Chord Qutest

[Ralf11]

Thanks much.  RME says they have improved some things in the ADI-2, but I have not heard one.  I do have an Audio Research euphonicisor in my system which may help with 'overly clinical' inputs...

[Marco78]

question for qutest owners, since it does not have a switch off button, are the LEDs going off after some time or does it remain lit like a Christmas tree all the time ? In the latter case I can't see it as a viable option in a bedroom setup.

[Tone Deaf]

46 minutes ago, Marco78 said:

question for qutest owners, since it does not have a switch off button, are the LEDs going off after some time or does it remain lit like a Christmas tree all the time ? 

It stays lit like a Christmas tree.  Nice analogy by the way.  It is small enough that you could easily toss a cloth or something over it to dim the lights....

[jamesg11]

Box of dimlights for the xmas tree then!

Hmm, dunno with this re barrows’ non-chinese tick-box ...

[asdf1000]

5 hours ago, Marco78 said:

question for qutest owners, since it does not have a switch off button, are the LEDs going off after some time or does it remain lit like a Christmas tree all the time ? In the latter case I can't see it as a viable option in a bedroom setup.

 

Have you already tried the following, which is from the manual....

 

[pl_svn]

4 hours ago, Tone Deaf said:

It stays lit like a Christmas tree.

 

... but you can set it to "dimmed"  

[ecwl]

11 hours ago, Ralf11 said:

Thanks much.  RME says they have improved some things in the ADI-2, but I have not heard one.  I do have an Audio Research euphonicisor in my system which may help with 'overly clinical' inputs...

I have listened to the ADI-2 at my local head-if meet. I think it’s an excellent DAC chip based DAC. I personally prefer Mojo to the ADI-2 but I think people are so used to listening to DAC chip based DACs that I can see they might prefer ADI-2.

For me, the “advantage” of ADI-2 is the lower noise floor which creates this extra sense of transparency. This is a result of Mojo’s reference power, noise shaper and the 4-element pulse array DAC. The ADI-2 problem compared to Mojo is the noise floor modulation which makes music sound digital/bright/harsh so smooth, warm vocals or instrumental sounds often sound extra exciting with ADI-2 whereas they sound more natural with Mojo. There is also the transient timing issue so when you hear finger snapping or drum strikes, Mojo is more realistic but ADI-2 still sounds dynamic because even though the strikes are less precise, the noise floor modulation makes it sound more exciting. And by natural and realistic, I mean compared to live unamplified performances.

I have not heard Qutest in any system yet but I’ve listened to Hugo 2 via headphones at a stereo store and it far surpasses Mojo and Hugo and 2Qute in everywhere. Hugo 2 is much closer to DAVE than Mojo. So I suspect Qutest would have the same transparency as ADI-2 and all the advantages of Chord DACs.

 

Ultimately though, when we are spending that much money on ADI-2 or Qutest, you have to like the sound and how it integrates into your system. I’ve seen people swear they prefer their R2R or DSD DACs. And DAC chip based DACs are so ubiquitous (Sonos, iphones, just basically everywhere), I think people are so used to hearing their favorite music with imprecise transients and noise floor modulation from DAC chip based DACs, they are frequently preferred over other types of DAC designs. I think the challenge is that we rarely listen to live unamplified music so we forget what natural realistic music sounds like and the DAC chip based DAC sound is now our reference instead of live music. I definitely think anyone thinking about purchasing a DAC in Qutest price range should audition the Qutest though. If it’s not your cup of tea, don’t buy it.

[matthias]

3 hours ago, ecwl said:

I have listened to the ADI-2 at my local head-if meet. I think it’s an excellent DAC chip based DAC. I personally prefer Mojo to the ADI-2 but I think people are so used to listening to DAC chip based DACs that I can see they might prefer ADI-2.

