That's disappointing, I wasn't expecting a Golfing panther but hey ho what the heck is this
The linearity is mostly low amplitude linearity. If you look at absolute value or interpret in absolute sense. 0.0x% is already very very low distortion. Too low distortion to cause any obvious differential or integral error in the linearity test.Not expecting Fiio to get headless panther, but this is an unbiased website.
Surprised to see high linearity but poor IMD tho, I always thought those are correlated but apparently not the case.
Add: Heck, I thought linearity is supposed to correlate with harmonic distortion according to EE lecture (HD coming from non-linearity) but the gap between SINAD and linearity result here is pretty huge.
This was my Q5s reviewed here
Do I hear some badass competition laughter? not pretty..Hahahaha
shit...
A5 was really good.Yikes! I own several other Fiio products, and previously considered the company to provide solid performance. This makes me reconsider things, since this really is broken.
This is terrible. In particular, is the DAC filter the worst of all the devices you have tested?This is a review and detailed measurements of the portable Fiio Q5s USB DAC and optional THX headphone amplifier (AM3D). It was kindly sent to me by a member. Q5s by itself costs US $300 from Amazon including Prime shipping. The AM3D THX AAA module costs an extra US $150 on Amazon including Prime shipping.
As it happens, the company was also kind enough to send one to me last year which I had yet to review! So I got a chance to test that version as well but briefly.
The form factor of Q5s is similar to other pricier stand-alone DAC/Headphone amps with chunky metal box:
View attachment 76539
There is a single micro USB connector at the bottom that provides both charging and also data connection to the PC. To use the THX module, you have to unscrew the standard one that comes with it ("AM3E") and plug in the THX one:
View attachment 76540
The volume control uses an ADC which converts is value and that is used to then adjust the level on both channels. This provided perfect channel matching but made me concerned about the performance of the volume control chip used in the middle.
On top you have Line In and Line Out. On the bottom on the THX module you have the 3.5 mm unbalanced jack and 2.5mm balanced.
DAC Audio Measurements
I connected the Q5s using USB and Line out and fired our normal dashboard test and got this disappointing result:
View attachment 76541
Look at the high spray of harmonic distortion to the right of our 1 kHz tone (the tall spike). I finished running all the DAC and headphone test which took two hours. I then went to graph the SINAD above in my spreadsheet and I see that I already had an entry for Q5s at SINAD of 100 db! What the heck?
I dig up the old unit with standard headphone amp and measure it and it too produces 90 dB SINAD. Frustration is starting to set in.
Aha! Maybe I used the headphone out. I plug the output cable into that and SINAD indeed shoots up to nearly 100 dB!
View attachment 76542
Back to the member unit with THX module it too measures well with headphone out:
View attachment 76544
With the headphone amp output, Q5s squeezes into second tier of all DACs tested:
View attachment 76545
At this point I realize all the tests are invalidated and I have to spend two more hours testing them all. Instead of punching my fist into the monitor, I calmed myself and ran the IMD test again, only to see no improvement whatsoever!
View attachment 76547
Why is IMD "SINAD" so poor at just 78 dB? The mystery was solved later when I ran the THD+N versus frequency:
View attachment 76548
Oh brother! Even getting rid of the effects of the poor DAC filter, there is a sharp rise in distortion as frequencies get lower. Dashboard SINAD is using a 1 kHz tone which doesn't see that. But IMD test tones do because it is the combination of 60 Hz and 7 kHz.
I fire up the analyzer at 20 Hz and get this horror show:
View attachment 76549
Our SINAD now is nearly half as much it was in the dashboard 1 kHz tone!
I ran a new test to show the issue in "time domain:"
View attachment 76550
The red sine wave is the 20 Hz tone. The blue one is the difference between it, and a pure 20 Hz sine wave, multiplied by 100 (otherwise the variations are too small to see). This is called the "residual" (null) signal. The highest variation is during negative swings of the power supply indicating poor capacity there.
Back to the DAC filter, this is what it looks like:
View attachment 76551
Gosh, here we go again catering to audiophile myths with such slow filters that allow a ton of ultrasonic junk to be produced by the DAC for no good reason. Worse yet, it causes the audible band to have an early droop:
View attachment 76552
As is typical, the roll off starts at 10 khz and by 20 kHz we are down a whopping 4 dB! The younger audience that uses this likely had the hearing that goes up that high so will be cheated out of some of the highs in their music. Maybe they will confused that for fidelity.
