always been important to me. I'm no audiophile but I'll be keeping up with this. Thanks!
Related
Anyone do any testing on the output impedance of the output jack yet? The S2 seemed to have quite a high amount which is why it had a hard time driving a lot of headphones (had to crank the volume very high just to get decent output).
Also another factor in low frequency response because if the output circuitry uses an output capacitor for DC blocking that also creates a highpass filter and there would be a trade off of low impedance vs. low frequency response.
I've been reading mixed reviews on the SQ of these units. I've read about the mod and it sounded like the sound just got louder. But is the sound fine and just flat sounding or is it something else? I was thinking a line driver would help or maybe it would just amplify the noise. I have some decent audio equipment and I don't want to feel I'm not getting the full potential out of them. I do use some WAV and FLAC files. I guess I can't tell if people are commenting on the on board audio or from their amps hooked up to the RCAs.
I have a 2016 Jeep Wrangler Unlimited. I had the factory 430N head unit, with the base 8 speaker system. I replaced 4 of the 8 speakers with Kickers and add a Kicker sub and amp (amp is just for the sub), and it sounded better but still not great.
THis weekend I added a Joying Jeep specific unit, and I feel like the sound of the overall system is vastly improved. I think it sounds pretty good, I primarily use Google Play Music and Pandora.
CadillacMike said:
I have a 2016 Jeep Wrangler Unlimited. I had the factory 430N head unit, with the base 8 speaker system. I replaced 4 of the 8 speakers with Kickers and add a Kicker sub and amp (amp is just for the sub), and it sounded better but still not great.
THis weekend I added a Joying Jeep specific unit, and I feel like the sound of the overall system is vastly improved. I think it sounds pretty good, I primarily use Google Play Music and Pandora.
Click to expand...
Click to collapse
Did you add a line driver at all to bump up the voltage on the pre-outs?
splxtreme said:
Did you add a line driver at all to bump up the voltage on the pre-outs?
Click to expand...
Click to collapse
I did not add anything. The sub and amp are built for Jeep, so it might be a little weird:
http://www.kicker.com/SWRA411
have any of you measure the pre-out voltage on the joying? On my tonghai create 2nd gen unit, pre-out voltage is only 1V. What this means is you will have a high noise floor going into your amplifier. Then your amp will amplify any noise your pre-outs have picked up. So if you use a line driver to raise the pre-out voltage prior to going into your amp, the higher voltage actually LOWERS the noise floor, hence eliminating noise that your RCAs may pick up.
My advice is you MUST have a line driver if you plan to use these units w/ a external amp. The difference will be night and day, particularly in the higher frequencies, where amplified noise is more pronounced.
They are dirt cheap too. I bought one on amazon for ~$30. It is actually a 3-way cross-over/line driver combo. It feels cheap but does it's job as a line driver very well.
edit: I say the above from personal experience, as I literally spent months trying to figure out why my tweeters always sounded so terrible. Line driver did the trick. Now I have clean highs and more clear, vibrant mids. I don't seem to notice much difference in my bass, so apparently noise interference isn't audible in the lower frequencies.
explain "line driver" to a stupid dumb dumb person, not me of course, just so other people understand
CadillacMike said:
explain "line driver" to a stupid dumb dumb person, not me of course, just so other people understand
Click to expand...
Click to collapse
line driver is just another word for a voltage step-up device. It's simply a way to increase the voltage of the RCA pre-outs coming out of your HU. Most amplifiers can read all the way down to about .5V pre-outs no problem. So a low pre-out voltage isn't much of a problem in theory.
But, an amplifier is also a voltage step-up device. It's going to take the pre-out voltage and multiply it by a pre-determined amount. You can adjust that amount by modifying gain. Your speakers are rated for a certain RMS power (80W for instance). So, if your speakers are 4 ohms rated for 80W RMS, the formula for the voltage it needs from your amp is sqrt(80 * 4) = 18V. So for your speaker to get the full rated power, it needs at least 18V from your amp. Higher pre-amp voltage = easier to reach the rated output to your speakers, since less gain is required.
Also, I mentioned noise floor. It's explained well here -
http://www.glasswolf.net/papers/preouts.html
It's not so easy to explain in laymans terms, but just know that higher voltage pre-outs = higher signal to noise ratio (SNR), meaning more signal & less noise = lower noise floor.
The article does mention however that having too high of a pre-out voltage is not good either, as it can lead to clipping. Best to buy an adjustable line driver (the $30 amazon model I bought had this feature... this one -
https://www.amazon.com/SX310-Pre-Am...70255121&sr=8-1&keywords=electronic+crossover). Most devices that accept pre-outs max out at around 6V, some as low as 4V. So don't crank up the line driver to max... I leave mine around 75-80%.
Hope that helps.
