Qudelix-5K Bluetooth DAC & Headphone Amp

[Eskil]

1: Yes, we've done vast amounts of blind tests, and to take an example: in a team working on power amps, we all could discern which one had the 6dB more loop gain in the midrange, which halved the 3rd order intermodulation products as well as halved the THD+n products (originally ~0.003%), as well as reduced output impedance. Pulling out the example, as it related to the THD levels we are talking about. I think the many blind tests taught me a bit on what level of difference I need to hear to be able to trust it is not psycoacoustics.
2: Amir stated: "So as a desktop DAC, it would not be a good choice although it would work.", based on the specs. So his opinion is that the specs should be better for that.
3: I think I'll end it here, and start to think that your insinuations are illusory.

 

[bobbooo]

1: Yes, we've done vast amounts of blind tests, and to take an example: in a team working on power amps, we all could discern which one had the 6dB more loop gain in the midrange, which halved the 3rd order intermodulation products as well as halved the THD+n products (originally ~0.003%), as well as reduced output impedance. Pulling out the example, as it related to the THD levels we are talking about. I think the many blind tests taught me a bit on what level of difference I need to hear to be able to trust it is not psycoacoustics.
2: Amir stated: "So as a desktop DAC, it would not be a good choice although it would work.", based on the specs. So his opinion is that the specs should be better for that.
3: I think I'll end it here, and start to think that your insinuations are illusory.

1. Extraordinary claims require extraordinary evidence. You've provided none whatsoever for your claims. As I said, you can easily rectify that by using the Distort program to generate and then ABX a distorted file to the original using Foobar2000, then post the log file and checksum of your successful passing of this test (at least 13/16 correct trials). The fact that you're not even willing to try this speaks volumes.

2. That's just from an engineering specs perspective, not in reference to audible sound quality - if you actually read the rest of the review you'd see the Qudelix was found to be audibly transparent with no discernable distortion at all. Plus the measurements that quote is referring to were taken with the incorrect sample rate setting on the Qudelix 5K, resulting in the source device resampling which increased distortion - the corrected measurements are in this post, showing ~5 dB higher SINAD than first posted.

3. There's no 'insinuation' here, just a request for evidence for extraordinary claims, on a science forum. And a presentation of the most likely explanation in lieu of none such evidence forthcoming - pricing (and other cognitive) biases. As suspected, like many who make such claims of audible differences between devices at miniscule levels of distortion, you are not willing to even attempt to back up these claims with scientifically valid evidence.

 

[Eskil]

May I remind you that a large part of scientific findings are started on interesting perceived observations or suspicions in the beginning, treated not like facts, but as something where you examine if there is actually a fire behind the smoke. Before you publish a final paper, you have of course done what you could to verify observations. It is not stated that this is a forum for papers only, and general posts here don't implicate that. It would be obvious to the skilled reader that my comment is not a published paper, but the start of a work, where validity will be investigated in the process.

I've said that I can normally recognize this level of performance in a real product, based on my prior work. I don't claim to be able to hear artificially generated THD. In fact my findings indicate that simple, low order THD isn't that bad (alone).
That this makes you think that you are a kind of judge with jurisdiction over me, who can sentence me to a day or two's free work for you (what I normally spend on an ABX test sequence), is rather ridiculous. And if I don't right away agree to your sentence, there's the worst possible reasons for that. I do have a plan to investigate the tool you mention, maybe I can learn from it. But I'm not promise here where I will finally get with it, and you don't have the right to sentence me to anything.

In my over 20 years of working with, designing for, and consulting senior and principal audio engineers all over the world in making the best sound quality for the least money, I've never met insults, rude behaviour and disbelief in my professional capabilities, like what you've presented to me from your anonymous acronym. I see on the heading of the topics in this forum that being teased is a central part of this forum, as this is not very constructive, it will not be a place where I spend my hours to help and maybe learn.

 

[bobbooo]

May I remind you that a large part of scientific findings are started on interesting perceived observations or suspicions in the beginning, treated not like facts, but as something where you examine if there is actually a fire behind the smoke. Before you publish a final paper, you have of course done what you could to verify observations. It is not stated that this is a forum for papers only, and general posts here don't implicate that. It would be obvious to the skilled reader that my comment is not a published paper, but the start of a work, where validity will be investigated in the process.

I've said that I can normally recognize this level of performance in a real product, based on my prior work. I don't claim to be able to hear artificially generated THD. In fact my findings indicate that simple, low order THD isn't that bad (alone).
That this makes you think that you are a kind of judge with jurisdiction over me, who can sentence me to a day or two's free work for you (what I normally spend on an ABX test sequence), is rather ridiculous. And if I don't right away agree to your sentence, there's the worst possible reasons for that. I do have a plan to investigate the tool you mention, maybe I can learn from it. But I'm not promise here where I will finally get with it, and you don't have the right to sentence me to anything.