For me, the “advantage” of ADI-2 is the lower noise floor which creates this extra sense of transparency. This is a result of Mojo’s reference power, noise shaper and the 4-element pulse array DAC. The ADI-2 problem compared to Mojo is the noise floor modulation which makes music sound digital/bright/harsh so smooth, warm vocals or instrumental sounds often sound extra exciting with ADI-2 whereas they sound more natural with Mojo. There is also the transient timing issue so when you hear finger snapping or drum strikes, Mojo is more realistic but ADI-2 still sounds dynamic because even though the strikes are less precise, the noise floor modulation makes it sound more exciting. And by natural and realistic, I mean compared to live unamplified performances.

I have not heard Qutest in any system yet but I’ve listened to Hugo 2 via headphones at a stereo store and it far surpasses Mojo and Hugo and 2Qute in everywhere. Hugo 2 is much closer to DAVE than Mojo. So I suspect Qutest would have the same transparency as ADI-2 and all the advantages of Chord DACs.

 

Ultimately though, when we are spending that much money on ADI-2 or Qutest, you have to like the sound and how it integrates into your system. I’ve seen people swear they prefer their R2R or DSD DACs. And DAC chip based DACs are so ubiquitous (Sonos, iphones, just basically everywhere), I think people are so used to hearing their favorite music with imprecise transients and noise floor modulation from DAC chip based DACs, they are frequently preferred over other types of DAC designs. I think the challenge is that we rarely listen to live unamplified music so we forget what natural realistic music sounds like and the DAC chip based DAC sound is now our reference instead of live music. I definitely think anyone thinking about purchasing a DAC in Qutest price range should audition the Qutest though. If it’s not your cup of tea, don’t buy it.

 

Did you encounter the noose floor modulation also when feeding the ADI-2 directly with higher rate DSD?

Thanks

 

Matt

[Ralf11]

Thanks ecwl.

 

Besides Matt's question can you also tell us if you heard the RMI & Qutest together in the same listening session?

[ecwl]

18 minutes ago, Ralf11 said:

Thanks ecwl.

 

Besides Matt's question can you also tell us if you heard the RMI & Qutest together in the same listening session?

I listened to my Chord Mojo, my Chord DAVE (no Blu2) and the RMI ADI-2 in the same session off headphones. I have never listened to Qutest. Closest was Hugo 2

[ecwl]

4 hours ago, matthias said:

 

Did you encounter the noose floor modulation also when feeding the ADI-2 directly with higher rate DSD?

Thanks

 

Matt

So nobody in my local head-fi meet upsamples from PCM to DSD512/DSD256. We were all listening straight off the DAC.

 

But more to the issue of noise floor modulation, it is definitely not something that many DAC designers talk about. And there seems to be a lot of people saying it doesn't happen or that when they upsample to DSD512 it solves the problem.

 

Truth is, noise floor modulation can happen at multiple stages. First, it's in the recording stage so it's possible for the original recording to have noise floor modulation, making it sound harsher. Second, it's in the digital playback stage so it's theoretically true that any DAC that takes say 44/16 and upsamples to DSD512 or 64-element 5MHz (Sabre DAC chip I think) can introduce noise floor modulation during the upsampling but I think that's actually exceedingly rare for any competent upsampling algorithm.

 

