• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Qudelix-5K Bluetooth DAC & Headphone Amp

ZolaIII

Addicted to Fun and Learning
Joined
Jul 28, 2019
Messages
968
Likes
560
Only their marketing materials; I haven't checked lately for exactly the reason you state. Unless I want to go all-in on their fledging ecosystem it's not relevant to me. It's merely the curiosity of: did they pull it off and does it actually matter/is it audible?
In the case of BT it just might be... LDAC tops around 100 dB in high mode 96/24.
Never run on to actual UAT measurements nor a Huawei one. Would be very interesting to explore that, at least to me.
 

ZolaIII

Addicted to Fun and Learning
Joined
Jul 28, 2019
Messages
968
Likes
560
I think if I needed something better than Aptx hd or LDAC, I would just use USB for reliable lossless.
Point is wireless small and portable and sure that's why I am more interested in R3Pro & BT playback control only (on device stored content, media server's & cetera). But knowing you have & better BT codec (if not gimmick) is a nice bonus.
 

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
1,191
Likes
1,168
In the case of BT it just might be... LDAC tops around 100 dB in high mode 96/24.
Never run on to actual UAT measurements nor a Huawei one. Would be very interesting to explore that, at least to me.
All that seems a bit theoretical for now. I know there is still some bandwith left with BT, but for audio applications it will be hard to squeeze more bitrate and make it reliable, especially with the antenna limitations of such a small device. IF the OS doesn't fully support UAT, and the BT SOC neither, a software player could potentially bypass the OS audio engine/driver, but it's sketchy. I want to see some real world use, not only measurment. If it stutters at one foot, we can have all the fidelityin the world but it's useless. I obviously don't know their algorythm, but to my ears, even LDAC don't sound better than AptX HD. Theoretically it is better, but we don't know the real amount of interpolations when sample don't reach timely. Bluetooth will never be Wifi. Now let's wait and see, It's unfair from me to not at least give them benefit of the doubt, but in my opinion trying to reach more fidelity for bluetooth applications have limited appeal. That said, we have to at least respect manufacturers that are at least trying to innovate, but selling this to the big guys at Google and Apple, which in the end is the only answer to real market penetration, will be a rough path.
 

ZolaIII

Addicted to Fun and Learning
Joined
Jul 28, 2019
Messages
968
Likes
560
All that seems a bit theoretical for now. I know there is still some bandwith left with BT, but for audio applications it will be hard to squeeze more bitrate and make it reliable, especially with the antenna limitations of such a small device. IF the OS doesn't fully support UAT, and the BT SOC neither, a software player could potentially bypass the OS audio engine/driver, but it's sketchy. I want to see some real world use, not only measurment. If it stutters at one foot, we can have all the fidelityin the world but it's useless. I obviously don't know their algorythm, but to my ears, even LDAC don't sound better than AptX HD. Theoretically it is better, but we don't know the real amount of interpolations when sample don't reach timely. Bluetooth will never be Wifi. Now let's wait and see, It's unfair from me to not at least give them benefit of the doubt, but in my opinion trying to reach more fidelity for bluetooth applications have limited appeal. That said, we have to at least respect manufacturers that are at least trying to innovate, but selling this to the big guys at Google and Apple, which in the end is the only answer to real market penetration, will be a rough path.
Thers no big audience nor ther will be one. All do I support open source that's highly unlikely to expect from Chinese, however you do get a deacent app with integrated libs and deacent interface for free on both platforms. It's even more important on the iOS platform which is lacking both LDAC & USB 2.0 audio. For the use case that's fairly good distribution method all do it won't probably ever become anything widely adopted. A tip HiBy software had a wide adoption in the past regarding many Chinese manufacturers and after they failed to deliver some manufacturers turn to their own path in developing their own like for instance Shanling. Today HiBy software is much more mature and doesn't really lack futures all do some things remained annoying (like no EQ preset saving) for a sake of wider device compatibility.
My 10¢ on the topic & of course it would be desirable to get some real measurements & move to the empirical from the cloud of theory.
 

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
1,996
@amirm any chance you could check if the output impedance is sane? Sorry to bother :D

Well I never thought I'd say this, but Head-Fi has come up trumps here - a member there measured the Qudelix 5K's output impedance to be 0.2 ohms. It looks like this isn't a worry then, even if their value isn't hugely accurate (the measurements on here seem to be up to 1 ohm off anyway).
 