But the story is not on the right. It is on the left! What is that notch? At first I thought the analyzer was loosing data but it was not. Zooming in we clearly see the issue:
View attachment 76553
I tested the above on my other Q5s and it had the same problem with line out. Thankfully this is one test that improved with headphone out without this problem with either output module.
Why did they put this in there? It can't be an accident. Was there some audible noise there and they decided to filter it out or something? Blood is starting to boil big time by this point.
I had run the 32-tone test signal simulating "music" and got this poor output:
View attachment 76554
I thought no problem, I just switch to headphone output and the problem would be gone. No it was not! The output was identical out of the "THX" module! That power supply/low frequency issue is causing severe problem with this test.
Not everything is bad news. Dynamic Range is fine:
View attachment 76555
As is jitter:
View attachment 76557
Linearity is perfect:
View attachment 76556
Headphone Amplifier Tests
Here is our 3.5mm headphone output dashboard using the THX AAA module:
View attachment 76558
77 dB SINAD? How can this be? A THX module with this kind of performance? And this is in low gain with output not going much past 1.1 volt -- something a dongle can do better.
Switching to balanced output improves things:
View attachment 76559
But still nothing impressive.
Signal to noise ratio is worse than using the DAC:
View attachment 76560
View attachment 76561
The low output voltage kills any prospects of good power delivery into 300 ohm load:
View attachment 76562
At max gain and allowing for the distortion to rise we only have 18 milliwatts. If I measure at the knee of the curve as I usually do, power will drop to just 3 or 4 milliwatts!
I compared this to the standard module in high gain:
View attachment 76563
The noise level is a bit higher but both suffer identically from distortion and produce the exact same power.
Switching to 300 ohm gives us more power but still nothing that great:
View attachment 76564
Fortunately balanced output does deliver more power:
View attachment 76565
But with lots of distortion again.
I tested the frequency response and as I noted earlier, it doesn't have that notch:
View attachment 76566
Listening Tests
Please have some common sense. In the mood I am at the end of the above measurements puts me in no mood to listen to this device...
Conclusions
This was a painful review. So many variations to test. So many odd results. Two different devices. Two different amps. Three different outputs. Two different inputs. Maybe there is some explanation for the odd results that I am seeing but frankly, I don't have the patience to dig in more to find out. I see the words "THX" on a headphone amp and I expect world class performance. How did these measurements qualify for THX logo? What explains simple failures in frequency response of the line out?
I should say the company is very nice and cordial to deal with. But ultimately I have to go by what I have in front of me. Sadly it is not a functional product in my view. Yes, it produces sound. Yes, most people won't hear these issues but you are paying for great performance and you deserve to get it.
Needless to say, I can't recommend the Q5s with or without THX amplifier module.
-----------
As always, questions, comments, recommendations, etc. are welcome.
The panthers are asking for "fall" clothing. I keep telling them that is two months away but they don't listen. They say all the sales are now. With so many of them, no way I can afford to get them all new outfits. So please help me by donating what you can using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
No, Hi-Res files will not be reproduced. They might be accepted as input but the reproduction is sub-CD standard.So when I see a Hi-Res AUDIO sticker on a product, it means Hi- Res files will be reproduced, but the quality of the sound is not guaranteed at all.
Planning for an ASR sticker is a must from now on...!!!!
Please, do.
That is the very reason for me to ask for another sticker....No, Hi-Res files will not be reproduced. They might be accepted as input but the reproduction is sub-CD standard.
It is the last filter in the list I think, the "Super" slow roll-off which is close to no processing/no filter at all. In any case one of the worst picks if you have to make one..This is terrible. In particular, is the DAC filter the worst of all the devices you have tested?
Oh, I'm not disagreeing with you, just quibbling about "Hi- Res files will be reproduced".That is the very reason for me to ask for another sticker....
Seems I did not make myself understood...
Their spec is a lie.But the story is not on the right. It is on the left! What is that notch? At first I thought the analyzer was loosing data but it was not. Zooming in we clearly see the issue:
Frequency Response 5 Hz~90 kHz(-3dB)
It is the last filter in the list I think, the "Super" slow roll-off which is close to no processing/no filter at all. In any case one of the worst picks if you have to make one..
me too,at least they're updating the software on the M series.Yikes! I own several other Fiio products, and previously considered the company to provide solid performance. This makes me reconsider things, since this really is broken.