The last stage of noise floor modulation that can be introduced is at the actual Digital-to-Analog conversion. According to the designer Rob Watts, this happens with virtually all DAC designs for a variety of reasons, ranging from RF noise, jitter from the clock but most importantly, the DAC design itself, be it DSD/R2R/multi-bit PWM/DSD, including most modern DAC chips. The issue is that every time your DAC switches from 0's to 1's, or in the case or R2R DACs 256 to 16 or 18 to 238, there is noise that's generated in the rest of the circuit. Moreover, simple switching from 0 to 1 generates a different noise than switching from 1 to 0. The example Rob Watts give is that let's say you have silence and in DSD256, you can encode it with 101010101010101010101010101010101010 or 11001100110011001100110011001100. Technically they are equivalent int he digital domain but the first sequence would involve 32 switches and the other would involve 16 switches so when you're dealing with an actual DAC circuit, the amount of noise generated by these two sequences would be different. Moreover, the frequency of switching would also be different, generating a different noise pattern, despite being "silent". This is the reason why with loud music, the noise floor tends to be higher and with quieter music, the noise floor tends to be lower, leading to noise floor modulation. In addition, specific to DAC chips, because all the DAC chip elements are so close to each other on silicon, they are much more prone to noise. This is the reason why a lot of DAC designers have moved to discrete elements, ranging from DSD DACs to multibit DACs, like dCS. The solution Chord uses in the Pulse Array DAC is that using the 10-element in Qutest to represent say signal levels of 1-9 (through the noise shaper), the DAC is constantly switching at 104 MHz by flipping either two elements at the same time to maintain the same signal level (switching 0 to 1 in 1 element or 1 to 0 in another element) or only 1 element switches (either 0 to 1 or 1 to 0) to change the signal level (from say 4 to 5, or 7 to 6). As a result, the noise level fluctuation is constant regardless of the loudness of music you're playing. 

 

So in answer to your question at the beginning, no I did not get to listen for noise floor modulation off the ADI-2 with DSD256 or DSD512. But the bigger issue is that by design, ADI-2, as with other DAC chips, would have noise floor modulation regardless of what signals you feed it.

[buonassi]

Here, Rob talks about the noise floor modulation and how it is also present in amps, conductors (wires), etc.  I have to say, I recently heard mojo with my favorite IEMs at a meet in Chicago earlier this month, and was very impressed.  So much so, I'm considering the qutest, or a second hand 2qute at half the price.....

 

 

 

[ecwl]

4 hours ago, buonassi said:

So much so, I'm considering the qutest, or a second hand 2qute at half the price.....

Having listened to Hugo 2 vs Mojo and having a friend own 2Qute, I would say if you can afford it, go for Qutest. There are so many technical improvements from Qutest that the sound quality is far superior. Here are the main ones that I think matter the most. 

Qutest has 49000 taps for 16fs upsampling compared to 2Qute 28000 8fs followed by simple 2x FIR upsampling to get to 16fs. 

Qutest then has an additional WTA filter to get to 256fs whereas 2Qute uses a very simple FIR filter and this contributes a lot to sound quality. 

Qutest has 10 elements and a 11th? Order noise shaper whereas 2Qute has 4 elements with a 5th order noise shaper. 

Qutest can output 1, 2, 3V so it won’t clip your preamp whereas 2Qute only puts out 3V.

[buonassi]

good points, thx.  I think I'll try and nab a qutest. The overwhelming feature for me is the variable voltage output.  I did wonder about the 2qute possibly clipping at least one of my headamps.  This way, I can lower the voltage without having to use digital volume control.  

 

Thanks @ecwl..... I think you helped me make up my mind.  This will be a big milestone for me.  The most expensive gear I own (besides some of my in ears - which are ridiculously priced btw) is a metrum amethyst.  I'm trying to find a way to justify keeping it if I get the qutest.  That NOS sound is special, though not as detailed as a very good linear phase DAC.  I was considering the ADI-2 DAC, but I really don't need all those features.  

 

The qutest seems to be, by all accounts, and Darko's words "The DAC to beat at 2k".  pretty sure I can get one used for $15-1600.  But they don't last for more than a day after being listed on head-fi.  I have to be quick.  I'll report back after the score!

[pl_svn]

@buonassi just so you get an idea: a couple of years ago I bought a 2Qute and was surprised by how close it was to my beloved Metrum Hex (almost three times the price. the Metrum, though, had some more “magic” and was millimetrical in instrument positioning on the stage)

 The 2Qute stayed (damn’d DSD malarky  ) and was afterwards replaced by a Qutest

Then, a couple months ago, one day I tried once more the Hex and... immediately ordered an Adagio

[Marco78]

On 12/30/2018 at 2:05 AM, pl_svn said:

@buonassi

Then, a couple months ago, one day I tried once more the Hex and... immediately ordered an Adagio

 

Could you please elaborate ? was the Metrum a match for the Qutest ( which can be quite bright in my opinion) ?