I own a few expensive headphones and IEM's that cost more than this phone.. Why does SUPER ALWAYS sound worse than standard in the hifi sound option? Super sounds muffled on every earpiece I used.. Its like its backwards or something..
so is standard the snapdragon DaC and Super is AKM dac? Or is both AKM? Mostly bought this phone for its audio capabilities, I ride over 100 miles a week on a cycle and always tune out.
HELP!!!
velvethammer said:
I own a few expensive headphones and IEM's that cost more than this phone.. Why does SUPER ALWAYS sound worse than standard in the hifi sound option? Super sounds muffled on every earpiece I used.. Its like its backwards or something..
so is standard the snapdragon DaC and Super is AKM dac? Or is both AKM? Mostly bought this phone for its audio capabilities, I ride over 100 miles a week on a cycle and always tune out.
HELP!!!
Click to expand...
Click to collapse
Also interested in the answer to this, I thought I was going insane as on my custom IEMS standard sounds better than super which seems muffled but if I swap to a pair of cheaper earphones then super gives better results.
Want to know what each of these modes do in terms of DAC or filter being used?
Super sounds waayyyyy better than default on my A2017G using OnePlus Icons.
I should know as a music producer and audiophile.
The sound is a lot deeper and thicker. Also the bass is a lot cleaner and the high frequencies are also more clear and less 'dull'
keessonnema said:
Super sounds waayyyyy better than default on my A2017G using OnePlus Icons.
I should know as a music producer and audiophile.
The sound is a lot deeper and thicker. Also the bass is a lot cleaner and the high frequencies are also more clear and less 'dull'
Click to expand...
Click to collapse
I'd like to know how others are finding the difference between standard and super headset hifi. Pls also post what earphones/headphones you are using for this.
For me, JH Roxannes and Sony EX1000 sound better with standard. If I use Westone UM30 pro then the opposite effect, super sound better.
I can't believe the phone can tell the difference between 2 earphones and would change super and standard accordingly.
beyerdynamics custom one pro.
no dolby.
with super the instruments sound a lot clearer (except for the increased volume in bass frequencies) as in i hear every single vibration of an instrument. e.g. when a hi hat is hit i audibly "see" how it swings.
I don't use dolby either but standard sounds better on my Roxannes than super. Same effects as you experience but on standard vs super. Highs/Mids/Bass sound clearer in standard. Super sounds muffled in comparison and bass is increased but muddy.
Unless standard and Super are reversed on the chinese version A2017?
May be my earlier post in other thread can help to answer to your question regarding DAC.
http://forum.xda-developers.com/axon-7/how-to/stock-music-player-t3474463/post69069304#post69069304
The only thing is not clear yet is the analog part. I suppose that it is in separated, already not named chip and both DACs drive same amplifier. The volume difference between Standard and Super would be because of different analog signal level. It is my opinion but either I can imagine that it is on purpose indicating that something happened.
Why do you hear standard sounds better than Super? I don't know. I checked how standard sounds (good) but I always use super mode.
I brought this up in a previous thread, glad it's still being discussed.
One issue lies in confirmation bias. Anyone doing testing on their own and knowingly switching between Standard and Super could have psychologically skewed the results. Since the software named one setting "Super", the weight of average users opinions may bias towards that. The only way to truly test it would be to have a full blind test done.
Since even I don't want to bother anyone to help me set up such an involved test, I did some rudimentary testing with my Shure SRH840, Sony 7506, and Triple Fi 10. I'm finding that Super increases the overall volume, as well as increasing mid-bass response and some treble.
I have an unsubstantiated theory on this. First, Super setting could be a sales tactic equalizer designed to "wow" the average consumer. It's a well known audio industry strategy that preys on the human brain's tendency to rate slightly louder audio sources as higher in quality.
Secondly, this setting could be designed to give perceived improvements using low quality headphones. The fact that several of us observe increased mid-bass and treble, areas that cheap headphones sometimes severly lack, may be an indicator. Let's not forget that in the box, the Axon comes with a cheap set of headphones that are many buyers' first audio experience on this phone, which ZTE would want to leave an impression.
If this is true, it makes sense that many of us with higher end headphones with more accurately balanced frqeuency response would find the Super setting to be inferior in use. Again, none of my claims can be confirmed unless professional testing equipment and a blind test method is used.
I remain skeptical in regards to questions of this setting toggling between a Snapdragon DAC and a Axon specific Hi-fi DAC. When quickly switching between Standard and Super, there's virtually no pause in the audio playback, with linear fade outs and fade ins. It seems odd that if the phone was truly switching between sending the audio to two different hardware chips, a more prominent pause in audio playback should occur.
So is it true that you cannot use Dolby and process audio using the AKM DAC at the same time? I read that somewhere, you can use either one or the other
trace1er said:
So is it true that you cannot use Dolby and process audio using the AKM DAC at the same time? I read that somewhere, you can use either one or the other
Click to expand...