Last edited:

Veri

Major Contributor
Joined
Feb 6, 2018
Messages
7,824
Likes
9,421
Well I never thought I'd say this, but Head-Fi has come up trumps here - a member there measured the Qudelix 5K's output impedance to be 0.2 ohms. It looks like this isn't a worry then, even if their value isn't hugely accurate (the measurements on here seem to be up to 1 ohm off anyway).
If correct that is absolutely OK. Odd that support couldn't answer that simple question then, though.
 

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
1,191
Likes
1,168
That’s great. Interestingly, measuring very low output impedance precisely is not as trivial as it might seem. My guess, from a previous posterhere that asked Qudelix, is not so much they didn’t want to confuse customers, nor that their number was too high, but simply that their own numbers weren’t consistant enough, just a guess. Now yeah, anything under or around 1 ohm is a non issue
 
Last edited:

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
1,996
That’s great. Interestingly, measuring very low output impedance is not as trivial as it might seem. My guess, from a previous posterhere that asked Qudelix, is not so much they didn’t want to confuse customers, nor that their number was too high, but simply that their own numbers weren’t consistant enough, just a guess. Now yeah, anything under or around 1 ohm is a non issue

It seems like a pretty straightforward measurement that anyone can do really from this post on hydrogenaudio:
You need: 60 Hz sine test signal, digital multimeter, dummy load.

Measure the voltage unloaded and with dummy load.
Calculate:
Zout = (Rload * (Vnoload - Vload)) / Vload

Example:
We measure 1V unloaded, 0.9V loaded and our dummy load is 30 ohm ==> (30 * (1 - 0.9)) / 0.9 = 3.33 ohm output impedance at 60 Hz.
 

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
1,191
Likes
1,168
It seems like a pretty straightforward measurement from this post on hydrogenaudio:
It seems like a pretty straightforward measurement that anyone can do really from this post on hydrogenaudio:

Yes it's straightforward calculation, but in this example it's 3 ohms, but when wondering if it's 0.2 or 0.3, you need a scope, and it's easy to get a 30%+ error.
Like in this exemple if one mesure 0.99 v, you'll get .3 ohms, if you read .98 v you'll get 0.6 ohms, yes you need to change the load accordingly, but you see how it can get tricky. and it's a sine, not DC
 
Last edited:

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
1,996
Yes it's straightforward calculation, but in this example it's 3 ohms, but when wondering if it's 0.2 or 0.3, you need a scope, and it's easy to get a 30%+ error.
Like in this exemple if one mesure 0.99 v, you'll get .3 ohms, if you read .98 v you'll get 0.6 ohms, yes you need to change the load accordingly, but you see how it can get tricky. and it's a sine, not DC

Absolute rather than percentage error is what matters in this context, and ±0.3 ohms is ok here. As you say anyway, just knowing it's under or around 1 ohm is fine. Output impedance measurements are more of a pass/fail test really. (Obviously you'd need to set the multimeter to AC for the measurement.)
 
Last edited:

AlexScan

Member
Joined
Jan 23, 2020
Messages
26
Likes
13
Location
France
Wow i have found my new portable amp !

LDAC better than USB, very interesting, Qudelix in the pocket with 2.5 balanced cable connected on my Toneking 9T ... and my Samsung S9+ in my hand without cable to handicap my movements... great ! (must be working in LDAC)
 

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
1,191
Likes
1,168
Absolute rather than percentage error is what matters in this context, and ±0.3 ohms is ok here. As you say anyway, just knowing it's under or around 1 ohm is fine. Output impedance measurements are more of a pass/fail test really. (Obviously you'd need to set the multimeter to AC for the measurement.)
It’s ok for us of course, but from a manufacturer perspective, you can’t go and say, the output impedance is o.2 ohms plus or minus 0.3 ohms... I’m joking, it is easy to evaluate better than that, but the point was why they wouldn’t disclose it. It,s just my guess that they just don’t know themselves and rather than answer: it’s somewhere between 0.2 and 0.4, they would have chose just not to answer the question. Just a guess of course, but it’s not uncommon that small manufacturers ( i believe they are small, never heard of them before?) Don’t really have access to state of the art labs, and it’s ok.
 