Ciao

[pl_svn]

2Qute (and Qutest afterwards) were both extremely close, re sound signature, to the Metrum Hex. the Hex, though, was a tad more transparent and, as I wrote, instrument positioning was millimetrical. Detail was almost on par (still, maybe, the Hex had a tiny little bit more).

As at the time I was intrigued by DSDs... in the end I ditched the Hex 🙄 as trade off was very small

The Adagio is in a whole different league 🙂

Never heard a Dave but I believe that's the one to compare it with. The Adagio also has a very effective volume control and an extremely clever -10db setting that perfectly suites my power amp (FirstWatt J2) + speakers (Klipsch Cornwall) allowing me to omit a preamplifier (all those I tried had too much gain) 😶

and... yes: I kissed DSDs bye-bye (search Google for "Rob Watts DSD vs PCM") 😛

 

 

cold, warm, bright, dark, in my experience, do not pertain to a given component but are rather results of the whole chain (when months back I replaced a 2A3 SET with the FW J2 my system went from neutral to overly warm. replacing the Qutest with a Vega and Don Garber's "Y" pre with a Taurus made it damn' cold and... with the Adagio everything is back to what I like)

[Miska]

11 minutes ago, pl_svn said:

and... yes: I kissed DSDs bye-bye (search Google for "Rob Watts DSD vs PCM") 😛

 

IMO,You shouldn't just jump on unsubstantiated FUD claims that appear from time to time...

 

I already debunked at least one of his claims that he didn't even try to prove in first place.

[Miska]

On 12/29/2018 at 3:02 PM, ecwl said:

Having listened to Hugo 2 vs Mojo and having a friend own 2Qute, I would say if you can afford it, go for Qutest. There are so many technical improvements from Qutest that the sound quality is far superior. Here are the main ones that I think matter the most. 

Qutest has 49000 taps for 16fs upsampling compared to 2Qute 28000 8fs followed by simple 2x FIR upsampling to get to 16fs. 

Qutest then has an additional WTA filter to get to 256fs whereas 2Qute uses a very simple FIR filter and this contributes a lot to sound quality. 

Qutest has 10 elements and a 11th? Order noise shaper whereas 2Qute has 4 elements with a 5th order noise shaper. 

Qutest can output 1, 2, 3V so it won’t clip your preamp whereas 2Qute only puts out 3V.

 

For comparison I can use for example 16000000 taps 512fs upsampling and 7th order adaptive noise shaper to a DSD DAC that has 32 elements...

 

[w1000i]

22 minutes ago, Miska said:

 

IMO,You shouldn't just jump on unsubstantiated FUD claims that appear from time to time...

 

I already debunked at least one of his claims that he didn't even try to prove in first place.

 

With all respect to your experience you still talk in digital domain where he see things until it reache the speaker. Also he design for other companies we don't know even their names for many many years with a lot of resources and $.

I see many engineers and I'm one of them gives ideas and suggestion that sound good but experienced ones see something else where things get complicated. And sometimes even an experience technician with a lot of experience he but a new engineer in the corner

[Miska]

1 hour ago, w1000i said:

With all respect to your experience you still talk in digital domain where he see things until it reache the speaker. Also he design for other companies we don't know even their names for many many years with a lot of resources and $.

I see many engineers and I'm one of them gives ideas and suggestion that sound good but experienced ones see something else where things get complicated. And sometimes even an experience technician with a lot of experience he but a new engineer in the corner

 

No, I'm not. I'm talking about about things until they reach the speaker too! I would say it is more Rob Watts who is talking about digital domain. I've posted a lot of my measurement results from DAC outputs on this forum for example.