Click to collapse
I believe that's not true, it may have come about from Poweramp users. If DVC is disabled in Poweramp then Dolby can be used. I assume that Poweramp uses the AKM dac.
I'm curious how the Axon 7 is setup. If only the AKM dac is enabled or if the S820 dac is also enabled. I can't work out if or how to switch between them. If there is any way to show which dac is in use that would be great as it's all guessing right now.
Kantana said:
I believe that's not true, it may have come about from Poweramp users. If DVC is disabled in Poweramp then Dolby can be used. I assume that Poweramp uses the AKM dac.
I'm curious how the Axon 7 is setup. If only the AKM dac is enabled or if the S820 dac is also enabled. I can't work out if or how to switch between them. If there is any way to show which dac is in use that would be great as it's all guessing right now.
Click to expand...
Click to collapse
This would be a good thing. Wonder if there's an app to do it, or possible to make one.
Kantana said:
. ... if the S820 dac is also enabled.
Click to expand...
Click to collapse
It isn't.
obladi64 said:
It isn't.
Click to expand...
Click to collapse
If that's the case then no concern over super vs standard. Both modes use the AKM dac and the difference being heard is software/filters/dsp being applied
We could push the argument to say that standard is the pure output out of the dac with no coloration being added so no surprise this sounds most transparent, clear with decent earphones.
Super would be for lower quality earphones to add back elevated bass/highs so artificially compensating for hardware defecits.
Kantana said:
If that's the case then no concern over super vs standard. Both modes use the AKM dac and the difference being heard is software/filters/dsp being applied
We could push the argument to say that standard is the pure output out of the dac with no coloration being added so no surprise this sounds most transparent, clear with decent earphones.
Super would be for lower quality earphones to add back elevated bass/highs so artificially compensating for hardware defecits.
Click to expand...
Click to collapse
Not at all.
I've tested with several headphones and the super settings adds a lot more depth. Default sounds a bit flat. Not in terms of bass or treble, but widening, clearance, and especially the low-end sounds a lot muddier.
Kantana said:
If that's the case then no concern over super vs standard. Both modes use the AKM dac and the difference being heard is software/filters/dsp being applied
We could push the argument to say that standard is the pure output out of the dac with no coloration being added so no surprise this sounds most transparent, clear with decent earphones.
Super would be for lower quality earphones to add back elevated bass/highs so artificially compensating for hardware defecits.
Click to expand...
Click to collapse
Not exatly. If you use flat mode there is no colorisatin in both. As I know 4491 is not able to oversample and convert 16 or 24 bit to 32. 4490 is able to. It should make sound better theoretically. Beyond these 4490 has many switchable filter option. Don't mix it with DSP. Usually it is called soft or sharp mode etc mode. Unfortunately ZTE doesn't make it availabe for user and I don't know what is the default selected option but difference between these option is so tiny that you can judge just after long term listening if you can at all.
Kantana said:
I don't use dolby either but standard sounds better on my Roxannes than super. Same effects as you experience but on standard vs super. Highs/Mids/Bass sound clearer in standard. Super sounds muffled in comparison and bass is increased but muddy.
Unless standard and Super are reversed on the chinese version A2017?
Click to expand...
Click to collapse
Something is wrong with your phone in particular. Reset it or send it back.
I saw a friend with Axon 7 dealing with exact same issue.
I got mine and its working like its supposed to. Super enhances the sound and dolby gives the open dynamics.
Kantana said:
If that's the case then no concern over super vs standard. Both modes use the AKM dac and the difference being heard is software/filters/dsp being applied
We could push the argument to say that standard is the pure output out of the dac with no coloration being added so no surprise this sounds most transparent, clear with decent earphones.
Super would be for lower quality earphones to add back elevated bass/highs so artificially compensating for hardware defecits.
Click to expand...
Click to collapse
Not both mode use AKM DAC but use different type of AKM DAC. 4490 is more enhanced than standard 4691. 4490 can oversample and convert bit depth but I already said it in earlier post. There is no software trick to make it "better" on cheap headphones. A7 was made for audiophiles too, why would they have done that? For this purpose a tone control would have been enough. I suggest to check how another A7 phones sound, may be yours has problem.
Otherwise today I listen music in standard mode too. Comparison was a bit difficult because I cannot set same volume level. The difference is 1.5 click between two mode. Never mind my intension was and long term test. I played only CD quality music and there is no question about that both DAC sound very good but differently. Standard sounds like typical CD player. It is a typical sterile digital sound. The Super is full with air and space. Everything is more clear and enjoyable. It sounds analog way. After this I don't know whether is it worth to compare with HD audio record too?
Now that would be a surprise to me as I saw AK4490 was used as DAC for playback but AK4961 was used as ADC for recording. I never considered they would have both wired for standard n super.