No-one's saying you have to write a scientific paper, just back up your subjective claims of audibility with objective data such as Foobar's log of a blind, level-matched ABX test. What were the test conditions of your 'prior work' that formed your belief that you can "recognize this level of performance"? Blind, or double-blind ABX? What exact devices were tested and what are their measurements? How many successful recognitions out of how many trials? Level-matched to within 0.1 dB? From this: "Though this is a nice product, it does rather recognizably, at uncalibrated volume sound less perfect than the 30x as expensive desktop products", it sounds like you haven't even level-matched, and your claims are presented as much stronger than 'suspicions' or a tentative hypothesis, which is what they should be at the very most.

That Distort software and Foobar's ABX tool allow for a simple, controlled way to determine distortion audibility with objective, verifiable logs, and is not limited to just low order THD - you can create a distortion profile very similar to the actual one measured here of the Qudelix 5K. What do you think is audibly different about artificially generated THD to THD generated by an actual device?

In my over 20 years of working with, designing for, and consulting senior and principal audio engineers all over the world in making the best sound quality for the least money, I've never met insults, rude behaviour and disbelief in my professional capabilities, like what you've presented to me from your anonymous acronym.

You need to stop taking this personally, I have not directed any insults at you whatsoever. It's a request for evidence on a science forum, nothing more. Appeals to authority are unscientific and irrelevant to subjective claims of audibility, if they weren't then we'd all be worshipping at the feet of Paul McGowan of PS Audio on here and his 46 years of experience in the audio industry designing and building audio equipment. But we don't, because he never backs up his subjective claims with scientifically valid, objective evidence. I look forward to you providing such evidence for your thus far unsubstantiated subjective claims that the Qudelix 5K is not audibly transparent.

 

[Atanasi]

Android only requires vendors to support SBC, and all other codecs are optional. Each manufacturer decides which codecs to include on their phones.

SBC is the required codec for all Bluetooth audio devices. Android compatibility guidelines recommend AAC and LDAC.
AAC support is required anyway on Android, and LDAC encoding is freely licensed to all Android OEMs.

 

[maramara]

Hi,

First post in this forum

Maybe dumb question, sorry...using my iPhone XS and connecting this device with the Apple USB/camera cable, can I listen to Qobuz?

I guess using local files would work ok with either the Apple player or anything else?

As I know nothing about wireless, which I don't prefer (I like cables, I'm old - and I know disconnections and odd noises are a thing with bluetooth), but I also don't like the massive Apple camera adapter...is there any kind of quality loss with wireless? CD quality is all I need.

I would really love to be able to apply the AutoEQ adjustments to my Ety ER4P/S using my iPhone without having to buy an Android DAP or bulky-ish external amps and after days of not finding literally anything with 10 band PEQ I finally found this

Thank you!

 

[m_g_s_g]

Maybe dumb question, sorry...using my iPhone XS and connecting this device with the Apple USB/camera cable, can I listen to Qobuz?

By all accounts you can. I received my recently purchased Qudelix only yesterday, and I have only tested it using BT/AAC with an iPad. My Android phone can play Spotify when connected as an external DAC/HP amp. I've ordered a "Lightning to Type-C OTG Cable" to use the 5K as an external DAC for my iPad, and it will take some time to arrive. I'll report back.

 

[m_g_s_g]

I have tested my Qudelix 5K with my Drop HD6xx, using Oratory and AutoEQ's PEQs. I like the sound better than playing without EQ (both from Android and iOS), but the volume doesn't get much higher than driving the headphones directly without amp and EQ, given that the settings impose a negative preamp gain (~ -7dB). This is with unbalanced cables and high gain. I guess I'll have to try with balanced cables.

Given the absurd / laughable prices for finished ones, I've decided to make my own cables. It's not that I can't afford the fancy finished ones , it's that I definitely don't want to contribute to this snake oil business.

 

[maramara]

By all accounts you can. I received my recently purchased Qudelix only yesterday, and I have only tested it using BT/AAC with an iPad. My Android phone can play Spotify when connected as an external DAC/HP amp. I've ordered a "Lightning to Type-C OTG Cable" to use the 5K as an external DAC for my iPad, and it will take some time to arrive. I'll report back.

Thanks for your reply, do you have Qobuz as well? My doubt is if they allow the use of external things like this, so it might work perfectly on Spotify but not on Qobuz. I will also try and see if there is any other place where to get this info.

 

[bobbooo]

As I know nothing about wireless, which I don't prefer (I like cables, I'm old - and I know disconnections and odd noises are a thing with bluetooth), but I also don't like the massive Apple camera adapter...is there any kind of quality loss with wireless? CD quality is all I need.