 

So far I've spent 20 years working on the DSP algorithms I use.

[pl_svn]

2 hours ago, Miska said:

 

IMO,You shouldn't just jump on unsubstantiated FUD claims that appear from time to time... 

 

I already debunked at least one of his claims that he didn't even try to prove in first place.

 

ok, so... let’s just say I’m extremely happy with Metrum DACs even though they don’t do DSD (and I can live without it) 😉

[Miska]

2 hours ago, pl_svn said:

ok, so... let’s just say I’m extremely happy with Metrum DACs even though they don’t do DSD (and I can live without it) 😉

 

I have no problem about that.

 

My Musette works OK when run at 352.8/384k. But I personally prefer Holo Audio more.

 

[buonassi]

@Miska, I have a few questions for you.  

 

1) Are all of your interpolation filters 'sinc' ?  I'm pretty sure at least a few are.  I don't have your software because I use AU plugins - otherwise, I'd be all about it.  The sinc is what Rob Watts goes on and on about.  That, and the noise floor modulation. 

 

2) I've had very good experiences upsampling in A+ using iZotope and Sox feeding my metrum amethyst, but have no idea what type of interpolation they use (cubic, sinc, etc).   Do you know?

 

3) I have to believe that your filters, being processed in the computer, contribute to lower noise floor modulation in the DAC.  Would that be a correct assumption?  And the fact that they use many more taps than the Qutest may contribute to better reconstruction of the original waveform (at least theoretically)? 

 

 

[Miska]

11 hours ago, buonassi said:

@Miska

1) Are all of your interpolation filters 'sinc' ?  I'm pretty sure at least a few are.  I don't have your software because I use AU plugins - otherwise, I'd be all about it.  The sinc is what Rob Watts goes on and on about.  That, and the noise floor modulation. 

 

Yes, there are quite a few, they have "sinc" in the name too. Although what Rob Watts exactly means by "sinc" is a bit ambiguous.

 

Reason why AU and VST are not supported is that neither one supports DSD and they commonly also have limitations for what PCM rates they can process (many don't support for example 352.8 or 768 kHz PCM). I want all my DSP features to work with any PCM and any DSD rate. I cannot accept to be limited to something like max 96 kHz PCM.

 

My modulators are specifically designed to avoid noise floor modulation. I just try to minimize amount of technical babble I put out about the DSP, so I don't talk about number of filter taps, noise floor modulation, and many many other things.

 

11 hours ago, buonassi said:

2) I've had very good experiences upsampling in A+ using iZotope and Sox feeding my metrum amethyst, but have no idea what type of interpolation they use (cubic, sinc, etc).   Do you know?

 

SoX is "sinc" and I'm pretty sure iZotope is too.

 

11 hours ago, buonassi said:

3) I have to believe that your filters, being processed in the computer, contribute to lower noise floor modulation in the DAC.  Would that be a correct assumption?  And the fact that they use many more taps than the Qutest may contribute to better reconstruction of the original waveform (at least theoretically)?

 

Number of taps is really meaningless number alone, you can have small number of "good taps" or high number of "bad taps". For example the "sinc-M" filter has >240 dB stop-band attenuation which is one of the important aspects. Filter design has many important aspects, many cannot be expressed by simple number but instead require entire mathematical formula to explain. Only way digital filter affects noise floor modulation is through intermodulation (or aliasing in case of badly done filter). Main contributor to noise floor modulation in SDM DACs (all modern DACs, including Chord) is through design of the modulator (noise-shaper). In R2R DACs like Metrum that are PCM and not SDM and thus don't utilize noise-shapers the only noise floor modulation path is intermodulation or aliasing in case of bad filter.

 

 

For me personally, main problem with current PCM content instead of DSD content is that it started it's life as SDM (DSD) and then was converted to low-rate high-bit PCM. Then again for playback it needs to be converted back. Using DSD content instead, like DSD256 recordings, you avoid all this back-and-forth shoveling.