Where can we find supporting documents for these 2 modes since even manual skips on details and just says you can change it within headset hifi?
Kantana said:
Now that would be a surprise to me as I saw AK4490 was used as DAC for playback but AK4961 was used as ADC for recording. I never considered they would have both wired for standard n super.
Where can we find supporting documents for these 2 modes since even manual skips on details and just says you can change it within headset hifi?
Click to expand...
Click to collapse
4691 is ADC and DAC in one chip while 4490 is only DAC. I haven't found well detailed document. All you can do in this question is that you read ZTE and AKM releases carefully. Also you can find some detail on site like Android Authority or Youtube etc.
obladi64 said:
4691 is ADC and DAC in one chip while 4490 is only DAC. I haven't found well detailed document. All you can do in this question is that you read ZTE and AKM releases carefully. Also you can find some detail on site like Android Authority or Youtube etc.
Click to expand...
Click to collapse
From ZTE site, it looks confusing:
Playback
Sound Playback Codecs: Dolby Atmos® + Dolby Digital Plus (7+1 Surround Sound)
Sampling Frequency: 384 kHz
THD+N: '112dB
S/N: 120 dB
Recording
Channel: 3 channel (HiFi),1 channel (VAD)
Resolution: 32 bit as per spec (24 bit as per performance)
Sampling Frequency: 96 kHz
MIC max input: 2.02 Vpp
S/N ADC: 102 dB
Conference Mode
Directional Recording: Dimensional and Omni Directional
Recording Range: Up to 7 meters (approximately 23 feet)
Dedicated Audio Chipset: AKM 4961 + 4490
- 4961 - HiFi main chipset to provide HiFi and recording features
- 4490 - Independent amplifier to enhance HiFi playback (AK380)
It makes it sound like 4961 is dac and 4490 is the amp. At least for the conferencing, for playback it only mentions the dolby which is software. It also adds (AK380) to the line mentioning 4490!
Important update 11/9/18:
Android Pie seems to have fixed the digital gain defaulting to max. This means that the stock implementation is now optimal and using the gain control app is not needed (unless you want more voltage from a Dragonfly Red).
This was not announced anywhere to my knowledge, but is very pleasant surprise. I found it accidentally while testing the new dongle (above link).
Updated: Edit--I did some further tests and unfortunately the gain optimization only seems to work with line in mode, not with a "forced" headset scenario. So with a headset inserted, the distortion bands are still present.
As of the Nov patch, Pie on the Pixel has fixed both the line out and headset mode distortion banding when using the Google DAC.
When looking at the tracings, assume that stock will look as good as the optimized.
Made a simple spreadsheet with data so far:
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
The above numbers are using my gain control app, android volume at 25/25 to produce the best possible output signal
Note, the voltage at distortion threshold is the point where harmonic banding becomes >5 dB.
The THD is the harmonic distortion at 250 mV, this would be about 90 dB using full size cans like the Senn 650/800 that have a sensitivity of 103 dB/V.
Will work on uploading the exact images if anyone is interested.
Stepped sine tests at higher output (750mV) - An attempt to see how high impedance, lower sensitivity full size cans will effect distortion.
Edit: I updated the measurements and comments
Edit: Tested Apple USB C Dac
Edit: Tested the HTC dongle
Edit: Tested the Razer Dac
Edit: Tested the Realtec Dac
Edit: Tested the Geekria Dac
Edit: Tested the Audiolab USB device
Edit: Brief test of Dragonfly red. I would consider this to be the gold standard, very low impedance (<.1), minimal crosstalk, high volt out (not on stock), very low distortion.
Edit: There appears to be 3 (at least) "modes" that the Google dongle will do. Depending on the impedance of the headset/line out device attached, the output voltage varies quite a bit.
Low impedance - .36 Volt (high efficiency iem's)
High impedance - .9 Volt (full size headphones)
Very high impedance- 1.8 Volt (line out)
What gets interesting is the effects on testing. The "reference" sites like GSMarena, are testing with a Y cable setup (as I was) and will see high voltage out even with headphones attached since the use case was taken from the line out initially. What made me aware of this was that my app was not reading the Spl correctly, as if the max voltage was about 7 dB down (which it was). Also of note, the reduction in output is not due to a change in the Dac digital gain (the one that adjust from 1 to 175), so the distortion bands are just as bad since they kept that gain at 175.
RMAA:
Used both a Senn 800S and Shure 846 as load, as well as no load.
Note the flat FR on the SE 846 which verifies the low impedance above.
So superb impedance, great voltage out, pretty good crosstalk (with low impedance iems).
But, significant distortion using stock gain settings:
And it has 25 volume steps (each tick is 2.3 dB).
going to subscribe for those that have invested in 3.5mm portable audiophile amps/iems.
Nice work. Subscribed!