Unfortunately iPhones don't support LDAC (which has bitrates up to 990 kbps), only supporting standard SBC, and the AAC codec over Bluetooth at ~250 kbps, which is way short of CD quality (1411 kbps). AAC Bluetooth performance wasn't measured specifically for the Qudelix 5K here, but I'd think it would be similar to other products using the codec, such as the Loxjie D30:

The 'shoulder' of noise around the 1 kHz test tone seems a bit concerning, but it's a perceptual-based codec so it may still be audibly transparent to you. Just be aware that if you're playing or streaming any non-lossless music not in AAC format, the iPhone will be transcoding already lossy music to AAC to be sent over Bluetooth, which will degrade quality further.

 

[bobbooo]

I have tested my Qudelix 5K with my Drop HD6xx, using Oratory and AutoEQ's PEQs. I like the sound better than playing without EQ (both from Android and iOS), but the volume doesn't get much higher than driving the headphones directly without amp and EQ, given that the settings impose a negative preamp gain (~ -7dB). This is with unbalanced cables and high gain. I guess I'll have to try with balanced cables.

The Qudelix app already has headroom set for the equalizer, switchable between -6 dB and -12 dB (under Equalizer > Option > Headroom), so you can take this into account when setting the negative preamp gain in order to get more volume. So in your case you can just set the headroom to -6 dB and the preamp gain to -1 dB (giving a total of -7 dB), resulting in 6 dB more (doubling) volume.

 

[m_g_s_g]

The Qudelix app already has headroom set for the equalizer, switchable between -6 dB and -12 dB

Thanks! I didn’t get what this setting was intended for.

 

[bobbooo]

Thanks! I didn’t get what this setting was intended for.

I think maybe it's for the more 'casusl' user who just wants to quickly adjust EQ without having to worry about setting negative preamp gain. I'd prefer it if there was an additional '0 dB' (off) setting for the headroom so you could just enter in the exact negative preamp you want of an EQ preset without having to adjust for the headroom to get the max volume. Maybe I'll ask the developer to include that on the Qudelix forum actually, they're pretty receptive to adding new features to the app/5K it seems.

 

[bobbooo]

Given the absurd / laughable prices for finished ones, I've decided to make my own cables. It's not that I can't afford the fancy finished ones , it's that I definitely don't want to contribute to this snake oil business.

Ridiculous isn't it? I couldn't even find a reasonably priced balanced cable on AliExpress for my (Chinese) HifiMan HE4XX.

 

[jhaider]

I never initially read past the first post, because headphone amps are generally uninteresting to me and there was no indication in Amir's initial review that there was any useful selling point to this one. However, now that I know it has PEQ, one's on the way.

Question - is there any downside to just using it between a computer and a headphone amp (here, HeadRoom Desktop Amp, but the model shouldn't matter). That is to say, plug this box into a Thunderbolt dock, run analog line level to the headphone amp's inputs. The benefit I see from doing so is combine PEQ with HeadRoom's excellent crossfeed circuit.

Also, this "balanced" connector thing still befuddles me. I have no idea what that really means. My headphone amp has unbalanced (RCA) inputs. I assume some old 3.5mm to RCA cable from the basement will be fine, or?

 

[m_g_s_g]

Question - is there any downside to just using it between a computer and a headphone amp

You would be using the 5K as a DAC + EQ. The DAC has not “the best SINAD in town”, but is probably in line with a good phone/PC one. But the PEQ feature is great (10 bands), and the app gives you a lot of options to play with.

I assume some old 3.5mm to RCA cable from the basement will be fine, or?

Yes. You have two sets of outputs, balanced and unbalanced.

As an amp, for high impedance headphones (like my HD6xx - 300 ohms), the balanced output really makes a big difference. I converted today my headphone cable to balanced (not my cleanest soldering job ever, with such a tiny 2.5mm connector ), and now my volume slider rests at 75% max. With an unbalanced cable, even with the help of @bobbooo, it was at 90%. And I don’t usually like to listen that loud.

 

[jhaider]

You would be using the 5K as a DAC + EQ. The DAC has not “the best SINAD in town”, but is probably in line with a good phone/PC one. But the PEQ feature is great (10 bands), and the app gives you a lot of options to play with.

Sonically I'm not worried. I'm sure the DAC is as good as or better than HeadRoom's ca. 2006 implementation.
https://www.stereophile.com/content/headroom-desktop-da-headphone-amplifier-measurements

I'm far more concerned with practical issues. Will the battery management support an always-plugged-in use case, for example?

 

[m_g_s_g]

Will the battery management support an always-plugged-in use case, for example?

I only received mine a couple days ago; maybe somebody more experienced can chime in, but from this app captures, it seems that it will!

 

[maramara]

Question - is there any downside to just using it between a computer and a headphone amp (here, HeadRoom Desktop Amp, but the model shouldn't matter).

I'm not a super expert, so the answer might interest me as well, but why are you not just using a software solution since you use a computer as source? There's plenty of PEQ, and crossfeed solutions available.

 

[somebodyelse]

Any way to configure the PEQ without the phone app?