Veri

Major Contributor
Joined
Feb 6, 2018
Messages
7,824
Likes
9,421
It’s ok for us of course, but from a manufacturer perspective, you can’t go and say, the output impedance is o.2 ohms plus or minus 0.3 ohms... I’m joking, it is easy to evaluate better than that, but the point was why they wouldn’t disclose it. It,s just my guess that they just don’t know themselves and rather than answer: it’s somewhere between 0.2 and 0.4, they would have chose just not to answer the question. Just a guess of course, but it’s not uncommon that small manufacturers ( i believe they are small, never heard of them before?) Don’t really have access to state of the art labs, and it’s ok.
Orrrr the customer service rep just dodged the question for no real reason :D
 

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
1,996
LDAC better than USB, very interesting

A surprising, and puzzling result for sure. Anyone have any idea how that can be? The only reason I can think of is this answer Qudelix support gave to a question on optimum sample rate settings on their forum (my emphasis):
USB DAC is a receiver, sink device.
It provides the supported sample rate and the device capability to the source device, PC, upon the initial connection.
A PC or laptop selects one of the sample rates and processes sample rate conversion if required.

If I understand this correctly, that means when playing a 44.1kHz file for example with the Qudelix 5K's sample rate set to 96kHz (as was the case for this review's measurements), the source device feeding the 5K will upsample, which may reduce SINAD - by how much is dependent on the quality of the resampler in the source device used and has nothing to do with the performance of the Qudelix. If it was correctly set to 44.1kHz sample rate a higher SINAD for the 5K over USB might have been recorded. I presume this wasn't an issue for the LDAC measurement however (the dashboard shows the correct 44.1kHz sample rate), so the 5K's USB DAC SINAD when set to the sample rate of the source (or to 44.1/48/88.2/96KHz mode which I believe is an option) could in fact be higher than the LDAC SINAD (as would be expected). Unless it's measured again set at 44.1kHz, how much higher the true SINAD may be if this is correct is unknown however.
 

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
1,191
Likes
1,168
A surprising, and puzzling result for sure. Anyone have any idea how that can be? The only reason I can think of is this answer Qudelix support gave to a question on optimum sample rate settings on their forum (my emphasis):


If I understand this correctly, that means when playing a 44.1kHz file for example with the Qudelix 5K's sample rate set to 96kHz (as was the case for this review's measurements), the source device feeding the 5K will upsample, which may reduce SINAD - by how much is dependent on the quality of the resampler in the source device used and has nothing to do with the performance of the Qudelix. If it was correctly set to 44.1kHz sample rate a higher SINAD for the 5K over USB might have been recorded. I presume this wasn't an issue for the LDAC measurement however (the dashboard shows the correct 44.1kHz sample rate), so the 5K's USB DAC SINAD when set to the sample rate of the source (or to 44.1/48/88.2/96KHz mode which I believe is an option) could in fact be higher than the LDAC SINAD (as would be expected). Unless it's measured again set at 44.1kHz, how much higher the true SINAD may be if this is correct is unknown however.
My guess is that Qudelix use the Qualcomm BT SOC as a usb receiver. They are quite bad at this task. it’s a side tought. Decent USB Dacs have A XMOs microcontroller, some receivers from TI or a few other chips good at this task. There is simply no room, or budget, to do it right on a product like that. and arguably, it’s a valid compromise. It’s clearly a portable device that most will use wireless.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
34,673
Likes
126,442
Location
Seattle Area
@amirm any chance you could check if the output impedance is sane? Sorry to bother :D
Sure. It is 0.4 ohm which is excellent. You can always tell the impedance is low when you have tons of power with 33 ohm load. Otherwise, most of it get lost in that test and results are poor which was not the case here.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
34,673
Likes
126,442
Location
Seattle Area
@amirm Any chance a quick remeasure of USB SINAD at 44.1 to put this question to bed?

Certainly a lot of member interest in this device..
Sure. Here you go:

Qudelix 5K Portable DAC and Headphone Amp USB 44100 kHz Audio Measurements.png


Strange thing is that if I change the sample rate above 48 kHz (e.g. 96 kHz), it reproduces the above tone at 500 Hz instead of 1000! Maybe this is the reason by default they don't enable other sample rates.
 
Top Bottom