By the way, very nice headphones. I'm looking forward to testing out LDAC on my MDR-1000Xs when my Pixel arrives.
Hi bjmrd,
So I was looking at the Phone arena review in regards to the audio quality through the dongle.
It showed this
Not sure if this means the 2XL is weaker than the 2?
What does your test conclude? Would this device provide a nice audio experience for the user ?
This is important to me as I use 3.5mm headphones.
Thanks in advance,
Previous devices are the V20 (superb) and the non DAC G6.
I double checked my voltage measurements and they seem correct (compared to a dragonfly red).
The dongle may contain a DAC (like apple lightening). It is possible that with root, ALSA, some of the distortion can be reduced if so. Certainly the voltage is good, impedance great.
Oh and each volume tick is about 2.5 dB.
Addendum - The dongle definitely has a DAC
I have ordered a USB audio pass through cable to see if onboard analog is enabled
bjrmd said:
I double checked my voltage measurements and they seem correct (compared to a dragonfly red).
The dongle may contain a DAC (like apple lightening). It is possible that with root, ALSA, some of the distortion can be reduced if so. Certainly the voltage is good, impedance great.
Oh and each volume tick is about 2.5 dB.
Addendum - The dongle definitely has a DAC
I have ordered a USB audio pass through cable to see if onboard analog is enabled
Click to expand...
Click to collapse
bjrmd when you have a minute can you explain some of the numbers and what they might mean to the layperson. Why is high voltage good and low impedance good.
jawmail1 said:
bjrmd when you have a minute can you explain some of the numbers and what they might mean to the layperson. Why is high voltage good and low impedance good.
Click to expand...
Click to collapse
The max voltage is what limits the overall volume, so if you are driving over the ear cans and need power, this is helpful. Not important for sensitive iem's.
Crosstalk--a mixing of the R and L channels- not intentional. Better crosstalk--more stereo separation.
Impedance- internal resistance of the amplifier--if this is too high, causes frequency response changes depending on the type of iem and the iem impedance.
Thanks for the definitions bjrmd!
A couple of key points:
The dongle is a USB DAC, so basically audio out as supplied by Google is determined by that cheap little item. You could get a better USB DAC like a dragonfly, but it is bigger.
The distortion is based on some sort of driver parameter issue. I think the reason will end up being that the dongle DAC is set for max digital gain which is not optimal (unless needed). The following test shows why:
I used the Dongle on my Pixel orig and Pixel 2 with the Neutron player usb driver and the signal was clean.
This is a tracing, the same on both devices.
This is a dragonfly red, virtually the same:
So, I think the Google designers did not properly design the software, but the hardware is solid. In the meantime, use the Neutron player USB driver and you will have a very clean signal.
bjrmd said:
The max voltage is what limits the overall volume, so if you are driving over the ear cans and need power, this is helpful. Not important for sensitive iem's.
Crosstalk--a mixing of the R and L channels- not intentional. Better crosstalk--more stereo separation.
Impedance- internal resistance of the amplifier--if this is too high, causes frequency response changes depending on the type of iem and the iem impedance.
Click to expand...
Click to collapse
Thanks:good:
Dongle optimization on the old Pixel:
Alsa installed (need root, so can't do on the new Pixel yet).
Dongle card output:
numid=2,iface=MIXER,name='Headphone Playback Switch'
; type=BOOLEAN,access=rw------,values=1
: values=on
numid=3,iface=MIXER,name='Headphone Playback Volume'
; type=INTEGER,access=rw---R--,values=2,min=0,max=175,step=0
: values=18,18
| dBminmax-min=-175.00dB,max=0.00dB
numid=4,iface=MIXER,name='Extension Unit Switch'
; type=BOOLEAN,access=rw------,values=1
: values=off
numid=1,iface=PCM,name='Playback Channel Map'
; type=INTEGER,access=r----R--,values=2,min=0,max=36,step=0
: values=0,0
It seems that the android software defaults the dongle digital gain at max:
This leads to the distortion bands previously seen.
But, if we cut the dongle DAC gain down, and raise the android Volume to net the same(or even a bit higher) voltage:
The bands are gone.
In addition we can get 1 volt out without too much distortion as well:
This is not that difficult to do, but root is needed.
I will try to do some more detailed testing over the weekend, but I wanted to get this up so if someone was on the fence due to audio issues, if you are willing to wait for root and a few mods, this is a reasonable solution.
BTW for 9 bucks the dongle is pretty good with the correct gain settings
do you have test the ldac on google pixel 2 xl ?
is it working? or any change?
Does anyone else think it's a bit ridiculous how large the adapter is? It's huge compared to Apple's.
Thanks for posting this! Would you say this adapter is superior to Apple's lightning adapter? There are a few audiophile sites that have posted measurements for it and say it's great, but I'm not sure how to compare it to this.
Here is one such review: http://www.kenrockwell.com/apple/lightning-adapter-audio-quality.htm
yangyangyanglei said:
do you have test the ldac on google pixel 2 xl ?
is it working? or any change?
Click to expand...
Click to collapse
See first post.
In order to get digital gain to change in the dongle Dac, root is needed, hence tested the dongle on original Pixel in post 11.
The results should be identical .
Nitemare3219 said:
Does anyone else think it's a bit ridiculous how large the adapter is? It's huge compared to Apple's.
Thanks for posting this! Would you say this adapter is superior to Apple's lightning adapter? There are a few audiophile sites that have posted measurements for it and say it's great, but I'm not sure how to compare it to this.
Here is one such review: http://www.kenrockwell.com/apple/lightning-adapter-audio-quality.htm
Click to expand...
Click to collapse
The dongle is a reasonable USB DAC, but the stock implementation (at least on my setup) is poor. Google has the DAC digital gain at max, which introduces significant distortion.
To backtrack, the net powerout of this device is dependent on the Android system volume + the digital gain of the DAC itself. So you could have a very low android volume and very high DAC gain (as it is now) or a higher android volume and lower DAC gain (optimal for distortion).
My dragonfly red will behave this way as well in regards to distortion.
So unless you want maximum power out, we want android volume higher and DAC gain lower.
The impedance is lower than Apple's, power out may be higher as well. Not sure about crosstalk but both are probably very good.
Size wise- I couldn't even tell there was circuitry in the dongle-they did a good job IMO in the size.
Bottom line-- the audio out of the new Pixel is basically dependent on what USB DAC used. The stock is fine with the above noted. In fact for all you know, another USB DAC may also default to max gain and distort as well.
bjrmd said:
The dongle is a reasonable USB DAC, but the stock implementation (at least on my setup) is poor. Google has the DAC digital gain at max, which introduces significant distortion.
To backtrack, the net powerout of this device is dependent on the Android system volume + the digital gain of the DAC itself. So you could have a very low android volume and very high DAC gain (as it is now) or a higher android volume and lower DAC gain (optimal for distortion).
My dragonfly red will behave this way as well in regards to distortion.
So unless you want maximum power out, we want android volume higher and DAC gain lower.
The impedance is lower than Apple's, power out may be higher as well. Not sure about crosstalk but both are probably very good.
Size wise- I couldn't even tell there was circuitry in the dongle-they did a good job IMO in the size.
Bottom line-- the audio out of the new Pixel is basically dependent on what USB DAC used. The stock is fine with the above noted. In fact for all you know, another USB DAC may also default to max gain and distort as well.
Click to expand...
Click to collapse
Here's the size comparison.
bjrmd said:
I double checked my voltage measurements and they seem correct (compared to a dragonfly red).
The dongle may contain a DAC (like apple lightening). It is possible that with root, ALSA, some of the distortion can be reduced if so. Certainly the voltage is good, impedance great.
Oh and each volume tick is about 2.5 dB.
Addendum - The dongle definitely has a DAC
I have ordered a USB audio pass through cable to see if onboard analog is enabled
Click to expand...
Click to collapse
Can you post a link of the pass through cable you are using or reccommend?
---------- Post added at 01:57 AM ---------- Previous post was at 01:50 AM ----------
I may be misunderstanding the tech involved here but what is the best way to get the best wired audio here? Root tweaking the existing dongle DAC or doing something like a pixel rooted with viper4audio with a pass through dongle?
Jayme Helgerud said:
Can you post a link of the pass through cable you are using or reccommend?
I may be misunderstanding the tech involved here but what is the best way to get the best wired audio here? Root tweaking the existing dongle DAC or doing something like a pixel rooted with viper4audio with a pass through dongle?
Click to expand...
Click to collapse
A pass through cable would be using the phones DAC/SOC (qualcomm) and analog audio out, just like with a headset jack. Unfortunately, @chdloc did a brief test with a Pixel 2 (not XL) with that type of cable and it did not play. Also, I bet the analog out would have higher impedance , worse crosstalk than a USB device. Either way you need an adapter, so a dongle or pass through.
So far the only way to get wired audio is through using a USB DAC. The dongle is a full fledged USB DAC. It has a fixed sample rate of 48K but otherwise excellent specs(by itself). Even if you wanted a high sample rate DAC, the android driver/audioflinger only puts out 48K.
As mentioned above, the stock gain settings cause distortion (at least in my case). After rooting this will be easy to fix. Right now, Neutron player has a custom USB driver that prevents that distortion (by also keeping the DAC gain lower) - that is what I am using.
Maybe Google will read this and fix the issue in the next update.
I can't give an opinion on Viper.
bjrmd said:
A pass through cable would be using the phones DAC/SOC (qualcomm) and analog audio out, just like with a headset jack. Unfortunately, @chdloc did a brief test with a Pixel 2 (not XL) with that type of cable and it did not play. Also, I bet the analog out would have higher impedance , worse crosstalk than a USB device. Either way you need an adapter, so a dongle or pass through.
So far the only way to get wired audio is through using a USB DAC. The dongle is a full fledged USB DAC. It has a fixed sample rate of 48K but otherwise excellent specs(by itself). Even if you wanted a high sample rate DAC, the android driver/audioflinger only puts out 48K.
As mentioned above, the stock gain settings cause distortion (at least in my case). After rooting this will be easy to fix. Right now, Neutron player has a custom USB driver that prevents that distortion (by also keeping the DAC gain lower) - that is what I am using.
Maybe Google will read this and fix the issue in the next update.
I can't give an opinion on Viper.
Click to expand...
Click to collapse
Will the neutron driver affect music played with other apps? I assume not...only music you've downloaded and are playing in the neutron app. It sounds like (haha) I'm going to be looking forward to some system level driver mods someone (hopefully) develops for a rooted pixel. Thanks for the insights!
Jayme Helgerud said:
Will the neutron driver affect music played with other apps? I assume not...only music you've downloaded and are playing in the neutron app. It sounds like (haha) I'm going to be looking forward to some system level driver mods someone (hopefully) develops for a rooted pixel. Thanks for the insights!
Click to expand...
Click to collapse
Sorry, no - only works in Neutron player.
Once root comes I plan on making something like this but will be much easier since only USB audio needs control. I would like some type of Spl level info as well to save our ears.
As an audio-lover coming from HTC's and later the Google Pixel devices, the Xperia 5 II got me very excited not only for retaining Sony's signature dual front-firing speaker setup but also for re-introducing the headphone jack. I understand that I am very much in the minority here with my wants and needs but, having access to a wealth of audio gear and competing audio-focused smartphones, I thought it would be helpful to draw some attention to one of the 5 II's best-performing features that hasn't been given justice yet online. Jump to the speaker quality section to skip the rambles.
To add context, I've owned the Xperia Z2 and loved it overall. However, while it ticked a lot of boxes, I was left dissatisfied with the performance of each feature - i.e. the cam was noisy and slow, the speakers very tinny, the headphone out lacking volume and I haven't returned to Sony until now. Until then, I was content with HTC's approach, their Boomsound stereo speakers were much richer and louder, the headphone jack implementations up to the U11 were also very good; the 10 especially had a dedicated audio section which delivered a very low noise floor, albeit a higher output impedance which made it a no-go for expensive multi-driver earphones.
Before the 5 II, I owned the Pixel 3 and 4, the Pixel 4 I found to offer the best speaker quality with good sound chamber design providing nice fullness and volume albeit at the cost of no headphone jack and abysmal battery size and life. So of course, audio is a priority for me. I have a wide collection of high-end earphones and headphones and like to travel light without an additional DAP, dongle or BT receiver. The 5 II is an audio lover's dream with regards to its speaker quality/volume/setup and headphone jack output. It also comes across as a substantially more refined Android experience than past Xperia's I've used to the extent that it rivals Pixel smartphones in fluidity.
Speaker Quality -
Though not the best on the market, of the devices I've owned, the Pixel 4 actually has the best speaker quality in terms of max volume, frequency balance and bass extension and I would say any user would be satisfied with the performance. The speaker setup on the 5 II is immediately more immersive and symmetrical - I'm not a huge fan of the popularisation of downward-firing woofer + earpiece tweeter as I find the volume discrepancy between the two distracting. The stereo setup is a big plus here for media consumption, and something I missed from my old HTC's and the PIxel 3. The 5 II, thankfully, also has far better speaker quality than previous Xperia's.
They are no longer tinny in the slightest with a nice body and balance to the midrange. With Dolby on, you even get a little bass punch in addition to a healthy jump in volume. There's more depth and fullness on the Pixel 4, but the difference was not as big as expected and I can personally forgive that given the real symmetry and stereo separation of the dual front-firing setup. Max volume is also excellent, easily as good as the Pixel and the audio reproduction is clearer and more balanced in the midrange. As such, vocals are easier to discern e.g. listening to videos while showering. I also don't find the 5 II to distort as much at max volume, where the Pixel 4 earpiece speaker became peaky and distorted on the top few notches. So though the Pixel may measure louder in some reviews, in use, the 5 II has the same useable volume range and both easily sufficient for listening to/sharing videos and music in loud environments.
It would be great to see Sony continue to work on the speaker quality in future models with a bit more bass extension, but this is definitely not an area where potential buyers should feel concerned that the 5 II is lacking.
Headphone Jack -
Onto the headphone jack, I have read that Sony have utilised custom amplifiers in their phones rather than the in-built audio from the Qualcomm chipset in previous models, and I wouldn't be surprised if this was also the case on the 5 II. In my eyes, a good audio source should provide 3 fundamental qualities:
- A low output impedance
- Ample driving power with low noise floor
- Linear frequency response
The 5 II ticks all boxes and it ticks them well for IEMs especially. It has a relatively low output impedance, just a few ohms at most. I tested by hooking up the Xperia and my desktop amp (THX789 with 1-ohm impedance) to the Campfire Audio Andromeda (5-driver BA earphone) with an in-line switcher. This revealed a slight bass drop-off, but pretty similar audio balance overall, a very good result. The Andro is notoriously sensitive to source impedance meaning that the vast majority of other IEMs, even high-end multi-drivers such as those from Empire Ears, Shure, etc, will sound accurate and faithful to their design on the 5 II - whereas, a higher OI will result in a shift in the sound signature.
Similarly, it has a flat response when measured using right RightMark Audio Analyser (Dolby off) and a black background with zero hiss even on the sensitive Andro. Of course, the detail retrieval and soundstage are not as strong as the THX desktop amplifier, but this is one of the best phones I've used in terms of audio output; it is simply clean and balanced. Even the LG V30 with its ESS audio hardware is not vastly more resolving and its output impedance is only slightly lower (measured at 1.5ohms) so both are equally good choices for the audio conscious. Power output is on the lesser side but easily adequate. I required 60% volume to get the Focal Clear to a comfortable listening volume leaving plenty of headroom for listeners preferring higher volumes. The THX amp obviously had a tighter, more extended bass but surprisingly, the Sony was not too bad, delivering a balanced sound and good soundstage. There was no bass drop-off and good detail retrieval. Impressive for a phone driving a high-end full-size headphone.
TLDR -
I think this is a pretty impressive smartphone all around and it's good to see that being recognised by critics and users. I am not personally hugely enamoured by the camera performance coming from the Pixel but there's definitely potential there especially if you like to tinker in post. The screen with warm colour temp in creator mode is a sight to behold with great accuracy to boot, and the battery life and audio performance are both sensationally good. In fact, 2-day battery life is very achievable with 3hrs SOT per day, not something I could say about any phone that I've owned prior. So hope this write up helps hobbyists on the fence. If you're looking for a good all-rounder smartphone with a clean headphone out, this is a great option!
Thanks a lot for this comprehensive writeup!
I'm in a similar situation to you, except I jumped ship from my previous Xperia's a bit later, ending up with a HTC U11+. Unfortunately the USB port has became faulty, meaning that I can't use the Usonic earphones that came bundled with the device. Flashing Android 10 to the device using a GSI image also means I can't even use a dongle anymore. Rough.
I had already planned to get the Xperia 5 last year, but managed to keep using my U11+ because it is still lasting me all day and I was afraid of having to face a worse camera in the X5 compared to the U11+.
With this post I think I'm pretty sure I'll love the X5 ii, it confirms everything I was thinking and have seen, from the display quality, the audio chipset, the speaker performance, etc.
I'm curious, on your profile it says you're a Physiotherapist but your level of understanding of tech is quite advanced for someone without any formal design/engineering background. I'm studying Electrical/Electronic Engineering at the moment, but I can say that a lot of my knowledge is from before I started my degree studies.
Shrenade514 said:
Thanks a lot for this comprehensive writeup!
I'm in a similar situation to you, except I jumped ship from my previous Xperia's a bit later, ending up with a HTC U11+. Unfortunately the USB port has became faulty, meaning that I can't use the Usonic earphones that came bundled with the device. Flashing Android 10 to the device using a GSI image also means I can't even use a dongle anymore. Rough.
I had already planned to get the Xperia 5 last year, but managed to keep using my U11+ because it is still lasting me all day and I was afraid of having to face a worse camera in the X5 compared to the U11+.
With this post I think I'm pretty sure I'll love the X5 ii, it confirms everything I was thinking and have seen, from the display quality, the audio chipset, the speaker performance, etc.
I'm curious, on your profile it says you're a Physiotherapist but your level of understanding of tech is quite advanced for someone without any formal design/engineering background. I'm studying Electrical/Electronic Engineering at the moment, but I can say that a lot of my knowledge is from before I started my degree studies.
Click to expand...
Click to collapse
Thanks! HTC made some great smartphones, wish they'd done better after the original One, such a goodie.
And yep, I studied Physio in my undergrad but tech/audio is my passion. I mostly lurk on forums reading the work of far more intelligent writers. On the side, I run an audio review website and through both I've picked up some knowledge - about audio especially.
I definitely am not a professional here in any way and love the perspective true pros bring to the industry, always a fascinating read. Hope you're enjoying your studies!
As a lurking audiophile, I wholeheartedly second this! I find that LG headphone jack still wins by virtue of having more volume steps. Sound quality wise, it is truly barely discernible given my portable usage of these two in my daily commutes.